Printer Friendly

Determination of sulpiride by capillary electrophoresis with end-column electrogenerated chemiluminescence detection.

Capillary electrophoresis (CE) [1] is an exceptional tool for performing rapid and highly efficient biochemical separations (1, 2). CE uses only picoliter to nanoliter sample volumes, and these small sample volumes require a very sensitive detection method. It is well known that laser-induced fluorescence detection, radiometric detection, and electrochemical detection are the most sensitive detection techniques available for CE, providing low detection limits. Although impressive analyses have been performed with available CE detection methods, there is still a need for new, highly sensitive CE detection methods. In addition to providing sensitivity, it would be desirable for new detection methods to be simpler and able to detect new types of analytes at trace concentrations. Chemiluminescence (CL) detection has been shown to be ideally suited for the challenging volume and detection limit requirements characteristic of CE separations. Various CL reagents and schemes have been reported for CE, using luminol (3, 4), peroxyoxalate (5, 6), and acridinium-based CL (7, 8). CL detection offers excellent detection limits because of the extremely low background noise. In addition, the instrumentation necessary to collect a CL signal is simple, consisting of a photomultiplier tube, a light-tight box, and minimal optics (9).

Electrogenerated CL (ECL) is the production of light by an oxidation or reduction reaction at an electrode surface. Compared with CL, ECL offers the ability to generate luminescence in a defined location on the surface of the electrode. Detection of photons emitted from an ECL reaction should provide an extremely sensitive means of detection for CE. Furthermore, the use of ECL should reduce interference from solution impurities and eliminate the need for complicated construction of postcolumn reactors for CL detection because the electrode replaces the catalyst added for CL (9).

One of the most efficient and thoroughly examined CL compounds is tris(2,2'-bipyridyl)ruthenium(II) [Ru[(bpy).sub.3.sup.2+]], which can emit light in both a CL and an ECL reaction. Ru[(bpy).sub.3.sup.2+] ECL is observed when Ru[(bpy).sub.3.sup.3+] reacts with Ru[(bpy).sub.3.sup.+] and yields an excited-state molecule, Ru[(bpy).sub.3.sup.2+] * (10). ECL emission can also be obtained when various oxidants and reductants react with the reduced or oxidized forms of Ru[(bpy).sub.3.sup.2+]; therefore, either the reductant or the oxidant can be treated as an analyte.

Ru[(bpy).sub.3.sup.2+] ECL has been widely used as a detection method for various amine-containing analyses in flowing streams, such as HPLC and flow injection analysis (11,12). Comparatively, the application of Ru[(bpy).sub.3.sup.2+] ECL detection in CE is very limited. To our knowledge, no more than 10 reports about CE-ECL have been published (13-19), and some of them suffer from poor separation efficiency, low sensitivity, and complicated instrumentation. In these reports, Ru[(bpy).sub.3.sup.2+] was added to the CE running buffer and converted to Ru[(bpy).sub.3.sup.3+] at the exit end of the capillary by use of a separate working electrode (13) or was added in a small reservoir at the interface between the separation capillary and the working electrode, which was then used to oxidize Ru[(bpy).sub.3.sup.2+] to Ru[(bpy).sub.3.sup.3+] (14,16,17). Additionally, Ru[(bpy).sub.3.sup.3+] has been generated in situ at the interface. In this approach, Ru[(bpy).sub.3.sup.2+] was transported to the interface by a syringe pump through a reaction tube, and the separation capillary was inserted into the reaction tube (15). Beta blockers, amino acids, and tripropylamine were used as analytes to characterize the CE-ECL systems. However, none of the above-mentioned studies explored the performance of CE-ECL for application to clinical analyses.

In our laboratory, CE-ECL detection has been applied to the analysis of biochemicals and pharmaceuticals. A Ru[(bpy).sub.3.sup.2+] ECL detection cell with end-column detection has been designed. In a separate experiment, we found that the detection limit for tripropylamine was 8 x [10.sup.-9] mol/L, which is four orders lower than that of the reported CE-ECL system for this compound (14). From these data, we concluded that the detector was very sensitive. In addition to sensitivity, the analysis system was compact and convenient. The inner diameter of separation capillary we used was 25 [micro]m, thus eliminating the need for on-column fractionation, which was used in the reported systems for separating CE current from the ECL detector (13-17). The diameter of the working electrode was 300 [micro]m, which was comparable to the outer diameter of the separation capillary; thus, alignment of the working electrode with the separation capillary outlet could be achieved easily under a microscope (x72 magnification).

Sulphide {5-(aminosulfonyl)-N-[(1-ethyl-2-pyrrolidinyl)methyl]-2-methoxybenzamide}, a selective dopamine [D.sub.2] antagonist with antipsychotic and antidepressant activity, is effective in the treatment of mental disorders, e.g., as a behavior regulator in the psychopathology of senescence, depression, and schizophrenia, and the frequency of extrapyramidal side effects is low (20, 21). The recommended oral dose of sulphide in the treatment of schizophrenia is 200-300 mg three times a day with a gradual increase to a maximum of 1200 mg daily (20). Various methods have been reported for the determination of sulphide; most of them were HPLC coupled to ultraviolet (22-25) or fluorescence detection (21, 26, 27). Gas chromatography with mass spectrometric detection (22, 28) and oscillopolarographic detection (29) methods have also been reported.

The goal of this work was to develop a CE separation method with end-column Ru[(bpy).sub.3.sup.2+] ECL detection for clinical analysis and to evaluate the performance of CEECL as a quantitative assay for determining sulphide concentrations in human plasma and urine. For this purpose, we used the newly designed Ru[(bpy).sub.3.sup.2+] ECL detection cell to detect sulphide after CE separation. The intra- and interday precision and accuracy of the assay were evaluated. This method was applied to the quantitative detection of sulphide in human body fluids, and seven patients were investigated for sulphide concentrations in plasma and urine 10 h after oral intake of 200 mg of sulphide.

In this experiment, the sulphide in the human plasma and urine samples was extracted by a double-step liquid-liquid extraction procedure using ethyl acetate-dichloromethane (5:1 by volume) as the extraction solvent.

Materials and Methods


All reagents and chemicals used were at least analytical reagent grade. Tris(2,2'-bipyridyl)ruthenium (II) chloride hexahydrate was purchased from Aldrich Chemical Co. Sulphide was purchased from Sigma Chemical Co. and was dissolved in 0.6 mol/L acetic acid at a concentration of 10 mmol/L for use as a stock solution. Before use, the stock solution was diluted to the desired concentrations. Sulphide tablets were obtained from Jinan Pharmaceutical Factory; each tablet contains 100 mg of sulphide. The buffers used in this experiment were sodium dihydrogen phosphate and disodium hydrogen phosphate. The buffer pH was adjusted with sodium hydroxide (98% purity; Sigma). D-Arginine, L-aspartic acid, L-glutamic acid, L-glycine, L-tyrosine, L-threonine, and L-valine were obtained from Shanghai Biochemical Co. Perphenazine, lidocaine hydrochloride, and megimide were purchased from Shanghai Zhaohui Pharmaceutical Factory. Haloperidol and phentolamine methanesulfonate were obtained from Shanghai Haipu Pharmaceutical Co. Promethazine hydrochloride was obtained from Shanghai Hefeng Pharmaceutical Co. Anisodamine hydrochloride and glyphylline were obtained from Shanghai No. 1 and Shanghai Xinyi Pharmaceutical Factory, respectively, and procaine hydrochloride was purchased from Beijing Yongkang Pharmaceutical Factory. All solutions were prepared with water purified by a Milli-Q system (Millipore) and stored at 4 [degrees]C in a refrigerator. Before use, all samples and buffer solutions were filtered through 0.22 [micro]m cellulose acetate filters (Shanghai Xinya Purification Material Factory). Drug-free human plasma and urine were obtained from Changchun blood bank (Changchun, China) and a healthy male volunteer, respectively, for use in constructing the calibration curve and preparing the samples for recovery studies.


To determine whether CE-ECL could give reliable results for identification and quantification of sulphide, plasma and urine samples were obtained from three male and four female patients newly admitted to Shandong Mental Health Center (Jinan, China), who were not on sulphide treatment until this study began. The diagnosis, according to the Chinese Classification and Diagnostic Criteria of Mental Disorders, 3rd edition (30), was schizophrenia in six patients and mental retardation in one patient. The patients' ages ranged from 19 to 43 years, and body weights were 45-68 kg. This study received approval from the Department of Medicine and Research Administration of Shangdong Mental Health Center, and patients enrolled in the study provided informed consent. The patients received 200 mg of sulphide (two 100-mg tablets) orally at night. Baseline blood and urine (~2 mL) samples were collected 0.5 h before oral intake of the drug, and blood and urine samples (~2 mL) were collected the next morning, 10 h after the dosing. Blood samples were collected in test tubes without anticoagulants and centrifuged at 1020g for 10 min; the plasma was then transferred to suitably labeled tubes. Urine samples were also collected in labeled tubes, and all the plasma and urine samples were stored at -20 [degrees]C until analysis.


Electrophoresis in the capillary was driven by a high-voltage power supply (CZE 1000R; Spellman). Uncoated fused-silica capillaries (25-[micro]m i.d.; 360-[micro]m o.d.) were purchased from Hebei Yongnian Optical Conductive Fiber Plant. A Model 800 Electrochemical Analyzer (CH Instruments) was used to provide potential for the oxidation of Ru[(bpy).sub.3.sup.2+] to Ru[(bpy).sub.3.sup.3+]. CL emission was detected by a Model BPCL Ultra-weak Luminescent Analyzer (Institute of Biophysics, Chinese Academy of Sciences, Beijing, China); the photomultiplier tube was operated in pulse mode and set at 900 V for detection.

A piece of capillary was cut to 50 cm in length and was used as separation capillary. An ~5-mm section of the polyimide coating was burned off at the distal end to allow light to pass. The capillary was placed between two buffer reservoirs. A high voltage was applied at the injection end, with the reservoir in the ECL detection cell held at ground potential.

ECL detection was carried out using a three-electrode system with a platinum disk working electrode, a 300-[micro]m diameter Ag/AgCl reference electrode (in saturated KCl solution), and a 1-mm diameter platinum wire auxiliary electrode. The working electrode was placed at the outlet of the separation capillary.

The detection cell and the detector (detailed below) were located in the light-tight box of the luminescent analyzer to prevent stray room light from contributing to background noise.


ECL detection at a constant potential with CE was performed using the end-column approach; a schematic diagram of the detection cell and light detection apparatus is given in Fig. 1. Axial alignment was achieved by adjusting the three nylon screws around the electrode. Once the electrode and capillary were axially aligned, the distance between the capillary and the electrode could be adjusted by adjusting the capillary holder under a microscope (x72 magnification). A distance of 70 [+ or -] 5 [micro]m between the capillary and electrode was found to be optimal. The reservoir (300 [micro]L) was refilled with Ru[(bpy).sub.3.sup.2+] solution and 50 mmol/L phosphate supporting electrolyte before each analysis. A piece of 1-mm thickness optical glass was mounted at the bottom of the reservoir; photons produced during the Ru[(bpy).sub.3.sup.2+] ECL reaction passed through the glass window and were detected by a photomultiplier tube.


The working electrode was constructed with a 300-[micro]m diameter platinum wire. Approximately 2.5 cm was cut and carefully soldered to a copper wire with soldering tin at the end of the electrode; it was then inserted into a polypropylene tube (1-mm i.d.; 3 mm-o.d.) until ~100 [micro]m of the platinum wire protruded from the tip of the tube. The tube was heated with continuous rolling above a micro flame alcohol burner until the plastic melted around and adhered tightly to the wire. Leakage of the electrolyte could be avoided by this technique. The electrode protruding from the plastic tube was cut and carefully polished with 1.0- and 0.3-[micro]m [alpha]-[Al.sub.2][O.sub.3] on a polishing cloth until a mirror-smooth surface was obtained. Before use, the platinum disk electrode was sonicated and rinsed with doubly distilled water for ~3 min.

After each analysis, the electrode was treated by cyclic voltammetry at a potential of -0.5 to 0 V and a scanning rate of 100 mV/s for 2 min. This treatment allowed the electrode to be used at least 1 week without any loss in activity for oxidizing Ru[(bpy).sub.3.sup.2+].



Cyclic voltammetry of 1 mmol/L Ru[(bpy).sub.3.sup.2+] and 5 mmol/L sulphide plus 50 mmol/L phosphate buffer (pH 8.0) was carried out in the same ECL detection cell with the same three-electrode system and electrochemical analyzer that were used in ECL detection.


A new fused-silica capillary was filled with 0.1 mol/L NaOH solution for 24 h before use and then flushed with filtered water and the corresponding running buffer for 15 min by means of a syringe. The running buffer in the experiment was 10 mmol/L phosphate solution. During the experiment, +15 kV (6 [micro]A) separation voltage was applied across the capillary, and the detection potential was applied at the working electrode. After the baseline CL signal reached a constant value (~10 min), electromigration injection was used for sample introduction, and the electropherogram was recorded. Between two electrophoretic runs, the capillary was rinsed with filtered water and running buffer for ~5 min, respectively. The number of theoretical plates (N), was calculated according to the equation:

N = 5.54(tm/ Wmz)z (1)

Where [t.sub.m] is the migration time, and [W.sub.1/2] is the width at half height of the electrophoretic peak.


We pipetted 100 [micro]L of human plasma or urine samples containing sulphide, enriched standards, and blank plasma or urine into 1.5-mL centrifuge tubes. The sample was alkalinized by adding 10 [micro]L of 1 mol/L NaOH solution; 1 mL of ethyl acetate-dichloromethane (5:1 by volume) was then added to the tube, and the tube was capped. The samples were vortex-mixed for 10 s in a Model G-560E vortex-mixer (Scientific Industries, INC.) with a 25 Hz frequency (5 mm amplitude), mixed at 120 rpm (20 mm amplitude) on a tube shaker (Shenzhen Guohua Co.) for 10 min, cooled to -20 [degrees]C for 10 min to break the emulsion formed during mixing, and then centrifuged at 17008 for 5 min. The top organic layer was transferred to a clean tube, and the process was repeated. The organic layer obtained this time was transferred to the same tube, and the combined organic layers were evaporated to dryness under a gentle stream of nitrogen in a water bath at 35 [degrees]C. Filtered water (100 [micro]L) was added to dissolve the residue, and the tube was vortex-mixed for 5 min at the same frequency as above. The sample was then injected into the CE-ECL system by electromigration.


Calibrators for sulphide were prepared by adding appropriate amounts of sulphide to a series of 100-[micro]L aliquots of drug-free human plasma or urine to achieve concentrations of 0.05, 0.1, 0.25, 0.5, 1.0, 2.5, 5.0, 10.0, and 25.0 [micro]mol/L, respectively; the samples were then extracted and analyzed using CE-ECL as described above. A calibration curve was obtained from the least-squares linear regression of the ECL peak intensities vs the concentrations of sulphide added. This curve was used to calculate sulphide concentrations in the unknown human plasma or urine samples based on the ECL peak intensities.

Two sets of samples, a control group and a recovery group, were prepared to determine the extraction recovery. The control samples were prepared as follows: 100-[micro]L aliquots of drug-free human plasma or urine were extracted by the liquid-liquid sample extraction procedure described above, and the residues were reconstituted with 100 [micro]L of 0.05, 0.1, 0.25, 0.5,1.0, 2.5, 5.0,10.0, or 25.0 [micro]mol/L sulphide solution, respectively. The recovery samples were prepared according to the procedure for preparing calibrators; the control group and recovery group samples were then analyzed by CE-ECL.

The absolute recovery was calculated as the ratio of measured ECL peak intensities for the recovery samples to that of the corresponding control samples at each sulphide concentration.


Several enriched human plasma or urine samples were tested for the presence of interfering compounds. The intra- and interday CVs and relative error of the mean were used to validate the precision and recovery of the assay, based on sulphide calibrators in plasma or urine. To assess intraday precision and recovery, we assayed five sets of controls at nine concentrations (0.05-25.0 [micro]mol/L) with one calibration curve in the same run. To assess interday precision and recovery, we evaluated five sets of control samples at nine different concentrations on 5 different days in 1 week (five calibration curves were prepared).



In the cyclic voltammograms, the Ru[(bpy).sub.3.sup.2+]/Ru[(bpy).sub.3.sup.3+] couple displayed reversible electrochemistry ([Delta]Ep = 59 mV) on the platinum disc electrode (Fig. 2, voltammogram III), whereas sulphide did not show an obvious electrochemical response in the potential range from 0 to +1.4 V (Fig. 2, voltammogram II). When cyclic voltammetry of Ru[(bpy).sub.3.sup.2+] was repeated in the presence of sulphide, the Ru[(bpy).sub.3.sup.3+] reduction disappeared and an electrochemical catalytic wave, produced by the reduction of Ru[(bpy).sub.3.sup.3+] by sulphide, was observed (Fig. 2, voltammogram IV).



The intensity of the emitted light is dependent on the rate of the light-emitting chemical reaction, and this reaction rate is dependent on the potential applied to the electrode (31). We evaluated the potential at which a maximum ECL signal was observed. Applied potentials of 1.0-1.3 V were explored, with 1.2 V producing the maximum ECL response. The ECL increased when the electrode potential was changed from 1.0 to 1.2 V and then decreased slightly after 1.2 V. We therefore set 1.2 V as the detection potential in our experiments.

As discussed above, there were roughly three modes for introducing Ru[(bpy).sub.3.sup.2+] at the electrode-capillary interface. In the end-column detection mode presented here, the reservoir is filled with Ru[(bpy).sub.3.sup.2+], and Ru[(bpy).sub.3.sup.3+] is generated on the surface of the working electrode. However, there is a strong flow of effluent from the electrophoresis capillary over the electrode, which may reduce the concentration of Ru[(bpy).sub.3.sup.3+], reducing the efficiency of light-producing reaction (9). In addition, only a fraction of the total amount of Ru[(bpy).sub.3.sup.2+] is converted to Ru[(bpy).sub.3.sup.3+] on the electrode surface. Thus, the electrogenerated Ru[(bpy).sub.3.sup.3+] is the limiting reagent for the ECL reaction (15, 32). The amount of Ru[(bpy).sub.3.sup.3+] at the capillaryelectrode interface could also limit the detection reaction. Thus, increasing the concentration of Ru[(bpy).sub.3.sup.2+] increases the concentration of Ru[(bpy).sub.3.sup.3+] and, thus, the ECL intensity (32). We found that the ECL intensity increased with increasing Ru[(bpy).sub.3.sup.2+] concentration from 0.2 to 5 mmol/L; we therefore selected a Ru[(bpy).sub.3.sup.2+] concentration of 5 mmol/L to achieve a high ECL signal.

Because the injection volume is one of the most important factors in the separation and detection, we observed the effect of injection voltage and time on the ECL intensity of sulphide to optimize the value of the injection voltage.

The ECL intensity increased, whereas N decreased, with increasing injection voltage over a range from 5 to 17.5 kV (Fig. 3A). The CL reaction occurs in the diffusion layer near the electrode when Ru[(bpy).sub.3.sup.2+] is oxidized to Ru[(bpy).sub.3.sup.3+] at the electrode surface and the oxidized Ru[(bpy).sub.3.sup.3+] contacts the analyte (32). Because the CL reaction rate is a function of the concentrations of species involved in the reaction and because emission intensity is dependent on the rate of the reaction (11), when more analyte in the effluent is present in the diffusion layer of the working electrode, a higher ECL signal can be obtained. On the other hand, when too much analyte gets into the diffusion layer, it frequently causes peak broadening. As a compromise, to obtain a high ECL signal and large N, we selected 10 kV as the optimal injection voltage. As illustrated in Fig. 3B, ECL intensity increases with increasing injection time, whereas N decreases. The reason for the change is similar to that for the effect of injection voltage (described above), and a 6-s injection time was chosen.


The effect of pH on the ECL intensity was investigated in a pH range from 4.5 to 10.0 in 0.5 pH units. With various amines, optimum pH values have been reported at pH 4-6 (33), and values remain essentially constant from pH 3 to 9 (34). In our present experiment, we observed that the ECL intensity for sulphide was dependent on pH. As shown in Fig. 4, at a pH range of 4.5-7.5, ECL intensity increased steadily and reached a maximum value of ~1500 at pH 8.0. At pH values >8.5, the ECL intensity decreased rapidly. A pH of 8.0 was chosen to obtain high ECL signals.


For human specimens, it has been reported that ethyl acetate-dichloromethane (5:1 by volume) is suitable as the extraction solvent for plasma because it produces minimal interference (27). In our experiments, we found that extraction with this solvent mixture provided maximum recoveries (95.6-101%) and minimum CVs for recovery (2.9-6.0%) and also produced clean extracts devoid of CE and ECL interference from the human plasma or urine. We therefore used ethyl acetate-dichloromethane (5:1 by volume) as the extraction solvent in this study.


Shown in Fig. 5 are typical electropherograms of the extracts from plasma and urine containing 10 [micro]mol/L sulphide under optimum experimental conditions: pH 8.0 phosphate buffer, injection for 6 s at 10 kV, and a detection potential of +1.2 V. Drug-free human plasma and urine showed no interference for sulphide.



To examine other potential interferences, we added the following substances and drugs to plasma or urine samples containing 10 [micro]mol/L sulphide: D-arginine (100 [micro]mol/L), L-aspartic acid (100 [micro]mol/L), L-glutamic acid (100 [micro]mol/L), L-glycine (100 [micro]mol/L), L-tyrosine (100 [micro]mol/L), L-threonine (100 [micro]mol/L), L-valine (100 [micro]mol/L), perphenazine (14 [micro]mol/L), lidocaine hydrochloride (10 [micro]mol/L), megimide (16 [micro]mol/L), haloperidol (13 [micro]mol/L), phentolamine methanesulfonate (26 [micro]mol/L), promethazine hydrochloride (10 [micro]mol/L), anisodamine hydrochloride (30 [micro]mol/L), glyphylline (40 F.i,mol/L), and procaine hydrochloride (10 [micro]mol/L). Approximately five additional peaks were observed within the 5-min migration time compared with the sample without these substances and drugs (Fig. 6). The sulphide peak was well resolved, and the recovery and migration time of sulphide were not affected.


Assay validation involved the estimation of intra- and interday imprecision and relative error in human plasma and urine. The results for plasma are presented in Table 1. The imprecision and relative error data for urine were similar and are not presented. The intra- and interday CVs were 2.9-5.3% and 4.3-6.0%,respectively. The intra-and interday relative errors of the mean were -4.38% to 1.78% and -3.38 to 1.61%, respectively.

The mean intraday migration time was 3.76 [+ or -] 0.02 min with a CV of 0.48%,and the mean interday migration time was 3.81 [+ or -] 0.03 min with a CV of 0.86%. All these data depicted the high precision of the analysis.



The sulphide calibration curve (Fig. 7) was linear (r >0.998) at 0.05-25.0 [micro]mol/L. The experimental detection limit, defined as the lowest concentration of sulphide that gave a signal-to-noise ratio 3, was 2.9 x [10.sup.-8] mol/L (9.9 [micro]g/L, or 0.029 [micro]mol/L). Thus, the assay is sufficiently sensitive and reliable for monitoring sulphide in human plasma or urine.


The method described here was applied to the determination of sulphide in plasma and urine from seven patients treated with sulphide. Sulphide was clearly detected in all samples examined, and no interfering peaks were observed from endogenous blank plasma or urine matrices (Fig. 8).


The regression equation for sulphide concentrations in patient plasma and urine samples, as measured by CEECL, was: y = 940.97x + 457.02, where x is the sulphide concentration ([micro]mol/L) and y is the ECL intensity (counts; see top panel in Fig. 7). The sulphide concentrations in the plasma samples from patients I-VII were 0.56, 1.00, 0.38, 0.43, 0.65, 0.44, and 0.58 [micro]mol/L, respectively, which were comparable to those obtained by HPLC [0.29-0.58 [micro]mol/L (100-200 [micro]g/L)] for the plasma samples collected at the same time (10 h) after the patients received the same oral dose (200 mg) (26). The discrepancy may be attributable to the variability of the bioavailability of sulphide among the patients. The urinary sulphide concentrations measured by CE-ECL were 6.71, 6.16, 8.98,10.35, 8.76,14.22, and 11.61 [micro]mol/L for patients I-VII, respectively.



CE is available as a clinical tool for measurement of serum protein and quantification of hemoglobin variants; for drug, vitamin, and isoenzyme analysis; and for analysis of metabolic disorders, porphyria, and DNA (35). The Ru[(bpy).sub.3.sup.2+] ECL is a well-established, sensitive detection technique for the determination of various amine-containing analytes without derivatization in flowing effluent streams, and the application of this method will be continue to grow in the biological and clinical sciences.

Sulphide possesses a chromophore and a fluorophore in its structure, which enable ultraviolet or fluorescent detection. Sulphide also has a tertiary amine group; it thus can be treated as an analyte when Ru[(bpy).sub.3.sup.2+] ECL detection is adopted. The electrochemical and ECL mechanisms for the reaction of Ru[(bpy).sub.3.sup.2+] with amines on different electrodes have been reported previously by other investigators (11, 36-38). In this study, when a platinum electrode was used and the amine was sulphide, we think that the mechanism was identical, based on the cyclic voltammograms (Fig. 2). The protonated sulphide radical can be formed by direct electrochemical oxidation of sulphide or by oxidation of Ru[(bpy).sub.3.sup.3+] (Fig. 2, voltammogram IV). The protonated sulphide radical then spontaneously loses a proton, forming a neutral sulphide radical. The main route for light emission is reduction of Ru[(bpy).sub.3.sup.3+] by the neutral sulphide radical to electronically excited Ru[(bpy).sub.3.sup.2+] *, which emits light when it relaxes to the ground state. According to this mechanism, our CE-ECL method can also be used to detect other amine-containing substances.

Sulphide is slowly and poorly absorbed from the gastrointestinal tract, with peak serum concentrations occurring within 2-6 h and 30% of an oral dose excreted unchanged in the urine in 48 h (20). Sulphide in human plasma or urine thus can be detected at an appropriate time after the administration of the drug. HPLC methods are routinely used for sulphide assays.

In this report we have presented, for the first time, the performance characteristics of CE-ECL as a quantitative method for the determination of sulphide in human plasma and urine and as a clinical assay for analyzing sulphide in patient plasma or urine samples. This CE-ECL method seems to offer several advantages compared with HPLC: it is rapid (run time <4 min), inexpensive (small buffer and reagent volumes; less-expensive fused-silica capillary and instrumentation), relatively robust, reproducible, and selective. We have provided optimal analytical conditions and data on reproducibility, linearity, and sensitivity. Using capillary zone electrophoresis separation with end-column Ru[(bpy).sub.3.sup.2+] ECL detection, we achieved good precision and high recoveries. Sulphide concentrations in patient plasma samples measured by CE-ECL correlated well with those determined by HPLC (26). The calibration curve was linear at 0.05-25.0 [micro]mol/L (r >0.998). This linear range was wider than those reported for HPLC methods: 14.7-293 [micro]mol/L (23), 0.029-4.4 [micro]mol/L (21), 0.029-2.9 [micro]mol/L (26), and 0.058-4.4 [micro]mol/L (27).

The detection limit for the CE-ECL method presented here is comparable to the detection limits reported for two fluorescence methods (2.9 x [10.sup.-8] mol/L) (25, 26), but lower than the detection limits reported for another fluorescence method (5.8 x [10.sup.-8] mol/L) (27), for a gas chromatographic-mass spectrometric method (7.31 x [10.sup.-8] mol/L) (22), for HPLC with ultraviolet detection [2.92 x [10.sup.-6] (23) and 3.5 x [10.sup.-7] mol/L (24)], and for a oscillopolarographic detection method (1 x [10.sup.-7] mol/L) (29) and is sensitive enough for clinical analysis. It should be noted that only 100 [micro]L of plasma or urine is necessary in the extraction procedure to achieve this low limit of detection and that this sample volume is much smaller than the volume needed for a HPLC method (>1 mL). Taking into consideration the low sample injection volume (3 nL), the mass limit of detection obtained by this CE-ECL method was only 8.7 x [10.sup.-17] mol, which is several orders of magnitude lower than the detection limit for the HPLC method, for which the sample injection volumes are ~100 [micro]L (21, 27).

In conclusion, we have developed and validated a new method combining CE with end-column Ru[(bpy).sub.3.sup.2+] ECL detection that is reproducible, reliable, and precise for the analysis of sulphide in human plasma or urine. In the clinical laboratory, this CE-ECL method may be of great value for drug monitoring, pharmacokinetic or bioavailability studies, and other clinical analysis.

We thank the staff of the Shandong Mental Health Center for obtaining patient samples. This project was supported by the National Natural Science Foundation of China.

Received October 11, 2001; accepted April 22, 2002.


(1.) Mikkers FEP, Everaerts FM, Verheggen TPEM. Concentration distributions in free zone electrophoresis. J Chromatogr 1979;169: 1-10.

(2.) Jorgenson JW, Lukacs KD. Zone electrophoresis in open-tubular glass capillary. Anal Chem 1981;53:1298-302.

(3.) Huang B, Li JJ, Zhang L, Cheng JK. On-line chemiluminescence detection for capillary ion analysis. Anal Chem 1996;68:2366-9.

(4.) Hashimoto M, Tsukagoshi K, Nakajima R, Kondo K. Compact detection cell using optical fiber for sensitization and simplification of capillary electrophoresis-chemiluminescence detection. J Chromatogr A 1999;832:191-202.

(5.) Shultz LL, Shippy S, Nieman TA, Sweedler JV. Peroxyoxalate chemiluminescence detection for capillary electrophoresis using membrane collection. J Microcolumn Sep 1998;10:329-37.

(6.) Tsukagoshi K, Okumura Y, Nakajima R. Migration behavior of dyestuff-containing liposomes in capillary electrophoresis with chemiluminescence detection. J Chromatogr A 1998;813:402-7.

(7.) Ruberto MA, Grayeski ML. Acridinium chemiluminescence detection with capillary electrophoresis. Anal Chem 1992;64:2758-62.

(8.) Ruberto MA, Grayeski ML. Investigation of acridinium labeling for chemiluminescence detection of peptides separated by capillary electrophoresis. J Microcolumn Sep 1994;6:545-50.

(9.) Gilman SD, Silverman CE, Ewing AG. Electrogenerated chemiluminescence detection for capillary electrophoresis. J Microcolumn Sep 1994;6:97-106.

(10.) Tokel NE, Bard AJ. Electrogenerated chemiluminescence IX. Electrochemistry and emission from systems containing tris(2,2'bipyridine)ruthenium (II) dichloride. J Am Chem Soc 1972;94:2862-3.

(11.) Lee W-Y. Tris(2,2'-bipyridyl)ruthenium (II) electrogenerated chemiluminescence in analytical science. Mikrochim Acta 1997;127:19-39.

(12.) Gerardi RD, Barnett NW, Lewis SW. Analytical application of tris(2,2'-bipyridyl)ruthenium (III) as a chemiluminescent reagent. Anal Chim Acta 1999;378:1-41.

(13.) Forbes GA, Nieman TA, Sweedler JV. On-line electrogenerated Ru[(bpy).sub.3.sup.3+] chemiluminescent detection of [beta]-blockers separated with capillary electrophoresis. Anal Chim Acta 1997;347:289-93.

(14.) Dickson JA, Ferris MM, Milofsky RE. Tris(2,2'-bipyridyl)ruthenium (III) as a chemiluminescent reagent for detection in capillary electrophoresis. J High Resol Chromatogr 1997;20:643-6.

(15.) Wang X, Bobbitt DR. In situ cell for electrochemically generated Ru[(bpy).sub.3.sup.3+]-based chemiluminescence detection in capillary electrophoresis. Anal Chim Acta 1999;383:213-20.

(16.) Hendrickson HP, Anderson P, Wang X, Pittman Z, Bobbitt DR. Compositional analysis of small peptides using capillary electrophoresis and [(bpy).sub.3.sup.3+]-based chemiluminescence detection. Microchem J 2000;65:189-95.

(17.) Bobbitt DR, Jackson WA, Hendrickson HP. Chemiluminescent detection of amines and amino acids using in situ generated Ru[(bpy).sub.3.sup.3+] following separation by capillary electrophoresis. Talanta 1998;46:565-72.

(18.) Wang X, Bobbitt DR. Electrochemically generated [(bpy).sub.3.sup.3+]-based chemiluminescence detection in micellar electrokinetic chromatography. Talanta 2000;53:337-45.

(19.) Arora A, Eijkel JCT, MorF WE, Manz A. A wireless electrochemiluminescence detector applied to direct and indirect detection for electrophoresis on a microfabricated glass device. Anal Chem 2001; 73:3282-8.

(20.) Tang G, ed. Modern pharmaceutics, 1st ed. Beijing: Chinese Medicine Technological Press, 1997:1036pp.

(21.) Tokunaga H, Kudo K, Jitsufuchi N, Ohtsuka Y, Imamura T. Sensitive determination of sulphide in human plasma by high-performance liquid chromatography. J Chromatogr B 1997;691: 203-7.

(22.) Rop PP, Sournac MH, Elie I, Burle J, Fornaris M. Toxicological analysis of sulphide in a lethal poisoning case. J Anal Toxicol 1999;23:294-6.

(23.) Yu X, Luo Z, Tang JS, Yu P. Determination of sulphide in human plasma by high performance liquid chromatography. Chin J Chromatogr 1997;15:546-7.

(24.) EI Walily AFM, EI Gindy A, Bedair MF. Application of first-derivative UV-spectrophotometry, TLC-densitometry and liquid chromatography for the simultaneous determination of mebeverine hydrochloride and sulphide. J Pharm Biomed Anal 1999;21:535-48.

(25.) Bressolle F, Bres J. Determination of sulphide and sultopride by high-performance liquid chromatography for pharmacokinetic studies. J Chromatogr 1985;341:391-9.

(26.) Nicolas P, Fauvelle F, Ennachachibi A, Merdjan H, Petitjean 0. Improved determination of sulpiride in plasma by ion-pair liquid chromatography with fluorescence detection. J Chromatogr 1986; 381:393-400.

(27.) Huang M-C, Ho H-O, Yeh G-C, Ke W-T, Lin L-C, Bruce Hsu T-M, et al. Development of a high-performance liquid chromatographic method for bioanalytical applications with sulpiride. J Chromatogr B 2001;763:157-63.

(28.) Frigerio A, Pantarotto C. Quantitative determination of sulpiride in biological samples from rats by gas-liquid chromatography and chemical ionization-mass fragmentography. J Chromatogr 1977; 130:361-7.

(29.) Zeng YH, Song Q. Oscillopolarographic determination of sulpiride and its electrochemical behavior. Chin J Anal Chem 1997;25:56-8.

(30.) Chinese Psychiatric Association. Chinese classification and diagnostic criteria of mental disorders, 3rd ed. Jian, China: Shandong Scientific and Technological Press, 2001:344pp.

(31.) Haapakka KE, Kankare JJ. The mechanism of the electrogenerated chemiluminescence of luminol in aqueous alkaline solution. Anal Chim Acta 1982;138:263-75.

(32.) Lee W-Y, Nieman TA. Evaluation of use of tris(2,2'-bipyridyl)ruthenium (III) as a chemiluminescent reagent for quantitation in flowing streams. Anal Chem 1995;67:1789-96.

(33.) Noffsinger JB, Danielson ND. Generation of chemiluminescence upon reaction of aliphatic amines with tris(2,2'-bipyridyl)ruthenium (III). Anal Chem 1987;59:865-8.

(34.) Downey TM, Nieman TA. Chemiluminescence detection using regenerable tris(2,2'-bipyridyl)ruthenium (II) immobilized in Nafion. Anal Chem 1992;64:261-8.

(35.) Jenkins, MA, Guerin MD. Capillary electrophoresis as a clinical tool. J Chromatogr B 1996;682:23-34.

(36.) Leland JK, Powell MJ. Electrogenerated chemiluminescence-an oxidative-reduction type ECL reaction sequence using tripropylamine. J Electrochem Soc 1990;37:3127-31.

(37.) Knight AW, Greenway GM. Relationship between structure attributes and observed electrogenerated chemiluminescence (ECL) activity of tertiary amines as potential analytes for the tris(2,2'-bipyridine)ruthenium (II) ECL reaction. Analyst 1996;121:101R-6R.

(38.) Wilson R, Akhavan-Tafti H, DeSilva, R, Schaap AP. Comparison between acridan ester, luminol, and ruthenium chelate electrochemiluminescence. Electroanalysis 2001;13:1083-92.

[1] Nonstandard abbreviations: CE, capillary electrophoresis; CL, chemiluminescence; ECL, electrogenerated CL; and Ru[(bpy).sub.3.sup.2+], tris(2,2'-bipyridyl)ruthenium(II).


State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin 130022, People's Republic of China.

* Author for correspondence. Fax 86-431-5689711; e-mail
Table 1. Intra- and interday precision of the
sulpiride assay. (a)


Amount added, Amount measured CV, % Relative
[micro]m?l/L (mean [+ or -] SD), error, %
(n = 5) [micro]m?l/L

0.1 0.0478 [+ or -] 0.0025 5.3 -4.38
0.1 0.0962 [+ or -] 0.0050 5.2 -3.81
0.3 0.241 [+ or -] 0.012 4.8 -3.41
0.5 0.491 [+ or -] 0.023 4.6 -1.88
1.0 1.01 [+ or -] 0.05 4.5 1.10
2.5 2.48 [+ or -] 0.11 4.3 -0.77
5.0 4.99 [+ or -] 0.21 4.2 -0.28
10.0 10.1 [+ or -] 0.3 3.5 0.65
25.0 25.4 [+ or -] ?.7 2.9 1 .78


Amount added, Amount measured CV, % Relative
[micro]m?l/L (mean [+ or -] SD), error, %
(n = 5) [micro]m?l/L

0.1 0.0483 [+ or -] 0.0029 6.0 -3.38
0.1 0.0988 [+ or -] 0.0059 5.9 -1.17
0.3 0.253 [+ or -] 0.015 5.8 1.18
0.5 0.505 [+ or -] 0.028 5.5 1.05
1.0 0.991 [+ or -] 0.052 5.2 -0.86
2.5 2.53 [+ or -] 0.12 5.0 1.05
5.0 5.06 [+ or -] 0.25 4.9 1.12
10.0 9.93 [+ or -] 0.43 4.3 -0.68
25.0 25.4 [+ or -] 1 .1 4.3 1.61

(a) Results were obtained from plasma validation; the
imprecision and relative error data in urine were similar
and are not presented.
COPYRIGHT 2002 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

 Reader Opinion




Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Drug Monitoring and Toxicology
Author:Liu, Jifeng; Cao, Weidong; Qiu, Haibo; Sun, Xiuhua; Yang, Xiurong; Wang, Erkang
Publication:Clinical Chemistry
Article Type:Clinical report
Date:Jul 1, 2002
Previous Article:Relationship between genetic polymorphisms of alcohol-metabolizing enzymes and changes in risk factors for coronary heart disease associated with...
Next Article:Evaluation of a no-pretreatment cyclosporin a assay on the dade behring dimension RxL clinical chemistry analyzer.

Related Articles
Capillary electrophoresis and the pharmaceutical industry: a new era.
SpectruMedix LLC Launches Mutation Discovery and SNP Genotyping Service; Company will offer SNP discovery and genotyping services based on the...
Determination of gabapentin in plasma by liquid chromatography with fluorescence detection after solid-phase extraction with a [C.sub.18] column.
Clinical analysis by microchip capillary electrophoresis.
Effect of sulfamethoxazole on clinical capillary zone electrophoresis of serum proteins.
Technique to remove interference caused by Radio-opaque agents in clinical capillary zone electrophoresis.
Caveats in carbohydrate-deficient transferrin determination.
Capillary zone electrophoresis for the diagnosis of congenital hemoglobinopathies.
Analysis of cyanine dye-labeled PCR product and restriction fragments by capillary electrophoresis and laser-induced fluorescence.
Diagnosis of Duchenne/Becker muscular dystrophy and quantitative identification of carrier status by use of entangled solution capillary...

Terms of use | Copyright © 2015 Farlex, Inc. | Feedback | For webmasters