Electrochemiluminescence from successive electro- and chemo-oxidation of rifampicin and its application to the determination of rifampicin in pharmaceutical preparations and human urine

Spectrochimica Acta Part A 67 (2007) 430–436

Electrochemiluminescence from successive electro- and chemo-oxidation of rifampicin and its application to the determination of rifampicin in pharmaceutical preparations and human urine Yao-Dong Liang a,b , Jun-Feng Song a,∗ , Min Xu a a

b

Institute of Analytical Science, Northwest University, Xi’an 710069, China Department of Chemistry and Chemical Engineering, Xi’an University of Science and Technology, Xi’an 710054, China Received 18 March 2006; received in revised form 26 July 2006; accepted 31 July 2006

Abstract A novel electrochemiluminescence (ECL) type was proposed based on successive electro- and chemo-oxidation of oxidable analyte, which was different from both annihilation and coreactant ECL types in mechanism. Rifampicin was used as a model compound. No any chemiluminescence (CL) was produced by either electrochemical oxidation or chemical oxidation of rifampicin in KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–dodecyl trimethyl ammonium chloride (DTAC) solution. However, an ECL was observed by electrochemical oxidization of rifampicin in the same solution in the presence of oxidant such as dissolved oxygen, activated oxygen and potassium peroxydisulfate (K2 S2 O8 ). The ECL was attributed to electrochemical oxidation of rifampicin to form semiquinone free radical, and then subsequently chemical oxidation of the formed radical by oxidant to form excited state rifampicin quinone. The proposed ECL type introduced additional advantages such as high selectivity, simple and convenient operation, and effective avoidance of side reaction that often took place in homogenous CL reaction, and will open a novel application field. In addition, with the ECL in the presence of K2 S2 O8 as oxidant, a flow injection ECL method for the determination of rifampicin was proposed. The ECL intensity was linear with rifampicin concentration in the range of 1.0 × 10−7 to 4.0 × 10−5 mol l−1 and the limit of detection (s/n = 3) was 3.9 × 10−8 mol l−1 . The proposed method was applied to the determination of rifampicin in pharmaceutical preparations and human urine. © 2006 Elsevier B.V. All rights reserved. Keywords: Rifampicin; Electrochemiluminescence (ECL); Flow injection

1. Introduction Electrochemiluminescence (ECL) had emerged as a useful analytical technique in which a luminescence was produced in the vicinity of an electrode surface when an appropriate electrochemical manner was used [1–3]. Compared with chemiluminescence (CL), ECL introduced two main advantages [4,5]. First, unstable CL reagents and intermediates can be generated and allowed to react in situ as soon as they are formed at the electrode surface. Second, because the reaction can be controlled and manipulated by employing electrochemical manner, it can be used for the selective determination of some substances without separation. So far, based on reaction mechanisms, general ECL has been classified into two types [3–5]. The first type of ECL is annihilation ECL including two kinds. One is based

∗

Corresponding author. Tel.: +86 29 88302077; fax: +86 29 88302573. E-mail address: [email protected] (J.-F. Song).

1386-1425/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2006.07.036

on the annihilation reaction between electro-generated oxidized and reduced forms of metal complexes or clusters to form the excited state that was able to emit light. The typical example was Ru(2,2 -bipyridine)3 3+ /Ru(2,2 -bipyridine)3 + system. The other is based on the annihilation reaction between the radical cation (A•− ) of one species and the radical anion (D•+ ) of same or different species. The second type is coreactant ECL, which involves the formation of an intermediate by electrochemical oxidation or reduction of a coreactant, and subsequent reaction between the intermediate and ECL luminophore to produce excited states. The typical examples are tertiary amine-Ru(2,2 bipyridine)3 2+ and peroxydisulfate-Ru(2,2 -bipyridine)3 2+ systems [6,7]. The coreactant must be easily electrochemically oxidized or reduced, and the intermediate, its oxidized or reduced product, should possess stronger reducing or oxidizing ability. In addition, great attention has been paid to luminol ECL in the literature [8]. In luminol ECL, luminol is electrochemically oxidized to be luminol radical, and the formed radical is then chemically oxidized to produce excited state 3-aminophtalate by

Y.-D. Liang et al. / Spectrochimica Acta Part A 67 (2007) 430–436

431

oxidant such as hydrogen peroxide, superoxide radical and hypobromite that is electrogenerated or added intentionally [9,10]. Besides luminol ECL systems, to best of our knowledge, only an example based on successive electro- and chemo-oxidation ECL, the ECL of indole and tryptophan in the presence of hydrogen peroxide as oxidant, has been reported [11]. The ECL type of luminol, indole and tryptophan is completely different from the general ECL types mentioned above in mechanism, and may be called as successive electro- and chemo-oxidation ECL. Although few examples of the proposed ECL type are reported, the proposed ECL phenomena may occur greatly. Thus, it is of significance to study the proposed ECL type and enlarge its application. Rifampicin, a semi-synthetic antibiotic drug, is widely used alone or in combination with other drugs such as isoniazid and pyrazinamide in the treatment of tuberculosis and other infectious diseases [12–14]. It is a hydroquinone derivative that contains three phenolic hydroxyl groups, which can be oxidized by both chemical and electrochemical methods [15–18]. Moreover, CL of rifampicin is observed based on chemical oxidation by such oxidant as N-bromosuccinimide, hexacyanoferrate and hydroxyl radical in alkaline solution [16–18]. Thus, rifampicin is selected as a model analyte to illustrate the proposed type of ECL. In this work, electrochemical oxidation of rifampicin formed an intermediate, and the formed intermediate was subsequently chemically oxidized to produce ECL. The ECL mechanisms were discussed in detail. Moreover, a flow injection ECL for the determination of rifampicin was proposed. This method had been evaluated by the analysis of rifampicin in pharmaceutical preparations and human urine.

injection valve, two Y-shaped mixing elements and a photomultiplier tube (PMT). PTFE tubing (0.8 mm i.d.) was used to connect all components in the flow system. Cyclic voltammograms were recorded on a model CHI660 electrochemical workstation (CH Instruments, USA). A commercial cylindroid glass cell was used as batch ECL cell and employed with a threeelectrode system consisting of a platinum coil (30 cm × 0.5 mm i.d.) as working electrode, a platinum wire as auxiliary electrode, and an Ag/AgCl/sat. KCl as reference electrode. ECL flow-through electrolysis cell, which was homemade, consisted of a glass coil (40 cm × 1.0 mm i.d.) with two-electrode system. Two platinum wires (30 cm × 0.5 mm i.d. and 2 cm × 0.5 mm i.d.), directly inserted from the inlet and the outlet of the glass tube, were used as working and auxiliary electrodes, respectively. The end distance between the two platinum electrodes that was inserted into the glass coil was about 1 cm. The constant direct current applied for electrolysis was achieved with a galvanostat (Model KLT-1 coulombmeter) (Jiangshu Electroanalysis Instrument Plant, China). The fluorescence spectra were monitored using Model RF-540 fluorescence spectrometer (Shimadzu, Japan).

2. Experimental

2.3.2. Procedure for ECL kinetic profile and ECL spectra The batch ECL cell with a three-electrode system was placed in front of the PMT. When the potential of 1.3 V was applied at the Pt working electrode, the ECL kinetic profile was recorded. The ECL spectra were achieved with a set of 11 narrow band interference filters (400–680 nm). The filters were set between the batch ECL cell and the PMT. When the potential of 1.3 V was applied at the Pt working electrode, the ECL signal was recorded at different wavelength bands.

2.1. Chemicals All reagents were of analytical reagent-grade. Twice glassdistilled water was used throughout the experiments. Rifampicin was of biochemical-reagent grade and was purchased from National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). The stock standard solution of rifampicin (1.00 × 10−3 mol l−1 ) was prepared by dissolving in methanol and was kept in the refrigerator at 4 ◦ C. Standard working solutions were prepared daily by appropriate dilution of the stock solution with water. 0.09 mol l−1 KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer, 0.10 mol l−1 K2 S2 O8 and 5.0 × 10−3 mol l−1 dodecyl trimethyl ammonium chloride (DTAC) solutions were routinely prepared. Rifampicin eye drop (Xi’an Pharmaceutical Plant, China) was purchased from local hospital. Urine samples were collected from three healthy individuals (from the Hospital of Northwest University, Xi’an). 2.2. Apparatus The ECL intensity was recorded by a Model IFFM-D-FICL analysis system (Xi’an Remax Electronic Science-Tech Co. Ltd., China). It consisted of two peristaltic pumps, a six-way

2.3. Procedure 2.3.1. Procedure for ECL profiles versus potential and cyclic voltammograms The batch ECL cell with a three-electrode system mentioned above was placed in front of the PMT. When the potential of the working electrode was scanned from 0 to 1.5 V at scan rate of 50 mV s−1 , ECL profiles versus potential and corresponding cyclic voltammograms were recorded simultaneously.

2.3.3. Procedure for flow injection ECL experiment The flow injection ECL system used in this work was shown in Fig. 1. By keeping the valve in washing position, KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer and K2 S2 O8 solution were continuously pumped into the manifold until the baseline was established on the recorder. Then 120 ␮l of a mixture of sample and DTAC solutions was injected into the KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer, and at the same time 10 mA direct current was applied. The buffer was then merged with K2 S2 O8 solution in the second Y-shaped mixing element (Y2 ) before the ECL flowthrough electrolysis cell. When the mixed solution flowed into the ECL flow-through electrolysis cell, ECL reaction occurred. The ECL signal produced was recorded. Calibration graphs were constructed by plotting the intensity (peak height) of the ECL signal versus the concentration of rifampicin (Scheme 1).

432

Y.-D. Liang et al. / Spectrochimica Acta Part A 67 (2007) 430–436

Fig. 1. Schematic diagram of the flow injection ECL manifold used for the determination of rifampicin: P1 and P2 , peristaltic pump; Y1 and Y2 , Y-shaped mixing element; V, six-way valve; F, ECL flow-through electrolysis cell; W, waste; WE, working electrode; AE, auxiliary electrode; NHV, negative high voltage; PMT, photomultiplier tube; R, computer; I, Galvanostat; (a) sample solution; (b) DTAC solution stream; (c) KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer; (d) K2 S2 O8 solution stream.

Scheme 1. Molecular structure of rifampicin.

3. Results and discussion 3.1. Characteristics of ECL The ECL characteristics of rifampicin in KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–DTAC solution were investigated in three cases: electrochemical oxidation, chemical oxidation and electrochemical oxidation in the presence of various oxidants. On the one hand, oxidation of rifampicin was carried out by electrochemical method. Only one fluorescence peak (λmax = 448 nm) in the range of 400–700 nm was observed in rifampicin–KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–DTAC solu-

tion, which was from rifampicin itself [19]. When an applied potential was scanned from 0 to 1.0 V (versus Ag/AgCl/sat. KCl) in the solution mentioned above after deoxygenated with pure nitrogen gas, only an irreversible oxidation peak of rifampicin appeared at 0.6 V while no any ECL peak appeared, and the fluorescence spectrum was no changed while the fluorescence peak intensity at 448 nm decreased slightly. It was reported that, in the absence of dissolved oxygen, electrochemical oxidation of phenolic hydroxyl group of rifampicin via one-electron transfer produced rifampicin semiquinone radical (Scheme 2(I)) [20–22]. The formed rifampicin semiquinone radical, a high active species, quickly performed a disproportionation reaction to give ground state rifampicin and ground state rifampicin quinone (Scheme 2(II)) [23–25]. As a result, no any CL was observed. On the other hand, chemical method was also used to oxidize rifampicin. When K2 S2 O8 was added into the same solution mentioned above, no any CL was produced as well, and the fluorescence spectrum and the peak intensity at 448 nm was almost the same as that without the addition of K2 S2 O8 , which was due to slow chemical oxidation of rifampicin by K2 S2 O8 in neutral aqueous solution [26]. These results revealed that no any CL was produced only by either electrochemical or chemical oxidation of rifampicin. When rifampicin was electrochemically oxidized in the presence of oxidant, however, the CL characteristics of rifampicin were different from that alone with electrochemical and chemical oxidation mentioned above. When the applied potential was still scanned from 0 to 1.0 V in the rifampicin–KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–DTAC solution without deoxygenation, a weak ECL was just observed (Fig. 2A(c)), and the irreversible oxidation peak current of rifampicin at 0.6 V was only a little decrease compared with that after deoxygenation (Fig. 2B(c)). Both the decrease of the oxidation peak current of rifampicin and the appearance of the weak ECL were related with dissolved oxygen. Previous studies suggested that, in the presence of dissolved oxygen, few of the formed rifampicin semiquinone radicals were chemically oxidized while most of them still performed the disproportionation reaction [27,28]. As no CL was produced during the disproportionation reaction, the weak ECL should be from excited state rifampicin quinone, chemical oxidation product of rifampicin semiquinone radical [16,29]. Further, when the applied potential was scanned from 1.0 to 1.5 V, an anodic current appeared (Fig. 2B(c)), producing activated oxygen by

Scheme 2. Molecular structure of rifampicin semiquinone radical (I) and rifampicin quinone (II).

Y.-D. Liang et al. / Spectrochimica Acta Part A 67 (2007) 430–436

Fig. 2. ECL profiles (A) and cyclic voltammograms (B) vs. potential in 0.09 mol l−1 KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–5.0 × 10−3 mol l−1 DTAC solution (a) and 1.0 × 10−5 mol l−1 rifampicin–0.09 mol l−1 KH2 PO4 – Na2 B4 O7 (pH 6.6) buffer–5.0 × 10−3 mol l−1 DTAC solution in the presence (b) and the absence of 0.10 mol l−1 K2 S2 O8 (c) without deoxygenation.

electrochemical oxidation of water [30,31], and the weak ECL was enhanced and the enhanced ECL reached a relative larger value at 1.3 V (Fig. 2A(c)). The fluorescence spectrum was still unchanged and only the intrinsic peak intensity (λmax = 448 nm) decreased compared with that before the potential was scanned from 0 to 1.5 V. These results indicated that activated oxygen involved in the enhanced ECL. It was well known that activated oxygen possessed much stronger oxidizing ability than dissolved oxygen. As soon as activated oxygen was produced, more rifampicin semiquinone radicals were chemically oxidized [31,32]. In this case, obviously, the emitting species in the enhanced ECL should also be excited state rifampicin quinone. When the potential was still scanned from 0 to 1.5 V in the same rifampicin–KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–DTAC solution in the presence of K2 S2 O8 that did not produce any oxidation and reduction peaks in the potential range examined, the oxidation peak current of rifampicin at 0.6 V decreased while the enhanced ECL intensity at 1.3 V increased and rose gradually with the increase of the concentration of K2 S2 O8 from 0 to 0.10 mol l−1 . When the concentration of K2 S2 O8 achieved 0.10 mol l−1 , the oxidation peak current at 0.6 V decreased approximately by 54% compared with that without the addition of K2 S2 O8 and the enhanced ECL intensity at 1.3 V reached a maximum value (Fig. 2B(b)). And the shape of the ECL profile in the presence of K2 S2 O8 (Fig. 2A(b)) was at all times the same as that in the absence of K2 S2 O8 (Fig. 2A(c)). At the same time, the fluorescence spectrum in the presence of K2 S2 O8 was the same as that in the absence of K2 S2 O8 , while the peak intensity (λmax = 448 nm) deceased much more sharply in the presence of K2 S2 O8 than that in the absence of K2 S2 O8 . These resulted indicated that almost all of electrogenerated rifampicin semiquinone radicals were chemically oxidized by K2 S2 O8 to form excited state rifampicin quinone, resulting in much stronger ECL. From these facts, therefore, it is obtained that:

433

Fig. 3. ECL dynamic response curves of 0.09 mol l−1 KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–5.0 × 10−3 mol l−1 DTAC solution (a), 1.0 × 10−6 mol l−1 rifampicin–0.09 mol l−1 KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–5.0 × 10−3 mol l−1 DTAC solution using electro-generated activated oxygen (c) and 0.10 mol l−1 K2 S2 O8 (b) as oxidant.

(1) Electrochemical oxidation of rifampicin in the presence of oxidant produced ECL, while chemical oxidation of rifampicin did not produce any CL. This result demonstrated that the first reaction for the ECL was the conversion of rifampicin to rifampicin semiquinone radical via electrooxidation. In order to confirm the existence of rifampicin semiquinone radical, three semiquinone radical scavengers [33,34], mannitol, ethanol and dimethyl sulfoxide, were separately used. When the potential was controlled at 1.3 V in the rifampicin–KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–DTAC solution in the absence and the presence of K2 S2 O8 , the enhanced ECL signals using activated oxygen or K2 S2 O8 as oxidant (Fig. 3c and b) decreased sharply after adding semiquinone radical scavengers (Table 1). (2) The enhanced ECL spectra using activated oxygen or K2 S2 O8 as oxidant showed one ECL peak band at 460–575 nm (Fig. 4a and b), which were in good agreement with the CL spectra of excited state quinone (λmax = 485–530 nm) that was produced by chemical oxidization of semiquinone radical by dissolved oxygen or hydroxyl radical [35,36]. These evidences confirmed that the emitting species in the proposed enhanced ECL was Table 1 The effect of semiquinone radical scavengers, 5.0 × 10−2 mol l−1 mannitol, 5.0 × 10−2 mol l−1 ethanol and 5.0 × 10−2 mol l−1 dimethyl sulfoxide on the enhanced ECL of 1.0 × 10−5 mol l−1 rifampicin in 0.09 mol l−1 KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–5.0 × 10−3 mol l−1 DTAC solution Semiquinone radical scavenger

The enhanced ECL relative intensity using activated oxygen as oxidanta

The enhanced ECL relative intensity using K2 S2 O8 as oxidantb

None Mannitol Ethanol Dimethyl sulfoxide

100 39 62 34

1004 343 426 258

a b

Using electro-generated activated oxygen as oxidant. Using 0.10 mol l−1 K2 S2 O8 as oxidant.

434

Y.-D. Liang et al. / Spectrochimica Acta Part A 67 (2007) 430–436

lytical conditions optimized included direct current, selection of oxidant, carrier and surfactant solutions, and the flow rate for the proposed flow injection ECL system. 3.2.1. Effect of direct current on the ECL The current-controlled electrolysis is used as it is thought to be simple in instrument. The effect of direct current on the ECL intensity of 1.0 × 10−6 mol l−1 rifampicin was examined over the range of 2.0–50.0 mA. The ECL intensity increased sharply as the direct current raised from 2.0 to 8.0 mA. With the direct current increasing from 8.0 to 50.0 mA, the ECL intensity reached a maximum value and almost kept constant. Therefore, 10 mA direct current was used throughout the experiments.

Fig. 4. The enhanced ECL spectra of 1.0 × 10−6 mol l−1 rifampicin in 0.09 mol l−1 KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–5.0 × 10−3 mol l−1 DTAC solution using electro-generated activated oxygen (a) and 0.10 mol l−1 K2 S2 O8 (b) as oxidant.

excited state rifampicin quinone [37]. Additionally, the ECL intensity of rifampicin increased with the increasing order of the oxidizing ability of the oxidant, dissolved oxygen < activated oxygen < K2 S2 O8 , while both oxidation peak current at 0.6 V and fluorescence peak intensity at 448 nm decreased with the same order [26]. The relativity of the ECL intensity with the oxidizing ability of the oxidant indicated that the same species, electrogenerated rifampicin semiquinone radical, was chemically oxidized to form excited state rifampicin quinone. Therefore, the subsequent reaction for the ECL was the conversion of the formed rifampicin semiquinone radical to excited state rifampicin quinone via chemical oxidation. From these experimental results mentioned above, the ECL of rifampicin can be described as follows: initially, electrochemical oxidation of rifampicin in KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer–DTAC solution formed rifampicin semiquinone radical. Then, the radical was further chemically oxidized by dissolved oxygen, activated oxygen ([O]) or K2 S2 O8 to form excited state rifampicin quinone. When the excited stat rifampicin quinone went back to its ground state, the weak or the enhanced ECL occurred. In its simple form, the ECL mentioned above was attributed to the following reactions: −e−

rifampicin −→rifampicin semiquinone radical, O2 , [O] or K2 S2 O8

rifampicin semiquinone radical −−−−−−−−−−−→ rifampicin quinone∗ , rifampicin quinone∗ → rifampicin quinone + hν 3.2. Optimization of experimental variables Based on the enhanced ECL of rifampicin, a flow injection ECL for the determination of rifampicin was proposed. The ana-

3.2.2. Selection of oxidant The effect of common oxidants, including potassium periodate, hydrogen peroxide, potassium permanganate, sodium bromate, potassium hexacyanoferrate and K2 S2 O8 , were investigated. It was found that the ECL was observed when the oxidants mentioned above were used. Among them, K2 S2 O8 possessed the strongest oxidizing ability [26], and thus gave the maximum ECL signal. The effect of K2 S2 O8 concentration on the ECL intensity was examined by using 1.0 × 10−6 mol l−1 rifampicin. The ECL intensity rose sharply as the K2 S2 O8 concentration increased from 1.0 × 10−4 to 0.08 mol l−1 . When the K2 S2 O8 concentration increased from 0.08 to 0.20 mol l−1 , the ECL intensity reached a maximum value and kept constant. Therefore, 0.10 mol l−1 K2 S2 O8 was used. 3.2.3. Selection of carrier solution Different carrier solutions such as H2 SO4 , HCl, HClO4 , KH2 PO4 –Na2 HPO4 , HAc–NaAc, KH2 PO4 –Na2 B4 O7 , NH3 ·H2 O–NH4 Cl, and Na2 SO4 solution were examined in present ECL system. Experiment showed that KH2 PO4 –Na2 B4 O7 buffer gave the maximum ECL signal, and also the best reproducibility for monitoring rifampicin. So KH2 PO4 –Na2 B4 O7 buffer was selected. The effect of pH value in the range of 5.8–9.2 on the ECL was examined in the KH2 PO4 –Na2 B4 O7 buffer. The maximum of the ECL intensity was observed at pH 6.6. The effect of the total concentration of KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer was also tested in the range of 0.015–0.18 mol l−1 . Excellent sensitivity and reproducibility were obtained when the total concentration of the buffer was higher than 0.06 mol l−1 . Accordingly, the 0.09 mol l−1 KH2 PO4 –Na2 B4 O7 (pH 6.6) buffer was used. 3.2.4. Selection of surfactant Some surfactants, including two neutral surfactants (Tween 80, Tween 20), three cationic surfactants (tetramethyl ammonium chloride, tetrabutyl ammonium chloride, DTAC), and one anionic surfactant (sodium dodecyl sulfate), were examined. It was found that all the surfactants enhanced the ECL of rifampicin. The surfactant media not only led to an increase of rifampicin concentration at the surface of electrode by adsorption [38], but also prolonged the lifespan of electrogenerated rifampicin semiquinone radical [39–41], which facilitated

Y.-D. Liang et al. / Spectrochimica Acta Part A 67 (2007) 430–436

more rifampicin semiquinone radical to be chemically oxidized and led to higher ECL intensity [42–44]. Experiment also showed that DTAC gave the maximum ECL signal, and the best reproducibility for monitoring rifampicin. So DTAC was selected. The effect of DTAC concentration on the ECL intensity was examined in the range of 0 to 10−2 mol l−1 . The results showed that the ECL signal reached its maximum value at 5.0 × 10−3 mol l−1 . Thus, 5.0 × 10−3 mol l−1 DTAC was used. 3.2.5. Effect of flow rate Pump P2 was used to deliver KH2 PO4 –Na2 B4 O7 buffer that contained sample. The effect of its flow rate on the ECL intensity was investigated in the range of 0.3–5.4 ml min−1 . The results showed that the ECL intensity was increased sharply with the increase of the flow rate in the range of 0.3–2.2 ml min−1 , and reached maximum value at 2.2 ml min−1 . When the flow rate was higher than 2.2 ml min−1 , the ECL intensity decreased because higher flow rate led to a decrease of electrolysis efficiency [45]. Thus, 2.2 ml min−1 was used. 3.3. Performance of the proposed method for rifampicin measurements Under the selected experimental conditions, the ECL intensity was linear with rifampicin concentration in the range of 1.0 × 10−7 to 4.0 × 10−5 mol l−1 . The detection limit was 3.9 × 10−8 mol l−1 (s/n = 3) and the relative standard deviation for 1.0 × 10−6 mol l−1 rifampicin (n = 9) was 1.3%. The linear regression equation was I = 7.48 + 3.47 × 108 C (where I is the ECL intensity and C is the rifampicin concentration, units are mV and mol l−1 , respectively) with a correlation coefficient of 0.9996 (n = 13). The sample measurement frequency was calculated to be about 30 samples h−1 . 3.4. Interferences study In order to assess the proposed method to the analysis of rifampicin in pharmaceutical dosage forms and urine sample, the interference of basic amino acids, commonly used excipients and additives, co-existing ions were examined. The tolerance limit was taken as the maximum concentration of the foreign substances which caused an approximately ±5% relative error for the determination of rifampicin. The results of interference tests were listed in Table 2. It can be seen that cysteine and ascorbic acid caused seriously negative interference. The interference was due to that they are more easily electrochemically oxidized

435

Table 2 The tolerable concentration ratios of some interfering species to 1.0 × 10−6 mol l−1 rifampicin Substance

Tolerable concentration ratio

Cation K+ , Na+ , Ca2+ , NH4 + Mg2+ , Al3+ , Zn2+ Fe2+ , Fe3+ , Cu2+ , Co2+ , Ni2+

1000 200 50

Anion Cl− , SO4 2− , PO4 3− , NO3 −

1000

Vitamin Thiamine hydrochloride (Vitamin B1 ) Riboflavin (Vitamin B2 ), folic acid (Vitamin Bc) Ascorbic acid (Vitamin C)

100 50 1

Amino acid Valine, serine, arginine, glutamic acid Threonine, phenylalanine, histidine Glycine, tyrosine, lysine Cysteine

500 100 50 10

Others Urea, starch Glucose, sucrose Oxalic acid, uric acid

200 100 50

than rifampicin [46]. Document reported that rifampicin was highly soluble in chloroform while cysteine and ascorbic acid were basically water-soluble [47]. The interference from them can be eliminated by extracting rifampicin with chloroform [47]. Other substance in urine sample in normal concentration range did not cause interference [48]. In addition, the major metabolites of rifampicin, including 3-formylrifamycin SV, rifampicin N-oxidation, 25-desacetyl rifampicin, were all the derivatives of 1,4-naphthalenediol [49], could also produce ECL signal under the proposed ECL system. Therefore, the proposed method could be used for the determination of the total content of rifampicin including its major metabolites in human urine. 3.5. Application In order to assess the validity of the proposed method, rifampicin in eye drop and human urine was determined. Rifampicin eye drops solution was directly diluted with water so that final concentration was in the working range. The determination result of rifampicin in eye drops was shown in Table 3, which agreed well with that obtained by spectrophotometry (λmax = 474 nm) (Pharmacopoeia method) [47]. Moreover,

Table 3 Results of determination of rifampicin in eye drops Amount labeled (mg/10 ml)

Proposed methoda (mg/10 ml)

Official methoda (mg/10 ml)

Added (×10−7 mol l−1 )

Founda (×10−7 mol l−1 )

Average recovery (%)

10

9.7 ± 0.2

9.8 ± 0.1

5.0 10.0 40.0

5.2 ± 0.1 9.7 ± 0.2 39.4 ± 0.2

104 97 99

a

Mean value ± S.D. (n = 5).

436

Y.-D. Liang et al. / Spectrochimica Acta Part A 67 (2007) 430–436

Table 4 Recovery tests in urine samples Sample

Concentration added (×10−7 mol l−1 )

References Founda (×10−7 mol l−1 )

Recovery (%)

1

5.0 10.0

5.1 ± 0.1 9.9 ± 0.2

102 99

2

5.0 10.0

4.9 ± 0.1 10.3 ± 0.1

98 103

3

5.0 10.0

4.8 ± 0.1 9.7 ± 0.1

96 97

a

Mean value ± S.D. (n = 5).

recovery studies were also carried out in samples to which known amounts of rifampicin were added. Each recovery was calculated by comparing the results obtained before and after the addition. As shown in Table 3, the recoveries were between 97 and 104% (n = 5). Each 1 ml of fresh urine sample was diluted to 10 ml with water. Then the diluted urine sample was extracted with 10 ml chloroform three times. The processed organic phases were combined and evaporated to near dryness on a water bath to obtain a residue. The residue was dissolved in 1 ml methanol, and diluted to 10 ml with water. The sample solution obtained was used for ECL determination of rifampicin. The determination results showed that no rifampicin was found in the urine samples. The recovery tests were carried out by adding known amounts of rifampicin to the fresh urine samples. The recoveries, shown in Table 4, were between 96 and 103%. 4. Conclusions This paper proposed a novel type of ECL, named as successive electro- and chemo-oxidation ECL, based on direct electrochemical oxidation of analyte to an intermediate free radical and subsequently chemical oxidation of the intermediate to an excited state species, which was different from annihilation and coreactant ECL in mechanism. As the intermediate was selectively electro-generated on-line, and the formed intermediate was selectively oxidized by various suitable oxidants to form an excited state species that could produce strong luminescence, the novel ECL offered several advantages such as high selectivity, simple and convenient operation, and effective avoidance of side reaction that often took place in homogenous CL reaction. A number of this type of ECL may occur naturally, it has potential application for analytical purpose. Acknowledgment Thanks for the financial support of the National Nature Science Foundation of China (Grant No. 20475043) for present work.

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49]

K.A. Fahnrich, M. Pravda, G.G. Guilbault, Talanta 54 (2001) 531. X.B. Yin, S.J. Dong, E.K. Wang, Trends Anal. Chem. 23 (2004) 432. M.M. Richter, Chem. Rev. 104 (2004) 3003. A.W. Knight, Trends Anal. Chem. 18 (1999) 47. A.W. Knight, G.M. Greenway, Analyst 119 (1994) 879. T.C. Richter, A.J. Bard, Anal. Chem. 67 (1996) 3140. X. Chen, C. Yi, M. Li, X. Lu, Z. Li, P. Li, X. Wang, Anal. Chim. Acta 466 (2002) 79. X.W. Zheng, Z.H. Guo, Z.J. Zhang, Anal. Chim. Acta 441 (2001) 81. S. Sakura, Anal. Chim. Acta 262 (1992) 49. H. Cui, Y. Xu, Z.F. Zhang, Anal. Chem. 76 (2004) 4002. G.N. Chen, R.E. Lin, Z.F. Zhao, J.P. Duan, L. Zhang, Anal. Chim. Acta 341 (1997) 251. R. Panchagnula, S. Agrawal, Int. J. Pharm. 271 (2004) 1. S. Isefuka, C.J. Joyner, A.H.R.W. Simpson, J. Orthop. Res. 19 (2001) 950. N. Duman, A. Cerikbag, C. Johansson, Int. J. Anticrob. Agents 23 (2004) 84. E. Hammam, A.M. Beltagi, M.M. Ghonim, Microchem. J. 77 (2004) 53. S.A. Halvatzis, M.M. Timotheou-Potamia, T.P. Hadjiioannou, Anal. Chim. Acta 272 (1993) 251. W.P. Yang, Y.T. Zhang, Z.J. Zhang, Acta Chim. Sin. 61 (2003) 303. W. Liu, B. Li, Z.J. Zhang, Chin. J. Anal. Chem. 30 (2002) 86. L. Gil, J.V.G. Mateo, J.M. Calatayud, Analyst 122 (1997) 139. G.F. Cheng, H.Y. Qu, D.M. Zhang, J.D. Zhang, P.G. He, Y.Z. Fang, J. Pharm. Biomed. Anal. 29 (2002) 361. Y.Z. Fang, J.C. Jiang, Chin. J. Anal. Chem. 24 (1996) 1371. H.Y. Ma, X.W. Zheng, Z.J. Zhang, Chin. J. Chem. 22 (2004) 279. A. Babaei, A.J. McQuillan, J. Electroanal. Chem. 462 (1999) 266. A. Babaei, A. McQuillan, J. Electroanal. Chem. 441 (1998) 197. T. Okajima, T. Ohsaka, J. Electroanal. Chem. 534 (2002) 181. M.A. Saleml, A.H. Gemeay, Monatsh. Chem. 131 (2000) 117. G. Cilento, W. Adam, Free Radic. Biol. Med. 19 (1995) 103. R. Munday, Free Radic. Biol. Med. 26 (1999) 1475. F. Yang, C. Zhang, W.R.G. Baeyens, X. Zhang, J. Pharm. Biomed. Anal. 30 (2002) 473. M. Yang, C. Liu, X. Hu, P. He, Y. Fang, Anal. Chim. Acta 461 (2002) 141. R. Amadelli, A.D. Battisti, D.V. Girenko, S.V. Kovalyov, A.B. Velichenko, Electrochim. Acta 46 (2000) 341. X.Y. Lia, Y.H. Cui, Y.J. Feng, Z.M. Xie, J.D. Gu, Water Res. 39 (2005) 1972. H. Utsumi, S.K. Han, K. Ichikawa, Water Sci. Technol. 38 (1998) 147. K. Satoh, H. Sakagami, H. Terasaka, Y. Ida, S. Fujisawa, Anticancer Res. 20 (2000) 1577. A. Brunmark, E. Cadenas, Free Radic. Biol. Med. 3 (1987) 169. J. Stauff, P. Bartolmes, Angew. Chem. Int. Ed. Engl. 9 (1970) 307. M. Mochizuki, S.I. Yamazaki, K. Kano, T. Ikeda, Biochim. Biophys. Acta 1569 (2002) 35. M.P. Soriaga, A.T. Hubbard, J. Am. Chem. Soc. 104 (1982) 2735. G.Z. Li, J.H. Mu, X.Z. Li, L.M. Zhai, T.X. Lu, G.L. Dai, Colloids Surf. A: Physicochem. Eng. Aspects 194 (2001) 263. Z. Cao, C. Lau, J. Lu, Analyst 129 (2004) 1262. J.M. Lin, M. Yamada, Trends Anal. Chem. 22 (2003) 99. S. De, A. Girigoswami, J. Colloid Interface Sci. 271 (2004) 485. D. Bruce, J. McCall, M.M. Richter, Analyst 127 (2002) 125. C. Cole, B.D. Muegge, M.M. Richter, Anal. Chem. 75 (2003) 601. J. Du, J. Li, L. Yang, J. Lu, Anal. Chim. Acta 481 (2003) 239. C. Zhao, J. Zhang, J.F. Song, Anal. Biochem. 297 (2001) 170. Editorial Committee of China Pharmacopoeia, China Pharmacopoeia Part II, People’s Health Press, Beijing, 1995, p. 301. M. Marquez, M. Silva, D. Perez-Benditom, Analyst 113 (1988) 1373. B. Bhavika, N. Sharda, S. Singh, J. Pharm. Biomed. Anal. 31 (2003) 607.