- Email: [email protected]
Vapour phase Fourier transform infrared spectrometric determination of l -cysteine and l -cystine
Available online at www.sciencedirect.com
Talanta 74 (2008) 753–759
Vapour phase Fourier transform infrared spectrometric determination of l-cysteine and l-cystine K. Kargosha ∗ , S.H. Ahmadi, M. Zeeb, S.R. Moeinossadat Chemistry and Chemical Engineering Research Center of Iran, P.O. Box 14335-186, Iran Received 21 February 2007; received in revised form 4 July 2007; accepted 5 July 2007 Available online 10 July 2007
Abstract A novel and selective procedure for the determination of l-cysteine and l-cystine based on vapour-generation Fourier transform infrared spectrometry is described. Potassium iodate solution was injected into a glass vessel containing l-cysteine and/or l-cystine. The evolved CO was swept by a stream of nitrogen to an infrared gas cell. The vapour phase FTIR spectra were continuously recorded, as a function of time, between 2240 and 2000 cm−1 , which includes the CO absorption band. The maximum absorbance at 2170 cm−1 was selected as a measurement criterion. The calibration curve was linear over the range 6–300 mg L−1 . The method provided a limit of detection of 2 mg L−1 of l-cysteine, a throughput of 12 samples h−1 and an R.S.D. of 1.76% for five independent analyses of a 75 mg L−1 l-cysteine solution. For the measurement of l-cysteine and l-cystine separately, after measuring total concentration of l-cysteine and l-cystine, l-cysteine was masked with p-benzoquinone at a pH of 3 and individual l-cystine was determined. The amount of l-cysteine was obtained by difference. The method was applied to the determination of l-cysteine and l-cystine in pharmaceutical and urine samples. Results obtained for real samples compared well with those obtained by a reference spectrometric method. © 2007 Elsevier B.V. All rights reserved. Keywords: FTIR spectrometry; Vapour phase; l-Cysteine; l-Cystine
1. Introduction l-cysteine (CSH) and l-cystine (CSSC) are important compounds in a wide range of samples such as biological tissues, body fluids, food products and medicines. l-Cysteine is a sulfur amino acid and contains a sulfhydryl group. When l-cysteine is exposed to air, it is oxidized to form l-cystine, which is a dimer of two l-cysteine molecules joined by a weak disulfide bond. Hepatic cystinuria and some other diseases are characterized by high concentration of l-cysteine and l-cystine in urine (usually > 400 mg L−1 ). The analysis of l-cysteine and l-cystine is essential and often difficult, although they can be determined by electrophoretic [1–5], chromatographic [6–11], electrochemical [12–14] and spectrometric [15–20] methods. However, some of these methods are insensitive, tedious, time consuming or expensive. The analysis of mixture of l-cysteine and l-cystine has been rarely
∗
Corresponding author. Fax: +98 21 44580762. E-mail addresses: [email protected], K [email protected] (K. Kargosha). 0039-9140/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2007.07.003
performed by non-chromatographic techniques. In addition, there are some interferences in electrochemical and spectrometric methods. For example, ethylenediaminetetraacetic acid (EDTA), metal ions, surfactants, and N-acetyl cysteine or other sulfhydryl thiols are the main interferences in electrochemical methods [12,13]. Most of chromatographic methods [7,9,11] are combined with pre- or post-column derivatization. These methods are time consuming and use expensive reagents and equipment. FTIR spectroscopy is very useful for carrying out determinations in gaseous or vapour samples owing to the high transparency of gases, the low background values and the possibilities offered by the use of multiple-pass cells, which can provide good sensitivity [21]. By using oxidation of lcysteine and/or l-cysteine with potassium iodate and generation of CO gas, we have developed a simple vapour phase FTIR analytical method for the determination of l-cysteine and lcystine in aqueous solutions. The system employed for vapour generation is a simple Pyrex glass reactor, into which liquid samples are introduced and potassium iodate solution is directly injected through a septum; the generated vapour is transported by an inert carrier gas flow and the correspond-
754
K. Kargosha et al. / Talanta 74 (2008) 753–759
ing transient signals are measured in the gas cell of an FTIR instrument. The aforementioned methodology was employed for the determination of l-cysteine and l-cystine in pharmaceutical and urine samples, without interference from other amino acids. 2. Experimental 2.1. Apparatus and reagents A vector 22 FTIR spectrometer from Bruker (Ettlingen, Germany) equipped with a DTGS mid-range detector, a KBr Ge/Sb2 S3 -coated beamsplitter and a Globar source was employed for spectral measurements; spectra were obtained by accumulating 20 scans at a resolution of 4 cm−1 , using a laboratory-made gas cell made of PTFE with a length of 10 cm and an internal diameter of 0.9 cm, equipped with ZnSe windows. Version 4 of Opus software, developed by Bruker, was employed to control the instrument, for data acquisition and also for processing the analytical results. The manifold employed for vapour-generation FTIR measurements (Fig. 1) was a single-channel manifold with a nitrogen carrier flow, which included a laboratory-made removable glass sample vessel of 20 mL internal volume with a gas inlet, a gas outlet and a septum. Sample vessel (reactor) was introduced inside a hot waterbath, the temperature of which was controlled by means of a thermocouple and operated using laboratory-made electrically controlled heater. To avoid the presence of water drops inside the measurement cell, at the exit of the glass reactor a small stainless steel u-form tube was introduced inside an ice-bath to
act as a water trap in which the excess water vapour condensed before passing through the gas cell. Other tubes employed in the manifold were made of PTFE. A trio 1000 GC–MS instrument from Fisons (Manchester, England) was employed for separating and detecting the generated gases. The column used for this separation was a Pro Plot-Q from Chrompack. All chemicals used were of analytical-reagent grade. lCysteine, l-cystine, potassium iodate, sodium metaperiodate, p-benzoquinone, citric acid monohydrate, tri-sodium citrate dihydrate and 37% (v/v) hydrochloric acid were obtained from Merck. A 1500 mg L−1 stock CSH solution was daily prepared by dissolving 0.15 g of l-cysteine in de-ionized water and diluting to 100 mL with citric acid buffer solution. A 1500 mg L−1 stock CSSC solution was daily prepared by dissolving 0.15 g l-cystine in 2 mL 37% (v/v) HCL and diluting to 100 mL with citric acid buffer solution. Working solutions of both analytes were prepared from stock solution by appropriate dilution. Citric acid buffer was obtained by dissolving 1.7963 g citric acid monohydrate and 0.4265 g tri-sodium citrate dihydrate in deionized water and diluting to 1 L. Nitrogen of 99.95% purity was employed as the carrier gas and de-ionized water was used throughout. 2.2. Vapour phase FTIR analysis A 2 mL volume of 4% (m/v) potassium iodate solution was injected into a glass vessel containing 10 mL of (6–300 mg L−1 ) l-cysteine and/or l-cystine solution. Prior to any measurement, an N2 carrier flow was passed through the empty vessel and the corresponding background was obtained. A new vessel containing the sample or standard solution was placed in a water-bath at 70 ◦ C; N2 flow was passed through the vessel; the potassium iodate solution was injected and the CO generated was transported to the gas cell of the FTIR spectrometer using a nitrogen carrier flow of 4.4 mL min−1 , without the need for an equilibrium time before measurement. FTIR spectra were obtained from 20 accumulated scans at a resolution of 4 cm−1 between 2240 and 2000 cm−1 which includes the CO absorption band. The maximum absorbance at 2170 cm−1 , corrected by a baseline established between 2240 and 2000 cm−1 , was selected as a measurement criterion. 2.3. Reference spectrometric method A spectrometric method based on the reaction of CSH with sodium arsenate and the absorbance measurement at 335 nm was applied as the reference method [15]. The reference method is specific for CSH and in mixtures of CSH and CSSC both amino acids can be determined separately by analysing two samples, one without and one with KCN. The colour is developed at room temperature for 2 h.
Fig. 1. Manifold employed for vapour generation and FTIR determination of l-cysteine and l-cystine: (1) nitrogen gas, (2) flow meter, (3) tube for entrance of nitrogen gas, (4) reactor, (5) water-bath, (6) heater, (7) tube for exit of nitrogen gas, (8) syringe for injection of oxidant, (9) water trap, (10) FTIR and (11) computer.
3. Results and discussion The oxidation of CSH and CSSC in aqueous solutions with potassium permanganate, potassium iodate, potassium
K. Kargosha et al. / Talanta 74 (2008) 753–759
755
iodide, sodium metaperiodate, sodium bismuthate and Nbromosuccinimide was evaluated. Appreciable amount of carbon monoxide was produced only when potassium iodate or sodium metaperiodate was used as oxidizing agent. Beside CO, CO2 was also generated by using these two oxidants. These gases were detected by injecting the evolved gases into a GC–MS spectrometer (see Fig. 2). The vapour phase FTIR spectrum of the gaseous species produced in sample vessel containing CSH, CSSC and either of these two oxidants are shown in Fig. 3. Figs. 2 and 3 reveal that the main gaseous species produced in the oxidation of CSH or CSSC by potassium iodate or sodium metaperiodate are CO and CO2 . Fig. 3(A) shows that the height of the CO absorbance signal is about twice of the CO2 signal in using potassium iodate as oxidant. Sodium metaperiodate is a stronger oxidant and can oxidize other amino acids (serine and theronine) and organic compounds with formation of CO2 . These amino acids and organic compounds are usually present in pharmaceutical and urine samples. On the other hand by selecting CO instead of CO2 as the final analyte, more precise results were obtained. However, potassium iodate was selected as optimum oxidant. Vapour phase FTIR spectra of eight l-cysteine standard solutions with different concentrations (6–300 mg L−1 ) present well defined and intense bands in the range 2240–2000 cm−1 (see Fig. 4). The resulting difference spectra show that the relative changes in absorbance appear to be linearly spaced. The oxidation of CSH and CSSC with potassium iodate solution, the generation of CO and the introduction of the evolved gas from reactor into the gas cell of the FTIR spectrometer are the most important steps of the proposed method. Different chemical and physical variables affect these steps. These variables must be optimized to introduce the highest and most reproducible amount of CO into the gas cell. Since the behaviour of CSH and CSSC toward these variables was found to be the same, only CSH samples were used through the rest of the work. 3.1. Effect of potassium iodate concentration on CO generation The concentration and volume of the oxidant necessary for the complete generation of CO from CSH was evaluated. Experiments were carried out in a monoparametric mode and the results demonstrated that, consistent data can be achieved by using excess amount of oxidant (4%, m/v) and 2 mL volume of KIO3 . 3.2. Effect of reactor temperature The oxidation reaction between l-cysteine or l-cystine and potassium iodate leading to the release of CO is very efficient but not very rapid and the effect of temperature is a decisive factor; it also affects the kinetics of the reaction and the rate of CO releasing out of the solution. As can be seen in Fig. 5, the absorbance of the transient peaks increases, as the reactor temperature increases. However, from an experimental point of view
Fig. 2. (A) vapour phase GC–MS of the gaseous spices produced in the oxidation l-cysteine or l-cystine by potassium iodate. (B)–(D) are mass spectra of each chromatogram component.
756
K. Kargosha et al. / Talanta 74 (2008) 753–759
Fig. 5. Effect of the reactor temperature on vapour phase FTIR determination of l-cysteine. Experimental condition: 10 mL of 75 mg L−1 l-cysteine, 20 mL reactor volume, 4.4 mL min−1 N2 flow rate, pH of 3 and 2 mL of 4% (m/v) potassium iodate were employed. Fig. 3. Vapour phase FTIR spectrum of the gaseous spices produced in the oxidation of l-cysteine or l-cystine by (A) potassium iodate and (B) sodium metaperiodate.
extremely fast measurement of CO creates problems. Therefore, a temperature of 70 ◦ C was selected in order to decrease water vapour transfer to a minimum and to obtain a compromise between speed, sensitivity, simplicity and reproducibility of the analytical signals. 3.3. Effect of reactor volume An increase in the reactor volume, which increases the CO dispersion by increasing the total volume of the system, has very small influence on the analytical signals obtained from the absorbance of the transient peaks. A reactor volume of 20 mL (the most readily available glass vial) is recommended for the determination of CSH and CSSC.
Fig. 4. Eight vapour phase co-added FTIR spectra obtained for eight standard solutions of l-cysteine with different concentrations (6, 15, 30, 45, 75, 150, 225 and 300 mg L−1 ). Spectra were obtained from 20 accumulated scans at a resolution of 4 cm−1 .
3.4. Effect of N2 carrier flow rate The flow rate of the carrier gas (N2 ) is a critical parameter in vapour phase FTIR spectrometric analyses [22–25], which affects the analytical sensitivity and the sampling frequency. This parameter controls the volatility of the analytes and the speed of vapour introduction into the measurement cell. As can be seen in Fig. 6, an increase in the carrier flow rate causes a decrease in sensitivity, but increases the sampling frequency. Hence a nitrogen flow rate of 4.4 mL min−1 , which allows a 12 h−1 sampling frequency, was selected in order to achieve a compromise between analytical sensitivity and sample throughput. 3.5. Effect of a fixed reaction time The possible sensitivity improvement was evaluated by carring out a series of additional experiments under the condition
Fig. 6. Effect of N2 carrier flow rate on vapour phase FTIR determination of lcysteine. Experimental condition: 10 mL of 75 mg L−1 l-cysteine, 20 mL reactor volume, reactor temperature of 70 ◦ C, pH of 3 and 2 mL of 4% (m/v) potassium iodate were employed.
K. Kargosha et al. / Talanta 74 (2008) 753–759
757
Table 1 Effect of pH on vapour phase FTIR determination of l-cysteine pH
Absorbance ± (sn − 1 × 10−4 )
0.5 1 1.5 2 2.5 3 3.5 4 4.5
0.00123 0.00239 0.00381 0.00634 0.00893 0.00898 0.00932 0.00912 0.00873
± ± ± ± ± ± ± ± ±
2.51 2.96 2.11 2.98 2.53 1.49 2.12 2.39 3.21
Experimental condition: 10 mL of 75 mg L−1 l-cysteine, reactor temperature of 70 ◦ C, 20 mL reactor volume, 4.4 mL min−1 N2 flow rate and 2 mL of 4% (m/v) potassium iodate were employed. The standard deviation values correspond to three independent measurements. Fig. 7. Effect of equilibrium time on vapour phase FTIR determination of l-cysteine. Experimental condition: 10 mL of 75 mg L−1 l-cysteine, reactor temperature of 70 ◦ C, 20 mL reactor volume, 4.4 mL min−1 N2 flow rate, pH of 3 and 2 mL of 4% (m/v) potassium iodate were employed.
of stop flow mode. In this mode of experiments, the stream of N2 carrier gas is stopped for a period of time to allow the oxidation reaction and CO release to reach an equilibrium. In fact in this mode, the oxidant is injected and after a fixed reaction time the evolved CO is introduced into the measurement cell by the stream of carrier gas. Times between 50 and 600 s were tried as fixed reaction times before the measurement. Fig. 7 reveals that an increase in fixed reaction time up to 300 s leads to an increase in the analytical signal and then to a decrease. Maximum obtainable sensitivity in 300 s fixed reaction time shows no improvement compared to sensitivity obtained by continuous flow mode (see Figs. 6 and 7). 3.6. Effect of pH and individual determination of l-cysteine and l-cystine The pH affects the chemical behaviour of KIO3 as an oxidant. The possibility of using potassium iodate as a selective oxidant to form more CO from CSH or CSSC by changing the pH was studied. To evaluate this possibility, nine samples (containing 75 mg L−1 CSH) with different pH values were tested. This study was assessed in the pH rang of 0.5–4.5. The pH of each solution was adjusted to a fixed value using appropriate buffer. As shown in Table 1, the signal increased as pH increased up to 3.5 and then decreased. However the pH of 3 (using citric acid buffer) was selected as most convenient pH because of two reasons: (1) the best reproducibility was obtained at this pH (see Table 1) and (2) the optimum pH for individual determination of CSH and CSSC was found to be 3. The resolution of l-cysteine and l-cystine mixtures is based on the fact that at pH 3 the p-benzoquinone (PBQ) reacts with the thiol group of l-cysteine and mask it to prevent its oxidation with KIO3 [18]. At this pH amino groups of CSH and CSSC are protonated and do not react with PBQ [13]. l-Cystine is a dimer of two l-cysteine molecules joined by a weak disulfide bond which cannot react with PBQ at the pH 3. The effect of pH on masking the l-cysteine with PBQ and the determination of l-cystine was
investigated. Study was performed by analysing samples with variable pH which contain 75 mg L−1 of each amino acid. Based on the results obtained in this study the pH of 3 was selected as an optimum pH value. This pH was established by using citric acid buffer. However, in mixtures of CSH and CSSC both amino acids can be determined separately by analysing two samples, one without and one with PBQ. l-Cystine is determined alone by analysing the sample with PBQ and the concentration of lcysteine plus l-cystine is determined by analysing the sample without PBQ. 3.7. Effect of interfering compounds An interference study, aimed at the determination of lcysteine and/or l-cystine in pharmaceutical preparations and urine samples, was performed. Samples containing fixed concentrations of l-cysteine (100 mg L−1 ) or l-cystine and various concentrations of the interfering compounds were subjected to the proposed method. The tolerance limit was defined as interfering compound and l-cysteine concentrations ratio causing a relative error of ≤5. The results are shown in Table 2. Table 2 Effect of various interfering compounds on the determination of 100 mg L−1 l-cysteine or l-cystine using optimum conditions reported under vapour phase FTIR analysis Substance
Tolerance ratioa ([substance]/ [cysteine or cystine])
Amino acids: Ile, Lys, Phe, Thr, Val, Arg, His, Gly, Ala, Pro, Ser, Tyr, Leu, Trp, Asp, Asn, Glu, Nty, Acy Citric, maleic, malic, tartaric, lactic, benzylic and mendelic acid Glycerol and ethylene glycerol Glocose and fructose Provic and glycolic acid Albomine, bilirubin, insulin, urea and uric acid Ascorbic acid N-Acetyl cysteine
250
220 200 170 160 150 50 20
a Maximum concentration ratio of interfering compounds to cysteine causing a relative error of ≤5%.
758
K. Kargosha et al. / Talanta 74 (2008) 753–759
Table 3 The declared content of Amino plasmal 10% solution (g L−1 )
4.2. Urine samples
Isoleucine Leucine Lysine hydrochloride Methionine Phenylalanine Theronine Tryptophan Valine Arginine Histidine Glycine Alanine Proline Aspartic acid Asparagine·H2 O Acetylcysteine Glutamic acid Ornithine hydrochloride Serine Tyrosine N-Acetyltyrosine
The accuracy of the proposed method was evaluated by the determination of l-cysteine and l-cystine in five different human urine samples. The urine samples were diluted by mixing it with the same volume of citric acid buffer solution. The values obtained for urine samples were compared with those found by a reference spectrometric method [15]. The results are summarized in Table 5.
5.10 8.90 7.00 3.80 4.10 4.10 1.80 4.80 9.20 5.20 7.90 13.70 8.90 1.30 3.72 0.68 4.60 3.20 2.40 0.30 1.23
4.3. Analytical figures of merit
4. Applications 4.1. Spiked pharmaceutical product Three samples of a two times diluted commercial pharmaceutical product, Amino plasmal 10% (purchased from B. Braun Melsungen AG Company, Germany), were spiked with different amounts of CSH and CSSC. Amino plasmal was diluted with citric acid buffer. The recovery study was performed by analysing these spiked samples with the proposed method. The composition of Amino plasmal is shown in Table 3. The results of this recovery study are shown in Table 4. The mean recoveries for l-cysteine and l-cystine were 101.57 and 101.43, respectively.
The main analytical characteristics of the method were established from typical calibration lines, obtained under the optimum experimental conditions and from the analysis of real samples. A typical expression for a calibration obtained from 6 to 300 mg L−1 of l-cysteine is y = 0.0001643 + 0.0001169x with r = 0.9997 where y is the absorbance of the transient peak and x is the concentration of l-cysteine (mg L−1 ). A limit of detection of 2 mg L−1 of l-cysteine was obtained by using the criterion, LOD = ksbl /m, where k is a factor (=2), sbl the standard deviation of the blank measurements and m is the calibration slope. The R.S.D. for five independent analyses of a standard sample with a concentration of 75 mg L−1 l-cysteine was 1.76%. The Same calibration curve and analytical figures of merit can be obtained in the determination of l-cystine. The precision of l-cysteine and l-cystine determination in real samples can be established from the relative standard deviation (R.S.D.) of three independent analysis. As can be seen from the data in Table 5 an average of R.S.D. <4% was obtained. The sampling frequency of the method is 12 h−1 , including the time required for the installation of glass sample vial in manifold.
Table 4 Recoveries obtained for l-cysteine and l-cystine in pharmaceutical samples by vapour phase FTIR spectrometry Pharmaceutical sample
1 2 3
l-Cysteine
l-Cystine
Added (mg L−1 )
Found (mg L−1 )
Recovery (%)
Added (mg L−1 )
Found (mg L−1 )
Recovery (%)
60 75 150
60.60 76.16 153.23
101.00 101.55 102.15
60 75 150
60.86 76.09 152.11
101.43 101.45 101.40
Mean
101.57
101.43
Table 5 Results obtained for the determination of l-cysteine and l-cystine in urine samples by vapour phase FTIR spectrometry and by a reference method [15] Urine sample
l-Cysteine (mg L−1 ) Proposed method ± sn − 1
1 2 3 4 5
52.03 66.20 45.12 60.02 53.22
± ± ± ± ±
1.55 2.06 1.16 1.61 1.31
l-Cystine (mg L−1 ) Reference method ± sn − 1 49.18 64.01 42.12 59.46 50.11
± ± ± ± ±
1.42 2.50 1.91 2.12 1.70
The standard deviation values correspond to three independent measurements.
Proposed method ± sn − 1 21.14 38.22 16.11 40.51 25.13
± ± ± ± ±
0.62 1.19 0.49 1.27 0.94
Reference method ± sn − 1 20.19 36.36 14.02 38.12 23.16
± ± ± ± ±
0.89 1.33 0.47 1.10 0.79
K. Kargosha et al. / Talanta 74 (2008) 753–759
5. Conclusion From the experimental results we conclude that both lcysteine and l-cystine can be oxidized by KIO3 to form carbon monoxide. At pH 3 thiol group of l-cysteine can be masked with PBQ and individual l-cystine can be oxidized to form CO. This is confirmed by recording the FTIR spectra of all gaseous species produced by this oxidation reaction. In addition, these results indicate that the proposed method can be useful for the determination of l-cysteine and l-cystine in pharmaceutical and urine samples and it provides excellent reproducibility and appropriate sensitivity. References [1] [2] [3] [4]
R.M. Latorre, S.H. Cassou, J. Saurina, J. Chromatogr. A 934 (2001) 105. K. Arlt, S. Brandt, J. Kehr, J. Chromatogr. A 926 (2001) 319. M. Ummadi, B.C. Weimer, J. Chromatogr. A 964 (2002) 243. Z. Shen, Z. Sun, L. Wu, K. Wu, S. Sun, Z. Huang, J. Chromatogr. A 979 (2002) 227. [5] R.M. Latorre, J. Saurina, S.H. Cassou, J. Chromatogr. A 976 (2002) 55. [6] P. Chaimbault, K. Petritis, C. Elfakir, M. Dreux, J. Chromatogr. A 855 (1999) 191.
759
[7] E. Bald, R. Glowacki, J. Drzewoski, J. Chromatogr. A 913 (2001) 319. [8] H. Yu, Y.S. Ding, S.F. Mou, P. Jandik, J. Cheng, J. Chromatogr. A 966 (2002) 89. [9] A. Namera, M. Yashiki, M. Nishida, T. Kojima, J. Chromatogr. B 776 (2002) 49. [10] J. Xu, Y. Wang, Y. Xian, L. Jin, K. Tanaka, Talanta 60 (2003) 1123. [11] K. Amarnath, V. Amarnath, K. Amarnath, H.L. Valentine, W.M. Valentine, Talanta 60 (2003) 1229. [12] F.G. Banica, J.C. Moreira, A.G. Fogg, Analyst 119 (1994) 309. [13] G. Hignett, S. Threlfell, A.J. Wain, N.S. Lawrence, S.J. Wilkins, J. Davis, R.G. Compton, M.F. Cardosi, Analyst 126 (2001) 353. [14] N. Spataru, B.V. Sarada, E. Popa, D.A. Tryk, A. Fujishima, Anal. Chem. 73 (2001) 514. [15] J. Chrastil, Analyst 115 (1990) 1383. [16] T.P. Ruiz, C.M. Lozano, V. Tomas, J. Carpena, Analyst 117 (1992) 1025. [17] M.C. Icardo, L.L. Zamora, M. Calatayud, Analyst 123 (1998) 1685. [18] D.A.M. Zaia, K.C.L. Ribas, C.T.B.V. Zaia, Talanta 50 (1999) 1003. [19] L. Nie, H. Ma, M. Sun, X. Li, M. Su, S. Liang, Talanta 59 (2003) 959. [20] F. Tanaka, N. Mase, C.F. Barbas, Chem. Commun. (2004) 1762. [21] P.L. Hanst, Fresenius, J. Anal. Chem. 324 (1986) 579. [22] A.P. Ponce, J.M. Garrigues, S. Garrigues, M. de la Guardia, Analyst 123 (1998) 1817. [23] K. Kargosha, M. Khanmohammadi, M. Ghadiri, Analyst 126 (2001) 1432. [24] M. Gallignani, M. Valero, C. Ayala, M. del Rosario Brunetto, A. Sanchez, J.L. Burguera, M. Burguera, Talanta 64 (2004) 1290. [25] N.M.M. Coelho, S. Garrigues, M. de la Guardia, Talanta 68 (2006) 836.
Talanta 74 (2008) 753–759
Vapour phase Fourier transform infrared spectrometric determination of l-cysteine and l-cystine K. Kargosha ∗ , S.H. Ahmadi, M. Zeeb, S.R. Moeinossadat Chemistry and Chemical Engineering Research Center of Iran, P.O. Box 14335-186, Iran Received 21 February 2007; received in revised form 4 July 2007; accepted 5 July 2007 Available online 10 July 2007
Abstract A novel and selective procedure for the determination of l-cysteine and l-cystine based on vapour-generation Fourier transform infrared spectrometry is described. Potassium iodate solution was injected into a glass vessel containing l-cysteine and/or l-cystine. The evolved CO was swept by a stream of nitrogen to an infrared gas cell. The vapour phase FTIR spectra were continuously recorded, as a function of time, between 2240 and 2000 cm−1 , which includes the CO absorption band. The maximum absorbance at 2170 cm−1 was selected as a measurement criterion. The calibration curve was linear over the range 6–300 mg L−1 . The method provided a limit of detection of 2 mg L−1 of l-cysteine, a throughput of 12 samples h−1 and an R.S.D. of 1.76% for five independent analyses of a 75 mg L−1 l-cysteine solution. For the measurement of l-cysteine and l-cystine separately, after measuring total concentration of l-cysteine and l-cystine, l-cysteine was masked with p-benzoquinone at a pH of 3 and individual l-cystine was determined. The amount of l-cysteine was obtained by difference. The method was applied to the determination of l-cysteine and l-cystine in pharmaceutical and urine samples. Results obtained for real samples compared well with those obtained by a reference spectrometric method. © 2007 Elsevier B.V. All rights reserved. Keywords: FTIR spectrometry; Vapour phase; l-Cysteine; l-Cystine
1. Introduction l-cysteine (CSH) and l-cystine (CSSC) are important compounds in a wide range of samples such as biological tissues, body fluids, food products and medicines. l-Cysteine is a sulfur amino acid and contains a sulfhydryl group. When l-cysteine is exposed to air, it is oxidized to form l-cystine, which is a dimer of two l-cysteine molecules joined by a weak disulfide bond. Hepatic cystinuria and some other diseases are characterized by high concentration of l-cysteine and l-cystine in urine (usually > 400 mg L−1 ). The analysis of l-cysteine and l-cystine is essential and often difficult, although they can be determined by electrophoretic [1–5], chromatographic [6–11], electrochemical [12–14] and spectrometric [15–20] methods. However, some of these methods are insensitive, tedious, time consuming or expensive. The analysis of mixture of l-cysteine and l-cystine has been rarely
∗
Corresponding author. Fax: +98 21 44580762. E-mail addresses: [email protected], K [email protected] (K. Kargosha). 0039-9140/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2007.07.003
performed by non-chromatographic techniques. In addition, there are some interferences in electrochemical and spectrometric methods. For example, ethylenediaminetetraacetic acid (EDTA), metal ions, surfactants, and N-acetyl cysteine or other sulfhydryl thiols are the main interferences in electrochemical methods [12,13]. Most of chromatographic methods [7,9,11] are combined with pre- or post-column derivatization. These methods are time consuming and use expensive reagents and equipment. FTIR spectroscopy is very useful for carrying out determinations in gaseous or vapour samples owing to the high transparency of gases, the low background values and the possibilities offered by the use of multiple-pass cells, which can provide good sensitivity [21]. By using oxidation of lcysteine and/or l-cysteine with potassium iodate and generation of CO gas, we have developed a simple vapour phase FTIR analytical method for the determination of l-cysteine and lcystine in aqueous solutions. The system employed for vapour generation is a simple Pyrex glass reactor, into which liquid samples are introduced and potassium iodate solution is directly injected through a septum; the generated vapour is transported by an inert carrier gas flow and the correspond-
754
K. Kargosha et al. / Talanta 74 (2008) 753–759
ing transient signals are measured in the gas cell of an FTIR instrument. The aforementioned methodology was employed for the determination of l-cysteine and l-cystine in pharmaceutical and urine samples, without interference from other amino acids. 2. Experimental 2.1. Apparatus and reagents A vector 22 FTIR spectrometer from Bruker (Ettlingen, Germany) equipped with a DTGS mid-range detector, a KBr Ge/Sb2 S3 -coated beamsplitter and a Globar source was employed for spectral measurements; spectra were obtained by accumulating 20 scans at a resolution of 4 cm−1 , using a laboratory-made gas cell made of PTFE with a length of 10 cm and an internal diameter of 0.9 cm, equipped with ZnSe windows. Version 4 of Opus software, developed by Bruker, was employed to control the instrument, for data acquisition and also for processing the analytical results. The manifold employed for vapour-generation FTIR measurements (Fig. 1) was a single-channel manifold with a nitrogen carrier flow, which included a laboratory-made removable glass sample vessel of 20 mL internal volume with a gas inlet, a gas outlet and a septum. Sample vessel (reactor) was introduced inside a hot waterbath, the temperature of which was controlled by means of a thermocouple and operated using laboratory-made electrically controlled heater. To avoid the presence of water drops inside the measurement cell, at the exit of the glass reactor a small stainless steel u-form tube was introduced inside an ice-bath to
act as a water trap in which the excess water vapour condensed before passing through the gas cell. Other tubes employed in the manifold were made of PTFE. A trio 1000 GC–MS instrument from Fisons (Manchester, England) was employed for separating and detecting the generated gases. The column used for this separation was a Pro Plot-Q from Chrompack. All chemicals used were of analytical-reagent grade. lCysteine, l-cystine, potassium iodate, sodium metaperiodate, p-benzoquinone, citric acid monohydrate, tri-sodium citrate dihydrate and 37% (v/v) hydrochloric acid were obtained from Merck. A 1500 mg L−1 stock CSH solution was daily prepared by dissolving 0.15 g of l-cysteine in de-ionized water and diluting to 100 mL with citric acid buffer solution. A 1500 mg L−1 stock CSSC solution was daily prepared by dissolving 0.15 g l-cystine in 2 mL 37% (v/v) HCL and diluting to 100 mL with citric acid buffer solution. Working solutions of both analytes were prepared from stock solution by appropriate dilution. Citric acid buffer was obtained by dissolving 1.7963 g citric acid monohydrate and 0.4265 g tri-sodium citrate dihydrate in deionized water and diluting to 1 L. Nitrogen of 99.95% purity was employed as the carrier gas and de-ionized water was used throughout. 2.2. Vapour phase FTIR analysis A 2 mL volume of 4% (m/v) potassium iodate solution was injected into a glass vessel containing 10 mL of (6–300 mg L−1 ) l-cysteine and/or l-cystine solution. Prior to any measurement, an N2 carrier flow was passed through the empty vessel and the corresponding background was obtained. A new vessel containing the sample or standard solution was placed in a water-bath at 70 ◦ C; N2 flow was passed through the vessel; the potassium iodate solution was injected and the CO generated was transported to the gas cell of the FTIR spectrometer using a nitrogen carrier flow of 4.4 mL min−1 , without the need for an equilibrium time before measurement. FTIR spectra were obtained from 20 accumulated scans at a resolution of 4 cm−1 between 2240 and 2000 cm−1 which includes the CO absorption band. The maximum absorbance at 2170 cm−1 , corrected by a baseline established between 2240 and 2000 cm−1 , was selected as a measurement criterion. 2.3. Reference spectrometric method A spectrometric method based on the reaction of CSH with sodium arsenate and the absorbance measurement at 335 nm was applied as the reference method [15]. The reference method is specific for CSH and in mixtures of CSH and CSSC both amino acids can be determined separately by analysing two samples, one without and one with KCN. The colour is developed at room temperature for 2 h.
Fig. 1. Manifold employed for vapour generation and FTIR determination of l-cysteine and l-cystine: (1) nitrogen gas, (2) flow meter, (3) tube for entrance of nitrogen gas, (4) reactor, (5) water-bath, (6) heater, (7) tube for exit of nitrogen gas, (8) syringe for injection of oxidant, (9) water trap, (10) FTIR and (11) computer.
3. Results and discussion The oxidation of CSH and CSSC in aqueous solutions with potassium permanganate, potassium iodate, potassium
K. Kargosha et al. / Talanta 74 (2008) 753–759
755
iodide, sodium metaperiodate, sodium bismuthate and Nbromosuccinimide was evaluated. Appreciable amount of carbon monoxide was produced only when potassium iodate or sodium metaperiodate was used as oxidizing agent. Beside CO, CO2 was also generated by using these two oxidants. These gases were detected by injecting the evolved gases into a GC–MS spectrometer (see Fig. 2). The vapour phase FTIR spectrum of the gaseous species produced in sample vessel containing CSH, CSSC and either of these two oxidants are shown in Fig. 3. Figs. 2 and 3 reveal that the main gaseous species produced in the oxidation of CSH or CSSC by potassium iodate or sodium metaperiodate are CO and CO2 . Fig. 3(A) shows that the height of the CO absorbance signal is about twice of the CO2 signal in using potassium iodate as oxidant. Sodium metaperiodate is a stronger oxidant and can oxidize other amino acids (serine and theronine) and organic compounds with formation of CO2 . These amino acids and organic compounds are usually present in pharmaceutical and urine samples. On the other hand by selecting CO instead of CO2 as the final analyte, more precise results were obtained. However, potassium iodate was selected as optimum oxidant. Vapour phase FTIR spectra of eight l-cysteine standard solutions with different concentrations (6–300 mg L−1 ) present well defined and intense bands in the range 2240–2000 cm−1 (see Fig. 4). The resulting difference spectra show that the relative changes in absorbance appear to be linearly spaced. The oxidation of CSH and CSSC with potassium iodate solution, the generation of CO and the introduction of the evolved gas from reactor into the gas cell of the FTIR spectrometer are the most important steps of the proposed method. Different chemical and physical variables affect these steps. These variables must be optimized to introduce the highest and most reproducible amount of CO into the gas cell. Since the behaviour of CSH and CSSC toward these variables was found to be the same, only CSH samples were used through the rest of the work. 3.1. Effect of potassium iodate concentration on CO generation The concentration and volume of the oxidant necessary for the complete generation of CO from CSH was evaluated. Experiments were carried out in a monoparametric mode and the results demonstrated that, consistent data can be achieved by using excess amount of oxidant (4%, m/v) and 2 mL volume of KIO3 . 3.2. Effect of reactor temperature The oxidation reaction between l-cysteine or l-cystine and potassium iodate leading to the release of CO is very efficient but not very rapid and the effect of temperature is a decisive factor; it also affects the kinetics of the reaction and the rate of CO releasing out of the solution. As can be seen in Fig. 5, the absorbance of the transient peaks increases, as the reactor temperature increases. However, from an experimental point of view
Fig. 2. (A) vapour phase GC–MS of the gaseous spices produced in the oxidation l-cysteine or l-cystine by potassium iodate. (B)–(D) are mass spectra of each chromatogram component.
756
K. Kargosha et al. / Talanta 74 (2008) 753–759
Fig. 5. Effect of the reactor temperature on vapour phase FTIR determination of l-cysteine. Experimental condition: 10 mL of 75 mg L−1 l-cysteine, 20 mL reactor volume, 4.4 mL min−1 N2 flow rate, pH of 3 and 2 mL of 4% (m/v) potassium iodate were employed. Fig. 3. Vapour phase FTIR spectrum of the gaseous spices produced in the oxidation of l-cysteine or l-cystine by (A) potassium iodate and (B) sodium metaperiodate.
extremely fast measurement of CO creates problems. Therefore, a temperature of 70 ◦ C was selected in order to decrease water vapour transfer to a minimum and to obtain a compromise between speed, sensitivity, simplicity and reproducibility of the analytical signals. 3.3. Effect of reactor volume An increase in the reactor volume, which increases the CO dispersion by increasing the total volume of the system, has very small influence on the analytical signals obtained from the absorbance of the transient peaks. A reactor volume of 20 mL (the most readily available glass vial) is recommended for the determination of CSH and CSSC.
Fig. 4. Eight vapour phase co-added FTIR spectra obtained for eight standard solutions of l-cysteine with different concentrations (6, 15, 30, 45, 75, 150, 225 and 300 mg L−1 ). Spectra were obtained from 20 accumulated scans at a resolution of 4 cm−1 .
3.4. Effect of N2 carrier flow rate The flow rate of the carrier gas (N2 ) is a critical parameter in vapour phase FTIR spectrometric analyses [22–25], which affects the analytical sensitivity and the sampling frequency. This parameter controls the volatility of the analytes and the speed of vapour introduction into the measurement cell. As can be seen in Fig. 6, an increase in the carrier flow rate causes a decrease in sensitivity, but increases the sampling frequency. Hence a nitrogen flow rate of 4.4 mL min−1 , which allows a 12 h−1 sampling frequency, was selected in order to achieve a compromise between analytical sensitivity and sample throughput. 3.5. Effect of a fixed reaction time The possible sensitivity improvement was evaluated by carring out a series of additional experiments under the condition
Fig. 6. Effect of N2 carrier flow rate on vapour phase FTIR determination of lcysteine. Experimental condition: 10 mL of 75 mg L−1 l-cysteine, 20 mL reactor volume, reactor temperature of 70 ◦ C, pH of 3 and 2 mL of 4% (m/v) potassium iodate were employed.
K. Kargosha et al. / Talanta 74 (2008) 753–759
757
Table 1 Effect of pH on vapour phase FTIR determination of l-cysteine pH
Absorbance ± (sn − 1 × 10−4 )
0.5 1 1.5 2 2.5 3 3.5 4 4.5
0.00123 0.00239 0.00381 0.00634 0.00893 0.00898 0.00932 0.00912 0.00873
± ± ± ± ± ± ± ± ±
2.51 2.96 2.11 2.98 2.53 1.49 2.12 2.39 3.21
Experimental condition: 10 mL of 75 mg L−1 l-cysteine, reactor temperature of 70 ◦ C, 20 mL reactor volume, 4.4 mL min−1 N2 flow rate and 2 mL of 4% (m/v) potassium iodate were employed. The standard deviation values correspond to three independent measurements. Fig. 7. Effect of equilibrium time on vapour phase FTIR determination of l-cysteine. Experimental condition: 10 mL of 75 mg L−1 l-cysteine, reactor temperature of 70 ◦ C, 20 mL reactor volume, 4.4 mL min−1 N2 flow rate, pH of 3 and 2 mL of 4% (m/v) potassium iodate were employed.
of stop flow mode. In this mode of experiments, the stream of N2 carrier gas is stopped for a period of time to allow the oxidation reaction and CO release to reach an equilibrium. In fact in this mode, the oxidant is injected and after a fixed reaction time the evolved CO is introduced into the measurement cell by the stream of carrier gas. Times between 50 and 600 s were tried as fixed reaction times before the measurement. Fig. 7 reveals that an increase in fixed reaction time up to 300 s leads to an increase in the analytical signal and then to a decrease. Maximum obtainable sensitivity in 300 s fixed reaction time shows no improvement compared to sensitivity obtained by continuous flow mode (see Figs. 6 and 7). 3.6. Effect of pH and individual determination of l-cysteine and l-cystine The pH affects the chemical behaviour of KIO3 as an oxidant. The possibility of using potassium iodate as a selective oxidant to form more CO from CSH or CSSC by changing the pH was studied. To evaluate this possibility, nine samples (containing 75 mg L−1 CSH) with different pH values were tested. This study was assessed in the pH rang of 0.5–4.5. The pH of each solution was adjusted to a fixed value using appropriate buffer. As shown in Table 1, the signal increased as pH increased up to 3.5 and then decreased. However the pH of 3 (using citric acid buffer) was selected as most convenient pH because of two reasons: (1) the best reproducibility was obtained at this pH (see Table 1) and (2) the optimum pH for individual determination of CSH and CSSC was found to be 3. The resolution of l-cysteine and l-cystine mixtures is based on the fact that at pH 3 the p-benzoquinone (PBQ) reacts with the thiol group of l-cysteine and mask it to prevent its oxidation with KIO3 [18]. At this pH amino groups of CSH and CSSC are protonated and do not react with PBQ [13]. l-Cystine is a dimer of two l-cysteine molecules joined by a weak disulfide bond which cannot react with PBQ at the pH 3. The effect of pH on masking the l-cysteine with PBQ and the determination of l-cystine was
investigated. Study was performed by analysing samples with variable pH which contain 75 mg L−1 of each amino acid. Based on the results obtained in this study the pH of 3 was selected as an optimum pH value. This pH was established by using citric acid buffer. However, in mixtures of CSH and CSSC both amino acids can be determined separately by analysing two samples, one without and one with PBQ. l-Cystine is determined alone by analysing the sample with PBQ and the concentration of lcysteine plus l-cystine is determined by analysing the sample without PBQ. 3.7. Effect of interfering compounds An interference study, aimed at the determination of lcysteine and/or l-cystine in pharmaceutical preparations and urine samples, was performed. Samples containing fixed concentrations of l-cysteine (100 mg L−1 ) or l-cystine and various concentrations of the interfering compounds were subjected to the proposed method. The tolerance limit was defined as interfering compound and l-cysteine concentrations ratio causing a relative error of ≤5. The results are shown in Table 2. Table 2 Effect of various interfering compounds on the determination of 100 mg L−1 l-cysteine or l-cystine using optimum conditions reported under vapour phase FTIR analysis Substance
Tolerance ratioa ([substance]/ [cysteine or cystine])
Amino acids: Ile, Lys, Phe, Thr, Val, Arg, His, Gly, Ala, Pro, Ser, Tyr, Leu, Trp, Asp, Asn, Glu, Nty, Acy Citric, maleic, malic, tartaric, lactic, benzylic and mendelic acid Glycerol and ethylene glycerol Glocose and fructose Provic and glycolic acid Albomine, bilirubin, insulin, urea and uric acid Ascorbic acid N-Acetyl cysteine
250
220 200 170 160 150 50 20
a Maximum concentration ratio of interfering compounds to cysteine causing a relative error of ≤5%.
758
K. Kargosha et al. / Talanta 74 (2008) 753–759
Table 3 The declared content of Amino plasmal 10% solution (g L−1 )
4.2. Urine samples
Isoleucine Leucine Lysine hydrochloride Methionine Phenylalanine Theronine Tryptophan Valine Arginine Histidine Glycine Alanine Proline Aspartic acid Asparagine·H2 O Acetylcysteine Glutamic acid Ornithine hydrochloride Serine Tyrosine N-Acetyltyrosine
The accuracy of the proposed method was evaluated by the determination of l-cysteine and l-cystine in five different human urine samples. The urine samples were diluted by mixing it with the same volume of citric acid buffer solution. The values obtained for urine samples were compared with those found by a reference spectrometric method [15]. The results are summarized in Table 5.
5.10 8.90 7.00 3.80 4.10 4.10 1.80 4.80 9.20 5.20 7.90 13.70 8.90 1.30 3.72 0.68 4.60 3.20 2.40 0.30 1.23
4.3. Analytical figures of merit
4. Applications 4.1. Spiked pharmaceutical product Three samples of a two times diluted commercial pharmaceutical product, Amino plasmal 10% (purchased from B. Braun Melsungen AG Company, Germany), were spiked with different amounts of CSH and CSSC. Amino plasmal was diluted with citric acid buffer. The recovery study was performed by analysing these spiked samples with the proposed method. The composition of Amino plasmal is shown in Table 3. The results of this recovery study are shown in Table 4. The mean recoveries for l-cysteine and l-cystine were 101.57 and 101.43, respectively.
The main analytical characteristics of the method were established from typical calibration lines, obtained under the optimum experimental conditions and from the analysis of real samples. A typical expression for a calibration obtained from 6 to 300 mg L−1 of l-cysteine is y = 0.0001643 + 0.0001169x with r = 0.9997 where y is the absorbance of the transient peak and x is the concentration of l-cysteine (mg L−1 ). A limit of detection of 2 mg L−1 of l-cysteine was obtained by using the criterion, LOD = ksbl /m, where k is a factor (=2), sbl the standard deviation of the blank measurements and m is the calibration slope. The R.S.D. for five independent analyses of a standard sample with a concentration of 75 mg L−1 l-cysteine was 1.76%. The Same calibration curve and analytical figures of merit can be obtained in the determination of l-cystine. The precision of l-cysteine and l-cystine determination in real samples can be established from the relative standard deviation (R.S.D.) of three independent analysis. As can be seen from the data in Table 5 an average of R.S.D. <4% was obtained. The sampling frequency of the method is 12 h−1 , including the time required for the installation of glass sample vial in manifold.
Table 4 Recoveries obtained for l-cysteine and l-cystine in pharmaceutical samples by vapour phase FTIR spectrometry Pharmaceutical sample
1 2 3
l-Cysteine
l-Cystine
Added (mg L−1 )
Found (mg L−1 )
Recovery (%)
Added (mg L−1 )
Found (mg L−1 )
Recovery (%)
60 75 150
60.60 76.16 153.23
101.00 101.55 102.15
60 75 150
60.86 76.09 152.11
101.43 101.45 101.40
Mean
101.57
101.43
Table 5 Results obtained for the determination of l-cysteine and l-cystine in urine samples by vapour phase FTIR spectrometry and by a reference method [15] Urine sample
l-Cysteine (mg L−1 ) Proposed method ± sn − 1
1 2 3 4 5
52.03 66.20 45.12 60.02 53.22
± ± ± ± ±
1.55 2.06 1.16 1.61 1.31
l-Cystine (mg L−1 ) Reference method ± sn − 1 49.18 64.01 42.12 59.46 50.11
± ± ± ± ±
1.42 2.50 1.91 2.12 1.70
The standard deviation values correspond to three independent measurements.
Proposed method ± sn − 1 21.14 38.22 16.11 40.51 25.13
± ± ± ± ±
0.62 1.19 0.49 1.27 0.94
Reference method ± sn − 1 20.19 36.36 14.02 38.12 23.16
± ± ± ± ±
0.89 1.33 0.47 1.10 0.79
K. Kargosha et al. / Talanta 74 (2008) 753–759
5. Conclusion From the experimental results we conclude that both lcysteine and l-cystine can be oxidized by KIO3 to form carbon monoxide. At pH 3 thiol group of l-cysteine can be masked with PBQ and individual l-cystine can be oxidized to form CO. This is confirmed by recording the FTIR spectra of all gaseous species produced by this oxidation reaction. In addition, these results indicate that the proposed method can be useful for the determination of l-cysteine and l-cystine in pharmaceutical and urine samples and it provides excellent reproducibility and appropriate sensitivity. References [1] [2] [3] [4]
R.M. Latorre, S.H. Cassou, J. Saurina, J. Chromatogr. A 934 (2001) 105. K. Arlt, S. Brandt, J. Kehr, J. Chromatogr. A 926 (2001) 319. M. Ummadi, B.C. Weimer, J. Chromatogr. A 964 (2002) 243. Z. Shen, Z. Sun, L. Wu, K. Wu, S. Sun, Z. Huang, J. Chromatogr. A 979 (2002) 227. [5] R.M. Latorre, J. Saurina, S.H. Cassou, J. Chromatogr. A 976 (2002) 55. [6] P. Chaimbault, K. Petritis, C. Elfakir, M. Dreux, J. Chromatogr. A 855 (1999) 191.
759
[7] E. Bald, R. Glowacki, J. Drzewoski, J. Chromatogr. A 913 (2001) 319. [8] H. Yu, Y.S. Ding, S.F. Mou, P. Jandik, J. Cheng, J. Chromatogr. A 966 (2002) 89. [9] A. Namera, M. Yashiki, M. Nishida, T. Kojima, J. Chromatogr. B 776 (2002) 49. [10] J. Xu, Y. Wang, Y. Xian, L. Jin, K. Tanaka, Talanta 60 (2003) 1123. [11] K. Amarnath, V. Amarnath, K. Amarnath, H.L. Valentine, W.M. Valentine, Talanta 60 (2003) 1229. [12] F.G. Banica, J.C. Moreira, A.G. Fogg, Analyst 119 (1994) 309. [13] G. Hignett, S. Threlfell, A.J. Wain, N.S. Lawrence, S.J. Wilkins, J. Davis, R.G. Compton, M.F. Cardosi, Analyst 126 (2001) 353. [14] N. Spataru, B.V. Sarada, E. Popa, D.A. Tryk, A. Fujishima, Anal. Chem. 73 (2001) 514. [15] J. Chrastil, Analyst 115 (1990) 1383. [16] T.P. Ruiz, C.M. Lozano, V. Tomas, J. Carpena, Analyst 117 (1992) 1025. [17] M.C. Icardo, L.L. Zamora, M. Calatayud, Analyst 123 (1998) 1685. [18] D.A.M. Zaia, K.C.L. Ribas, C.T.B.V. Zaia, Talanta 50 (1999) 1003. [19] L. Nie, H. Ma, M. Sun, X. Li, M. Su, S. Liang, Talanta 59 (2003) 959. [20] F. Tanaka, N. Mase, C.F. Barbas, Chem. Commun. (2004) 1762. [21] P.L. Hanst, Fresenius, J. Anal. Chem. 324 (1986) 579. [22] A.P. Ponce, J.M. Garrigues, S. Garrigues, M. de la Guardia, Analyst 123 (1998) 1817. [23] K. Kargosha, M. Khanmohammadi, M. Ghadiri, Analyst 126 (2001) 1432. [24] M. Gallignani, M. Valero, C. Ayala, M. del Rosario Brunetto, A. Sanchez, J.L. Burguera, M. Burguera, Talanta 64 (2004) 1290. [25] N.M.M. Coelho, S. Garrigues, M. de la Guardia, Talanta 68 (2006) 836.