Determination of inorganic anions in saliva by electroosmotic flow controlled counterflow isotachophoretic stacking under field-amplified sample injection

Journal of Chromatography B, 935 (2013) 75–79

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

Determination of inorganic anions in saliva by electroosmotic flow controlled counterflow isotachophoretic stacking under field-amplified sample injection Haixia Ren a,b , Hongdeng Qiu a , Xiaojing Liang a , Xusheng Wang a , Shengxiang Jiang a,∗ a Key Laboratory of Chemistry of Northwestern Plant Resources, CAS and Key Laboratory for Natural Medicine of Gansu Province, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, China b University of the Chinese Academy of Sciences, Chinese Academy of Sciences, Beijing 100049, China

a r t i c l e

i n f o

Article history: Received 18 March 2013 Accepted 23 July 2013 Available online 31 July 2013 Keywords: Capillary ion electrophoresis Electrokinetic injection Preconcentration Stacking

a b s t r a c t Under a strong counter-electroosmotic flow, five salivary inorganic anions, bromide, iodide, nitrite, nitrate and thiocyanate were determined by field-amplified sample injection in combination with isotachophoretic stacking. Separation and concentration conditions were investigated. A terminating electrolyte, 5 mM borate, was added in the sample. Under the optimized conditions, Br− , I− and SCN− were concentrated online using 150 mM HCl–Tris buffer at pH 7.8 in a bare fused capillary, providing more than ten thousand of sensitivity enrichment compared with normal injections. The relative standard deviations (RSDs, n = 5) were less than 1% in migration times, 8% in peak areas. Using direct UV detection at 200 nm and 226 nm, the limits of detection (LODs, S/N = 3) were of 0.002–0.01 ␮M. Unfortunately, NO2 − and NO3 − could be observed in purified or deionized water. Therefore, a low dilution factor was applied to saliva samples. Due to the matrix effect, samples were injected in a shorter time, and standard addition method was applied to determine all the five inorganic anions in saliva. The RSDs of the migration times and peak areas were in a range of 0.2–0.4% and 3.0–4.0%, respectively. The LODs were 0.2–2.0 ␮M. The salivary levels of the anions obtained were in accord with the reference data. The external standard method can not be adapted to real samples due to biases caused by electrokinetic injection and errors from high dilutions © 2013 Elsevier B.V. All rights reserved.

1. Introduction Capillary electrophoresis is a powerful analytical tool for the analysis of ionic and neutral compounds, such as small ions [1–3], drugs [4], particles, cells and microorganisms. In capillary electrophoresis, samples can be delivered onto column hydrodynamically or electrokinetically [5]. Electrokinetic injection (EKI) is advantageous for mobile components and able for viscous sample and gel electrophoresis [6]. EKI is first demonstrated as a sample stacking technique by Chien and Burgi in 1991 [7], and theoretically explained by Chien [8]. Analytes were enriched when the conductivity of the sample is much lower than that of the background electrolytes (BGE). It is also called field-amplified sample injection (FASI), allowing hundreds or thousands of improvements [9,10]. During the last decades, EKI stacking methods, combined with solid phase extraction [11,12], pH conjunction [13,14], sweeping

∗ Corresponding author. Tel.: +86 931 4968266; fax: +86 931 8277088. E-mail address: [email protected] (S. Jiang). 1570-0232/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2013.07.022

and isotachophoresis (ITP), have achieved extremely high sensitivity enrichments. For example, cation – or anion – selective exhaustive injection – sweeping, can achieve up to million-fold sensitivity increase [15]. A combination of EKI and ITP, one of the most powerful preconcentration techniques, has demonstrated over 100,000-fold enrichment [16,17]. The terminating electrolyte (TE) is either located after the sample zone, or added in samples. As there is bias for low mobility component, a low concentration of TE is usually added into the sample [18]. Hence, an ITP state is generated during EKI, where compounds migrate at a velocity the same as the leading ion, rather than its own electrophoretic mobility. Based on this fact, million fold sample stacking was achieved with a microchip under electroosmotic flow suppression condition [19]. Later, Breadmore and Quirino have further developed the method, a strong electroosmotic flow (EOF) controlled counterflow isotachophoretic stacking boundary system, where extremely long injection time can be applied [20], with an enrichment factor up to 100,000-fold [17]. Incredible improvements have been made in these studies. However, few of them have involved its application in biological samples.

76

H. Ren et al. / J. Chromatogr. B 935 (2013) 75–79

Saliva contains many important substances, including electrolytes, mucus, antibacterial compounds and various enzymes. Determination of inorganic anions in saliva is important in diagnosis of physiological and biological conditions of an individual. Br− is a sedative hypnotic. I− is involved primarily in the synthesis of thyroid hormones. They are substrates of lactoperoxidase, which plays a vital role in innate immune system [21,22]. NO2 − and NO3 − can be reduced to NO and other bioactive nitrogen oxide species [23]. The SCN− is a detoxification product of cyanide in the liver, and is considered to be a marker of distinguishing between smokers and nonsmokers [24]. Besides, it is a substrate of peroxidase [22,24], an inhibitor on NO formation [25]. Many methods have been developed to determine NO2 − , NO3 − [26] and/or SCN− [24] in saliva, such as potentiometric sensors [27], ion chromatography [28–30], capillary ion chromatography [31], short end injection [32], capillary electrophoresis [33]. In order to reduce protein adhesion, micellar electrokinetic chromatography [34] or BGE additive was applied for saliva analysis. Besides, the capillary can be either coated with polymer layers [35] or washed with SDS in capillary zone electrophoresis [36]. Few studies report the concentration of salivary I− . Even though direct injection of saliva was applied, I− could not be detected in most cases [37]. A high level of I− was reported by capillary ITP [38]. Xu et al. [39] described a more sensitive transient isotachophoretic method for the separation and quantification of these five inorganic anions in saliva. Under strong counter EOF, the mobilities of Br− , I− , NO2 − , NO3 − and SCN− are higher than the EOF, which ensures detection of these anions in negative polarity. In this study, we are aimed to apply the counterflow isotachophoretic stacking to determination of these anions in saliva. 2. Materials and methods 2.1. Chemicals and reagents Tris(hydroxymethyl)aminomethane (Tris) and KBr were supplied from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) was obtained from Bio Basic Inc. (Toronto, Canada). MES was purchased from Amresco (Solon, USA). Boric acid was from Fengyue Chemicals Co., Ltd. (Tianjin, China). NaI, NaNO2 and KNO3 were from Shanghai Chemical Reagent General Works (Shanghai, China). KSCN was obtained from Beijing Chemical works (Beijing, China). All analytes and other chemicals were of analytical grade. BGEs were prepared by dilution of HCl to 150 mM, and adjusted to pH 7.8 with Tris. The TEs were prepared from MES, HEPES and boric acid at a concentration of 100 mM, and the pH was adjusted to 7.8 with Tris. 2.2. Standards and sample preparation Stock standard solutions of 10 mM KBr, 10 mM NaI, 10 mM NaNO2 , 10 mM KNO3 and 10 mM KSCN were prepared in purified water. Working standard solutions were diluted in TE as required. Saliva samples were collected from a healthy volunteer. 2 mL saliva was diluted with acetonitrile to 4 mL. The mixture was centrifuged at 8000 rpm for 10 min and filtered through a 0.45 ␮m membrane filter. 2.3. Capillary electrophoresis Experiments were performed on An Agilent CE system (Agilent Technologies, Beijing, China) equipped with a UV DAD utilizing ChemStation (Rev.A 09.03). Detection was monitored at 200 and 226 nm. The capillary cartridge was thermostated at 20 ◦ C. Electrophoresis was carried out at −20 kV.

A fused capillary (Hebei Yongnian Country Reafine Chromatography, China) of 50 ␮m i.d. with a length of 68.5 cm, 60 cm to detector, was used for separation. Before use, the capillary was flushed (≈930 mbar) with methanol (10 min), 1 M HCl (20 min) and 1 M NaOH (20 min). Between runs, the capillary was rinsed with 0.1 M NaOH for 2 min followed by 3 min BGE. The analysis was performed by external standard and standard addition methods. For the former method, 1.6 mL standard solution was used for injection at −10 kV, 200 s. In addition, 10 ␮L saliva–acetonitrile mixture was diluted to 5 mL in 5 mM borate for injection, giving a high dilution factor of 1000 for saliva. For the latter one, 500 ␮L saliva–acetonitrile was mixed with 50 ␮L 100 mM borate and 500 ␮L standard solutions, and diluted to 1 mL. An aliquot of 200 ␮L mixture was then injected at −10 kV, 15 s. 3. Results and discussion 3.1. Optimization of separation conditions A strong EOF to the cathode can be generated in the fused-silica capillary. Under counter flow conditions, only anions with mobility greater than the EOF can migrate into the capillary towards the anode. The pH value has little effect on ionization of the inorganic acid in a pH rang 7–8.6. Supplementary Fig. 1 shows the influence of pH on effective mobility and EOF. Anions showed the highest electrophoretic mobilities at pH 7.8, where the maximum peak efficiency was observed (Supplementary Fig. 2) The EOF was almost constant over this pH range. Therefore, this decrease is referred to interactions between the five anions with the background electrolyte, such as ion paring with Tris. The resolution between Br− and I− increased as the increase of pH. However, almost constant resolution was observed between NO2 − and NO3 − over the pH range. The separation pH was compromised at pH 7.8 in the following study. Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jchromb. 2013.07.022. The UV spectra were tested from 190 to 250 nm. The I− has a maximum absorption at 226 nm. The Br− , NO2 − , NO3 − and SCN− exhibit higher absorption at shorter wavelength. As the background absorption of Tris is considered, 200 nm and 226 nm were selected for detection. The BGE concentration plays an important role in peak heights. The concentration of the buffer was studied in a range of 50–200 mM. Peak heights increased with the increase of concentrations from 50 to 150 mM, while decreased at a concentration larger than 150 mM. Larger conductivity difference between BGE and sample, higher stacking improvement will be obtained during EKI. Besides, the stacking factors are dependent upon the concentration of the leading in ITP process. Therefore, higher concentration of the BGE favours the concentration of the anions. However, high ionic strength of the BGE generates high current, leading to a significant Joule heating effect. Thus, the concentration was set at 150 mM in the following studies. 3.2. Counterflow isotachophoretic stacking When the conductance of the sample is much lower than that of the BGE, ions move faster than that inside the BGE during EKI, resulting in a stacked zone at the inlet of the capillary, in the form of FASI. When an amount of TE is added into the sample and the capillary is filled with leading electrolyte, an ITP boundary will be generated, where the anions migrate at a velocity of the leading ion. As the mobilities of bromide and iodide are higher than that

H. Ren et al. / J. Chromatogr. B 935 (2013) 75–79

Fig. 1. Influence of terminators on signal to noise ratios of the target compounds. Conditions: sample, 0.02 mM Br− , 0.004 mM I− and 1 mM SCN− ; injection, −10 kV, 5 s; capillary, 50 ␮m i.d., 38.5 cm total length; voltage, −16 kV.

77

Fig. 2. Electropherogram of a standard solution at 200 nm (upper trace) and 226 nm (lower trace). Conditions: sample, 0.016 ␮M Br− , 0.0016 ␮M I− and 0.4 ␮M SCN− , containing 5 mM borate; injection, −10 kV, 200 s; capillary, 50 ␮m i.d., 68.5 cm total length; voltage, −20 kV.

3.3. Saliva analysis by standard addition of BGE chloride, the two ions are lack of leader. Once the separation is performed, the TE and other less mobile components will be removed by the EOF. HEPES (pKa = 7.5) and borate (pKa = 9.0), which would contribute to buffer capacities, were tested as TEs besides MES (pKa = 6.1). Concentrations of MES, HEPES and borate on peak intensities were studied from 1 to 10 mM. The peak heights increased as the concentration of HEPES and MES reduced, which was consistent with the reported study [17]. The highest peaks were observed at a concentration of 1 mM, either HEPES or MES. The optimized concentration with the highest peaks for borate were found at 5 mM. It is inferred that at a concentration lower than 5 mM, the ionic strength is too low to produce an ITP boundary. Unfortunately, NO2 − and NO3 − are ubiquitous. At an injection time of 200 s, we have found NO2 − and NO3 − in distill water, deionized water and purified water. Therefore, only Br− , I− and SCN− were considered in this method. As shown in Fig. 1, the optimized signal to noise ratios for Br− , I− and SCN− were all observed in 5 mM borate. Usually, the lower conductivity of the sample, the more analytes could be injected; while the larger mobility variation between analytes and terminators, the more broaden peak would be observed. Here, the conductivity order was 5 mM borate <1 mM HEPES <1 mM MES (calculated by Peakmaster software), while their mobilities were in a reversed order. Obviously, at a given injection time, the field enhanced procedure plays an important part in stacking of Br− , I− and SCN− . Thus a lower conductivity sample in 5 mM borate was more advantageous. It is worth mentioning that with 2 successive injections of a 1.6 mL sample at 200 s, the peak area deviations were 3–4%, 5–10% and 13–20% in 1 mM HEPES, 5 mM borate and 1 mM MES, respectively. This is corresponding to their buffer capacities. Subsequently, 5 mM borate was selected the terminator. The effect of injection time on peak height was investigated from 100 s to 300 s as well. Peak heights of Br− , I− and SCN− increased with injection time. When the time was longer than 200 s, peak distortion of Br− occurred and peak height of I− decreased. It is worth mentioning that the peak efficiency reduced over time. Therefore, 200 s was used to determinate Br− , I− and SCN− . Electropherogram of a standard solution was shown in Fig. 2. The calibration curve was constructed using standard solutions, and the results were listed in Table 1. The linear correlation coefficients were larger than 0.995. The relative standard deviations (RSDs) of migration time and peak area for the three anions were less than 1% and 8%, respectively.

Saliva samples were deproteinized with acetonitrile, as the concentration of NO3 − and SCN− would decrease at room temperature [35]. In order to determine NO2 − and NO3 − , the saliva–acetonitrile mixture was diluted to 2-fold. As shown in Fig. 3, when the saliva–acetonitrile mixture is diluted by a low factor, there is a significant matrix effect on EKI. The highest peak efficiency and signal to noise ratio (Supplementary Fig. 3) were obtained at an injection time of 15 s, when neither NO2 − nor NO3 − could be observed in water or 25% acetonitrile. Therefore, samples were injected at −10 kV, 15 s in the following. Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jchromb. 2013.07.022. Fig. 4 shows a typical electropherogram of anions in a saliva sample. The performance was studied by standard addition method and the results were summarized in Table 2. The RSDs of the anions were found below 0.4% and 4% for the migration time and peak area respectively. The levels of the five inorganic anions were listed in Table 3, along with reference data reported previously. The salivary levels of the anions were among the range given by the reported studies, except Br− . The concentrations of Br− , I− and SCN− obtained with the external standard method were listed in

Fig. 3. Electropherograms of a standard solution mixed 1:1 (v/v) with water (upper trace) and saliva–acetonitrile (lower trace) at 200 nm. The standard solution contains 0.04 mM bromide, 0.005 mM iodide, 1 mM thiocyanate and 10 mM borate. The sample was injected −10 kV, 15 s.

78

H. Ren et al. / J. Chromatogr. B 935 (2013) 75–79

Table 1 Linearity, limit of detection (LOD, S/N = 3), repeatability and enrichment factor (EF) obtained from the standard solutions. Anions

Bromide Iodide Thiocyanate

Correlation coefficient

Slope

0.999 0.9989 0.994

Intercept

456.3 711.5 408.9

Range (␮M)

7.610 2.451 −21.28

Repeatability (%RSD)

0.02–0.2 0.002–0.04 0.1–1

Time

Area

0.77 0.79 0.94

7.7 1.6 2.2

LOD (␮M)

a

EF

Area 0.010 0.002 0.008

Height

21,127 25,573 15,385

11,646 15,723 27,515

a EF = Is × Cn / Cs /In , where Cn and Cs are the concentration used for normal and EKI injections, respectively; Is and In are peak intensity (area or height) obtained from normal and EKI injections, respectively.

Table 2 Linearity and repeatability of the standard addition method. Anions

Correlation coefficient

Slope

Intercept

Bromide Iodide Nitrite Nitrate Thiocyanate

0.9969 0.9968 0.9954 0.9951 0.9965

2021.3 2314.3 1087.2 1823.7 1413.0

12.545 4.2615 68.262 831.15 758.25

a

Range (mM)

0.002–0.01 0.001–0.005 0.05–0.2 0.08–0.4 0.2–1

a

Repeatability (%RSD)

Time

Area

0.3 0.3 0.4 0.2 0.4

2.0 3.8 2.6 4.0 2.6

LOD (␮M)

1.2 0.9 2.5 2.6 7

Obtained from five injections of sample spiked with 0.01 mM bromide, 0.005 mM iodide and 1 mM thiocyanate; 0.05 mM nitrite; 0.4 mM nitrate.

Table 3 Salivary concentrations and recoveries of anions obtained from standard addition method (A) and external standard method (E).

Br− I− NO2 − NO3 − SCN−

Concentration (mM)

Recovery (%)

A

E

a

0.0124 0.0037 0.1256 0.8662 1.2521

0.0162 0.0032

104.6 99.8 108.1 106.3 110.5

1.6405

A

Reference (mM) b

E

109.5 98.7

104.0

[37] 0.042–0.077 ND–0.009 0.043–0.626 0.313–1.327 0.070–2.558

[39] 0.068 0.003 0.076 0.318 0.742

± ± ± ± ±

0.018 0.008 0.115 0.255 0.247

a

Saliva–acetonitrile was mixed with 0.005 mM bromide, 0.0025 mM iodide and 0.5 mM thiocyanate; 0.075 mM nitrite; 0.32 mM nitrate. Other conditions were as described in Section 2.3. b 1000-fold diluted saliva was spiked with 0.005 ␮M bromide, 0.0025 ␮M iodide and 0.5 ␮M thiocyanate.

Table 3 as well, which were approximately 30% higher than those obtained with the standard addition method except I− . The discrepancy may refer to mobility, matrix and/or instrumental biases caused by electrokinetic injection [40]. The mobility bias is significant when the total mobility of injection is different with that of separation, which is the case for SCN− . However, the mobility bias is eliminated if mobilities are the same during the injection and separation, which is the case for bromide and iodide. Different

dilutions of saliva lead to the matrix bias. A source of the instrumental bias comes from the sample volume. A volume of 200 ␮L sample in a micro vial was used for injection with the standard addition method, smaller than that of 1.6 mL with the external method. Besides, the small sampling (10 ␮L) is an origin of analytical errors. Therefore, the external method cannot be simply adapted to the 1000-fold diluted sample. The Br− concentration was lower than the reference values. Another study reported a range of 0.001–0.053 mg/g in saliva [29]. It is inferred the low level of Br− is attributed to diet of the volunteer who seldom eats seafood, which has high levels of bromine. The NO3 − concentration varies widely depending on the variety and preparation of the food [41]. A very high NO3 − concentration up to 6.83 mM was reported [30]. 4. Conclusion

Fig. 4. Electropherograms of inorganic anions in saliva at 200 nm (upper trace) and 226 nm (lower trace). Conditions: sample, 500 ␮L saliva–acetonitrile was mixed with 50 ␮L 100 mM borate, and diluted to 1 mL; injection, −10 kV, 15 s. Other conditions were as described in section 2.3.

The strong electroosmotic flow controlled counterflow isotachophoretic stacking, with more than ten thousand sensitivity increments, has been applied to saliva analysis of inorganic anions. The matrix effect can be reduced by a high dilution up to 1000-fold, requiring only 1.6 ␮L saliva for one injection. However, analytical errors occur subsequently. By considering the omnipresent NO2 − and NO3 − , samples should not be handled in the high dilution. Besides, biases caused by EKI could not be neglected. Therefore, the method cannot be simply adapted to real samples. Standard addition method with lower sample dilution and shorter injection time is an alternative due to the matrix effect.

H. Ren et al. / J. Chromatogr. B 935 (2013) 75–79

Acknowledgements This work was supported by the Natural Science Foundation, China (21105107, 21175143) and the National Science & Technology Major Project, China (2011ZX05010, 2011ZX05011-004). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]

I. Ali, V.K. Gupta, H.Y. Aboul-Enein, Electrophoresis 26 (2005) 3988. I. Ali, H.Y. Aboul-Enein, Crit. Rev. Anal. Chem. 32 (2002) 337. I. Ali, H.Y. Aboul-Enein, Anal. Lett. 35 (2002) 2053. I. Ali, H.Y. Aboul-Enein, V.K. Gupta, S.F. Li, J. Capill. Electrophor. Microchip Technol. 9 (2006) 85. I. Ali, H.Y. Aboul Enein, V.K. Gupta, Anal. Lett. 39 (2006) 2345. S.F.Y. Li, Capillary Electrophoresis: Principles, Practice and Applications, Elsevier Science, Amsterdam, 1992. R.-L. Chien, D.S. Burgi, J. Chromatogr. A 559 (1991) 141. R.L. Chien, Anal. Chem. 63 (1991) 2866. S.L. Simpson Jr., J.P. Quirino, S. Terabe, J. Chromatogr. A 1184 (2008) 504. J.P. Quirino, S. Terabe, J. Chromatogr. A 902 (2000) 119. B. Claude, R. Nehmé, P. Morin, Anal. Chim. Acta 699 (2011) 242. H. Ye, S. Xia, L. Yu, X. Xu, C. Zheng, H. Xu, L. Wang, X. Liu, Z. Cai, G. Chen, Electrophoresis 32 (2011) 2823. D.J. Weiss, K. Saunders, C.E. Lunte, Electrophoresis 22 (2001) 59. Y. Zhao, C.E. Lunte, Anal. Chem. 71 (1999) 3985. J.P. Quirino, S. Terabe, Anal. Chem. 72 (2000) 1023. Z.Q. Xu, K. Nakamura, A.R. Timerbaev, T. Hirokawa, Anal. Chem. 83 (2011) 398. M.C. Breadmore, J.P. Quirino, Anal. Chem. 80 (2008) 6373. P. Jandik, W.R. Jones, J. Chromatogr. A 546 (1991) 431.

[19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41]

79

B. Jung, R. Bharadwaj, J.G. Santiago, Anal. Chem. 78 (2006) 2319. M.C. Breadmore, Electrophoresis 29 (2008) 1082. http://en.wikipedia.org/wiki/Lactoperoxidase (last accessed 18.10.12). P.G. Furtmüller, W. Jantschko, G. Regelsberger, C. Jakopitsch, J. Arnhold, C. Obinger, Biochemistry 41 (2002) 11895. E. Weitzberg, M. Hezel, J.O. Lundberg, Anesthesiology 113 (2010) 1460, 1410.1097/ALN.1460b1013e3181fcf1463cc. M. Valdés, M. Díaz-García, Crit. Rev. Anal. Chem. 34 (2004) 9. C.A. Palmerini, F. Marmottini, G. Arienti, J. Biochem. Mol. Toxicol. 26 (2012) 87. D.Y. Boudko, J. Chromatogr. B 851 (2007) 186. S.S.M. Hassan, W.H. Mohmoud, A.-S.S.H. Elgazwy, N.M. Badawy, Electroanalysis 18 (2006) 2070. M.I.H. Helaleh, T. Korenaga, J. Chromatogr. B: Biomed. Sci. Appl. 744 (2000) 433. ˙ Polkowska, J. Namie´snik, J. Chromatogr. B 875 (2008) 419. I. Demkowska, Z. Z.F. Chen, B.W. Darvell, V.W.H. Leung, Arch. Oral Biol. 49 (2004) 863. L. Rong, Z. Liu, M. Ma, L.W. Lim, T. Takeuchi, Chromatographia 73 (2011) 501. L. Vitali, V.T. Fávere, G.A. Micke, J. Chromatogr. A 1218 (2011) 2327. D.K. Lloyd, J. Chromatogr. B 866 (2008) 154. M. Mori, W. Hu, J.S. Fritz, H. Tsue, T. Kaneta, S. Tanaka, Fresenius J. Anal. Chem. 370 (2001) 429. Y. Tanaka, N. Naruishi, H. Fukuya, J. Sakata, K. Saito, S.-i. Wakida, J. Chromatogr. A 1051 (2004) 193. A. Gáspár, P. Juhász, K. Bágyi, J. Chromatogr. A 1065 (2005) 327. M. Mori, T. Yamamoto, M. Kaseda, S. Yamada, H. Itabashi, J. Chromatogr. B 887–888 (2012) 1. ´ Talanta 59 (2003) 643. J. Sádecká, J. Polonsky, Z. Xu, T. Doi, A.R. Timerbaev, T. Hirokawa, Talanta 77 (2008) 278. Z. Krivácsy, A. Gelencsér, J. Hlavay, G. Kiss, Z. Sárvári, J. Chromatogr. A 834 (1999) 21. J.J. Doel, M.P. Hector, C.V. Amirtham, L.A. Al-Anzan, N. Benjamin, R.P. Allaker, Eur. J. Oral Sci. 112 (2004) 424.