Journal of Chromatography A, 1407 (2015) 216–221
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Short communication
Simple determination of fluoride in biological samples by headspace solid-phase microextraction and gas chromatography–tandem mass spectrometry Sun-Myung Kwon a , Ho-Sang Shin b,∗ a b
Department of Environmental Science, Kongju National University, Kongju, 314-701, South Korea Department of Environmental Education, Kongju National University, Kongju 314-701, South Korea
a r t i c l e
i n f o
Article history: Received 7 April 2015 Received in revised form 27 June 2015 Accepted 29 June 2015 Available online 2 July 2015 Keywords: Gas chromatography–tandem mass spectrometry Headspace solid phase microextraction Biological samples Fluoride
a b s t r a c t A simple and convenient method to detect fluoride in biological samples was developed. This method was based on derivatization with 2-(bromomethyl)naphthalene, headspace solid phase microextraction (HSSPME) in a vial, and gas chromatography-tandem mass spectrometric detection. The HS-SPME parameters were optimized as follows: selection of CAR/PDMS fiber, 0.5% 2-(bromomethyl)naphthalene, 250 mg/L 15-crown-5-ether as a phase transfer catalyst, extraction and derivatization temperature of 95 ◦ C, heating time of 20 min and pH of 7.0. Under the established conditions, the lowest limits of detection were 9 and 11 g/L in 1.0 ml of plasma and urine, respectively, and the intra- and inter-day relative standard deviation was less than 7.7% at concentrations of 0.1 and 1.0 mg/L. The calibration curve showed good linearity of plasma and urine with r = 0.9990 and r = 0.9992, respectively. This method is simple, amenable to automation and environmentally friendly. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Fluoride ion (F− ) is a ubiquitous element present in various environmental matrices [1]. Many communities add fluoride to their drinking water to promote dental health, and hydrogen fluoride is used in the production of aluminum and chlorofluorocarbons. Fluoride poisoning occurs by inorganic fluoride salts in public water systems [2,3], ingestion of fluoride-containing products [4–6] and hydrogen fluoride exposure [7,8]. Acute inhalation exposure to gaseous hydrogen fluoride can cause respiratory damage, such as severe irritation and pulmonary edema, in humans. Severe ocular irritation and dermal burns may occur following eye or skin exposure in humans [6]. Otherwise, chronic exposure to fluoride at low levels has the beneficial effect of dental cavity prevention and may also be useful for the treatment of osteoporosis [6–8]. Many methods have been proposed for the chromatographic determination of fluoride in various matrices. Ion chromatography [9–11] is commonly used for the determination of fluoride, but the sensitivity is not adequate to detect g/L levels in biological samples. High performance liquid chromatography (HPLC) methods are performed using derivatives of the F− -La3+ -alizarin
∗ Corresponding author. Tel.: +82 41 850 8811; fax: +82 41 850 8998. E-mail address: [email protected] (H.-S. Shin). http://dx.doi.org/10.1016/j.chroma.2015.06.066 0021-9673/© 2015 Elsevier B.V. All rights reserved.
complexone ternary complex [12], and gas chromatography (GC) methods are performed following the reaction with silylation reagents, such as trimethylchlorosilane and trimethylimidazolesilane [13–17]. When the derivatized fluoride is identified based on its retention time in the case of HPLC and GC, a high rate of false-identification may be introduced, especially in the analysis of complex samples, such as biological samples. GC-mass spectrometry (MS) methods were published for the determination of fluoride in human whole blood and urine after derivatization with pentafluorobenzyl bromide [18] and pentafluorobenzyl p-toluenesulphonate [19], but these methods do not have a sufficient detection limit to detect fluoride in real samples. Headspace (HS) GC–MS following ethylation with triethyloxoniumtetrachloroferrate (III) was recently developed for the sensitive analysis of fluoride in tap water, seawater and urine [20]. The derivatization reagent, however, cannot be obtained commercially, and it takes a long time to synthesize the reagent. Recently, a derivatization method of fluoride in biological samples with 2(bromomethyl)naphthalene was developed and validated, and it was favorable for sensitive detection by GC–MS [21]. For all cases, the derivatization product is volatile enough to use headspace solid-phase microextraction (HS-SPME). HS-SPME is a popular sample treatment technique used to analyze volatile compounds and requires no solvent for the sample extraction, and is environmentally friendly.
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
The present study aimed to develop a highly sensitive HS-SPME GC–tandem mass spectrometry (MS-MS) method to detect fluoride in biological samples and to apply the new method to urine and plasma samples. This is the first HS-SPME-GC-MS-MS method being reported for fluoride analysis in biological samples. 2. Experimental 2.1. Materials All organic solvents used were HPLC grade. 2(Bromomethyl)naphthalene(96%), 2-fluoronaphthalene (99%), 15-crown-5-ether (98%) and potassium fluoride (99.9%) were obtained from Sigma–Aldrich (St. Louis, MO, USA). Commercially available SPME fibers, 50/30 m-divinylbenzene-carboxenpolydimethylsiloxane (DVB-CAR-PDMS), 100 m-polydimethylsiloxane (PDMS), 65 m-polydimethylsiloxane-divinylbenzene (PDMS-DVB), 85 m-polyacrylate (PA) and 85 m-carboxenpolydimethylsiloxane (CAR-PDMS), were purchased from Supelco (Bellefonte, PA, USA). The fiber was initially conditioned according to the instructions of the manufacturer to remove contaminants and to stabilize the SPME fibers. Conditioning was performed in an extra split/splitless port with helium carrier gas prior to each adsorption. The pure water used in this study was purified by a Milli-Q Reagent-Grade water system (ZD20) and had a resistivity of over 17 M. 2.2. Biological samples Urine samples were collected from 10 volunteers and plasma samples were collected from three volunteers before and after the intake of mineral water. Blank urine and plasma, in which the analytes were detected in the low concentration with our method, were used. Spiked samples of human urine and plasma were prepared with the addition of standard solutions at concentrations of 0.02–2.0 mg/L and the internal standard solution at a concentration of 1.0 mg/L. 2.3. Extraction/derivatization procedure Sample preparation (extraction and derivatization) was conducted in 10 ml headspace vials with screw caps. To a solution containing 1.0 ml of plasma or urine sample, 200 l of 2-(bromomethyl)naphthalene solution (10.0%), 100 l of 15crown-5-ether (1.0%), and 25 l of 1.0 mg/L 2-fluoronaphthalene as an internal standard were added. The pH was adjusted to 7.0 with 1.0 ml of 0.1 M phosphate buffer and the total volume of the solution was adjusted to 4.0 ml with pure water. Derivatization and adsorption was conducted simultaneously in a headspace vial with continuous shaking, and the derivatives were desorbed in the injection port and were passed onto the column for analysis. Derivatization was performed for different SPME adsorption times (10, 15, 20, 25, 30 and 35 min), at different reaction temperatures (40, 50, 60, 70, 80, 90 and 95 ◦ C), at different desorption temperatures (230, 250, 270, 290 and 310 ◦ C), at different concentrations of 2-(bromomethyl)naphthalene (0.1, 0.5, 1.0, 1.5 and 2.0%), and at different pH (5.0, 6.0, 7.0, 8.0, 9.0. 10.0 and 11.0). The optimum conditions for the derivatization of fluoride were determined by the amounts of 2-(fluoromethyl)naphthalene formed. 2.4. Apparatus All mass spectra were obtained with an Agilent 7890/7000B instrument. The ion source was operated in the electron ionization mode (EI; 70 eV). Full-scan mass spectra (m/z 40–400) were recorded for analyte identification. An HP-5MS capillary column
217
(30 m × 0.25 mm I.D. × 0.25 m film thickness) was used. Samples were injected in the pulsed splitless mode. The flow rate of helium as the carrier gas was 1.0 ml/min. The injector temperature was set at 290 ◦ C. The oven temperature programs were set as follows. The initial temperature of 50 ◦ C was held for 1 min and then increased to the final temperature of 250 ◦ C at 15 ◦ C/min. The MS-MS detection was performed in a multiple reaction monitoring (MRM) mode. Nitrogen was used as collision gas and collision energy was optimized for the derivatives of 2-(fluoromethyl)naphthalene and 2-fluoronaphthalene. Quantification of 2-(fluoromethyl)naphthalene was carried out monitoring the transition m/z 160 → 159 using collision energy of 25 V, whereas identification was achieved with the transition m/z 160 → 139 and 160 → 133 using collision energy of 40 and 35 V, respectively. The 2-fluoronaphthalene, used as internal standard, was quantified monitoring the transition m/z 146 → 120 using collision energy of 25 V, whereas identification was achieved with the transition m/z 146 → 99 and 146 → 96 using collision energy of 40 and 30 V. 2.5. Calibration and Quantification The calibration curve for fluoride was established by derivatization after adding 0, 10, 25 and 250 l of fluoride standard (2.0 mg/L in reagent water), 50, 100 l of fluoride standard (20 mg L−1 ), 200 l of 2-(bromomethyl)naphthalene solution (10.0%), 100 l of 15-crown-5-ether (10.0 mg/ml), and 25 l of 1.0 mg/L 2-fluoronaphthalene as an internal standard in 1.0 ml of blank urine or plasma. The total volume of the solution was adjusted to 4.0 ml with pure water. The corresponding concentrations of the standards were 0.02, 0.05, 0.5, 1.0 and 2.0 mg/L in urine or plasma. The spiked samples were derivatized by the method in the above reaction procedures and were injected into the GS–MS–MS system. The ions selected for quantitation were m/z 159 for 2(fluoromethyl)naphthalene and m/z 120 for 2-fluoronaphthalene considering sensitivity and selectivity. The ratio of the peak area of the standard to that of the internal standard was used to quantify the compound. The limit of detection (LOD) and limit of quantitation (LOQ) of the fluoride coupled derivatization and extraction method were calculated as 3.14 and 10 times the standard deviation obtained from the data of seven replicate measurements. 3. Results and discussion 3.1. Optimization of SPME fibers and derivatization conditions 2-(Bromomethyl)naphthalene was used for the derivatization of fluoride in biological samples in our preceding study [21]. In this study, we proposed its application to HS-SPME GS–MS–MS detection of fluoride through derivatization in biological samples. The fluoride is substituted with the bromide of 2(bromomethyl)naphthalene to form 2-fluoromethylnaphthalene, which is vaporized and adsorbed in SPME fibers and to be used for fluoride quantification using GC–MS–MS. To study the optimal conditions for the HS-SPME, the SPME fiber selection, the effects of the reagent concentrations, acidity, reaction temperature, reaction time and the desorption temperature were examined. Every experiment was performed at the concentration of 1.0 mg/L fluoride. Five SPME fibers were evaluated to select a suitable fiber for detecting fluoride. The adsorption efficiencies on the SPME fibers were evaluated by comparing the peak areas of the fluoride derivative. The highest efficiency among the five fibers was obtained using 85 m-CAR-PDMS, as shown in Fig. 1; therefore, CAR-PDMS was selected for detecting fluoride in biological samples. In the
218
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
derivative was maximized at 290 ◦ C. When 15-crown-5-ether as a phase transfer catalyst was used in the biological samples, the signal from fluoride increased approximately two times [21]. Therefore, 15-crown-5-ether was added to all biological samples. In light of these results, the optimal HS-SPME conditions were achieved with the selection of a CAR/PDMS fiber, 0.5% 2-(bromomethyl)naphthalene, 250 mg/L 15-crown-5-ether as a phase transfer catalyst, extraction/derivatization temperature of 95 ◦ C, heating time of 20 min, and pH of 7.0. 3.2. Chromatography and mass spectrometry
Fig. 1. Extraction efficiencies of fluoride in relation to various solid-phase microextraction fibers (this experiment was performed at a reaction time of 20 min and at 95 ◦ C).
study of the reaction yield of fluoride in relation to the amount of 2-(bromomethyl)naphthalene added in the biological sample, a maximum area of the fluoride derivative was achieved by 0.5% 2(bromomethyl)naphthalene (Fig. 2a). The derivatization was tested at various pH values from 5.0 to 11.0. The reaction showed good response at a pH value of 7.0. The response declined rapidly above or below pH 7.0, as shown in Fig. 2b; therefore, the pH of 7.0 was selected. The pH condition coincided with our previous study [21], indicating that the derivatization reaction and vaporization to the headspace were maximized at the selected pH. The reactivity of fluoride with 2-(bromomethyl)naphthalene in relation to the reaction temperature and time was studied. The formation of the derivative was studied for 20 min over the reaction temperature range of 40, 50, 60, 70, 80, 90 and 95 ◦ C and then over the reaction time range of 2, 5, 8, 10, 20, 30, 40 and 50 min at 95 ◦ C. The results indicated that the reaction was maximized at approximately 20 min at 95 ◦ C (Fig. 3). Although the increase of the response continues at reaction temperatures over 95 ◦ C, it is desirable to keep the temperature under the boiling point of water. The desorption temperature was tested in the range of 270 to 290 ◦ C. The response of the
The optimum HS-SPME conditions were applied to the analysis of fluoride in biological samples. Fig. 4 shows a GS–MS–MS chromatogram after the derivatization-HS-SPME of fluoride. For the GC separation of the derivative, the use of a non-polar stationary phase (HP-5MS) was efficient. The derivative of fluoride showed a sharp peak, and the ratio of the peak area of the standard to that of the internal standard was used to quantify the analyte. The full scan mass spectra of 2-(fluoromethyl)naphthalene and 2-fluoronaphthalene by electron ionization at 70 eV are shown in Fig. 5. The molecular ion at m/z 160 and the diagnostic ions at m/z 141, 139, 133, 127 and 115 in Fig. 5a indicated 2(fluoromethyl)naphthalene. The molecular ion at m/z 146 and the diagnostic ions at m/z 126, 125 and 120 in Fig. 5b indicated 2fluoronaphthalene. 3.3. Validation of the assay The combination of a high derivatization yield and the high sensitivity of the derivative by HS-SPME EI–MS–MS (MRM) permitted the detection of fluoride at concentrations well below or similar to those reported previously. The LOD and LOQ in this study were calculated as 11 and 34 g/L in urine and 9 and 28 g/L in plasma using a 1.0 ml sample volume. Examination of the typical standard curve by computing a regression line of the peak area ratios of 2-(fluoromethyl)naphthalene to 2-fluoronaphthalene on the concentration using a least-squares fit demonstrated a linear
Fig. 2. Reaction yield of fluoride in relation to the concentration of 2- (bromomethyl)naphthalene and pH in 1.0 ml of urine (this experiment was performed at a reaction time of 20 min and at 95 ◦ C).
Fig. 3. Effect of the reaction temperature (A) and time (B) on the reaction of fluoride with 2-(bromomethyl)naphthalene.
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
219
Fig. 4. Gas chromatography–tandem mass spectrometry chromatograms of the blank biological samples (A), a sample spiked at the concentration of 0.05 mg/L fluoride (B), and real samples detected in the concentration of 0.48 mg/L in urine and 0.07 mg/L in plasma (C) (2-fluoronaphthalene (IS) = 5.9 min; 2-(fluoromethyl)naphthalene = 7.8 min).
relationship with a correlation coefficient of 0.999. The line of best fit for fluoride is y = 0.5355x–0.0035 over a range of 0.02–2.0 mg/L, where x is the analyte concentration (mg/L) and y is the peak area ratio of the analyte to the internal standard. The trueness and accuracy were assessed by determining the recovery in spiked samples. Intra-day accuracy was evaluated by five spiked samples at concentrations of 0.1 and 1.0 mg/L, and interday accuracy was determined by the recovery on five different days. The trueness and accuracy of the assay were good, as shown in Table 1. The trueness was in the range of 101.7–105.6%, and precision of the assay was less than 7.7%. To test the matrix influence on the determination of fluoride, the accuracy and precision of the method were assessed in real biological samples, in which the analyte was at low concentration. No interference peak was found near the retention time of the derivative. 3.4. Application To test the applicability of the proposed method in the analysis of real samples, the optimized approach was applied to urine and plasma samples. Whole-day urine samples were collected from
10 volunteers after intake of mineral water containing fluoride at 0.7 mg/L and from the same volunteers after intake of mineral water containing no fluoride for 10 days. Plasma samples were collected from three volunteers before and after intake of mineral water containing fluoride at 0.7 mg/L. Fluoride was detected in a concentration range of 0.44–1.08 mg/L (mean 0.83 ± 0.23 mg/L) in the ten urine samples after intake of mineral water containing fluoride and in a concentration range of 0.06–0.58 mg/L (mean 0.31 ± 0.16 mg/L) in the ten urine samples after intake of mineral water containing no fluoride. Otherwise, fluoride was detected in a concentration range of 0.033–0.037 mg/L (mean 0.035 ± 0.002 mg/L) and 0.041–0.075 mg/L (mean 0.060 ± 0.017 mg/L) in the three plasma samples before and after intake of mineral water containing fluoride, respectively. The fluoride concentrations in the urine and plasma collected after intake of the mineral water were approximately 2.7 and 1.7 times higher than those collected after intake of mineral water containing no fluoride, respectively. The proposed method sensitively determines the fluoride concentration in biological samples by HSSPME GS–MS–MS and permits reliable analysis of trace levels of fluoride in biological samples.
220
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
Fig. 5. The full mass spectra of 2-fluoromethylnaphthalene (A) and 2-fluoronaphthalene (B).
Table 1 Intra- and inter-day laboratory trueness and precision results for the analysis of fluoride in plasma or urine (n = 5). Matrix
Matrix blank (mg/L)
Spiked Conc. (mg/L)
Intra-day measured value
Inter-day measured value
Mean ± SD (mg/L)
Trueness (%)
Precision (%)
Mean ± SD (mg/L)
Trueness (%)
Precision (%)
Plasma
nd
0.1 1.0
0.095 ± 0.002 1.004 ± 0.043
95.0 100.4
1.8 4.3
0.102 ± 0.008 1.056 ± 0.047
101.7 105.6
7.7 4.4
Urine
nd
0.1 1.0
0.098 ± 0.007 0.992 ± 0.048
97.7 99.2
6.7 4.8
0.104 ± 0.006 1.039 ± 0.036
103.7 103.9
6.0 3.4
4. Conclusion
References
In this paper, we present a simple, amenable to automation and environmentally friendly to detect fluoride in biological samples. Derivatization was performed by the reaction of fluoride and 2-(bromomethyl)naphthalene in a headspace vial. The formed volatile 2-(fluoromethyl)naphthalene was vaporized and simultaneously adsorbed in an SPME fiber and then desorbed in GC–MS-MS. All reagents were spiked in the preparation step, and the analysis was performed automatically after capping. The proposed method can be used for the routine analysis of fluoride in biological samples. The major advantages of this method are as follows: (1) This method has high sensitivity and low interference due to use HS-SPME and GC–MS–MS. (2) This method offers easier manipulation of samples and shorter analysis duration and is amenable to automation. (3) This method requires no solvent for the sample extraction and is environmentally friendly. The developed method provides an important tool for evaluating trace levels of fluoride in biological samples.
[1] USEPA, Toxicological profile for fluorides, hydrogen fluoride, and fluorine (TP-91/1), US Department of Health and Human Services, Agency for toxic substances and diseases registry, Atlanta, Georgia (1993). [2] R.L. Vogt, L. Witherell, D. LaRue, D.N. Klaucke, Acute fluoride poisoning associated with an on-site fluoridator in a Vermont elementary school, Am. J. Public Health 72 (1982) 1168–1169. [3] B.D. Gessner, M. Beller, J.P. Middaugh, G.M. Whitford, Acute fluoride poisoning from a public water system, N. Engl. J. Med. 330 (1994) 95–99. [4] A. Poklis, M.A. Mackell, Disposition of fluoride in a fatal case of unsuspected sodium fluoride poisoning, Forensic Sci. Int. 41 (1989) 55–59. [5] P.A. Monsour, B.J. Kruger, A.F. Petrie, J.L. McNee, Acute fluoride poisoning after ingestion of sodium fluoride tablets, Med. J. Aust. 141 (1984) 503–505. [6] G.M. Whitford, Acute and chronic fluoride toxicity, J. Dent. Res. 71 (1992) 1249–1254. [7] J. Braun, H. Stöß, A. Zober, Intoxication following the inhalation of hydrogen fluoride, Arch. Toxicol. 56 (1984) 50–54. [8] M. Meldrum, Toxicology of hydrogen fluoride in relation to major accident hazards, Regul. Toxicol. Pharm. 30 (1999) 110–116. [9] A.L.H. Müller, C.C. Müller, F.G. Antes, J.S. Barin, V.L. Dressler, E.M.M. Flores, E.I. Müller, Determination of bromide, chloride, and fluoride in cigarette tobacco by ion chromatography after microwave-induced combustion, Anal. Lett. 45 (2012) 1004–1015. [10] V. Kontozova-Deutsch, F. Deutsch, L. Bencs, A. Krata, R. Van Grieken, K. De Wael, Optimization of the ion chromatographic quantification of airborne fluoride, acetate and formate in the Metropolitan Museum of Art, New York, Talanta 86 (2011) 372–376. [11] H. Yiping, W. Caiyun, Ion chromatography for rapid and sensitive determination of fluoride in milk after headspace single-drop microextraction with in situ generation of volatile hydrogen fluoride, Anal. Chim. Acta 661 (2010) 161–166. [12] X.R. Xu, H.B. Li, J.-D. Gu, K.J. Paeng, Determination of fluoride in water by reversed-phase high-performance liquid chromatography using
Acknowledgments This work was supported by the South Korean Ministry of Environment (MOE) as part of “the Environmental Health Action Program.”
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
[13]
[14]
[15]
[16]
F− -La3+ -alizarin complexone ternary complex, Chromatographia 59 (2004) 745–747. G. Yamamoto, K. Yoshitake, T. Kimura, T. Ando, Gas chromatographic determination of traces of fluoride with several alkylsilane extractants, Anal. Chim. Acta 222 (1989) 121–126. W. Tashkov, I. Benchev, N. Rizov, A. Kolarska, Fluoride determination in fluorinated milk by headspace gas chromatography, Chromatographia 29 (1990) 544–546. G. Wejnerowska, A. Karczmarek, J. Gaca, Determination of fluoride in toothpaste using headspace solid-phase microextraction and gas chromatographyflame ionization detection, J. Chromatogr. A 1150 (2007) 173–177. D.-M. Hui, M. Minami, Monitoring of fluorine in urine samples of patients involved in the Tokyo sarin disaster, in connection with the detection of other decomposition products of sarin and the by-products generated during sarin synthesis, Clin. Chim. Acta 302 (2000) 171–188.
221
[17] M. Kaykhaii, M. Ghalehno, Rapid and sensitive determination of fluoride in toothpaste and water samples using headspace single drop microextractiongas chromatography, Anal. Methods 5 (2013) 5622–5626. [18] S. Kage, K. Kudo, N. Nishida, H. Ikeda, N. Yoshioka, N. Ikeda, Determination of fluoride in human whole blood and urine by gas chromatography-mass spectrometry, Forensic Toxicol. 26 (2008) 23–26. [19] M. Sakayanagi, Y. Yamada, C. Sakabe, K. Watanabe, Y. Harigaya, Identification of inorganic anions by gas chromatography/mass spectrometry, Forensic Sci. Int. 157 (2006) 134–143. [20] E. Pagliano, J. Meija, J. Ding, R.E. Sturgeon, A. D’Ulivo, Z. Mester, Novel ethylderivatization approach for the determination of fluoride by headspace gas chromatography/mass spectrometry, Anal. Chem. 85 (2013) 877–881. [21] S.M. Kwon, H.S. Shin, Sensitive determination of fluoride in biological samples by gas chromatography-mass spectrometry after derivatization with 2-(bromomethyl) naphthalene, Anal. Chim. Acta 852 (2014) 162–167.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Short communication
Simple determination of fluoride in biological samples by headspace solid-phase microextraction and gas chromatography–tandem mass spectrometry Sun-Myung Kwon a , Ho-Sang Shin b,∗ a b
Department of Environmental Science, Kongju National University, Kongju, 314-701, South Korea Department of Environmental Education, Kongju National University, Kongju 314-701, South Korea
a r t i c l e
i n f o
Article history: Received 7 April 2015 Received in revised form 27 June 2015 Accepted 29 June 2015 Available online 2 July 2015 Keywords: Gas chromatography–tandem mass spectrometry Headspace solid phase microextraction Biological samples Fluoride
a b s t r a c t A simple and convenient method to detect fluoride in biological samples was developed. This method was based on derivatization with 2-(bromomethyl)naphthalene, headspace solid phase microextraction (HSSPME) in a vial, and gas chromatography-tandem mass spectrometric detection. The HS-SPME parameters were optimized as follows: selection of CAR/PDMS fiber, 0.5% 2-(bromomethyl)naphthalene, 250 mg/L 15-crown-5-ether as a phase transfer catalyst, extraction and derivatization temperature of 95 ◦ C, heating time of 20 min and pH of 7.0. Under the established conditions, the lowest limits of detection were 9 and 11 g/L in 1.0 ml of plasma and urine, respectively, and the intra- and inter-day relative standard deviation was less than 7.7% at concentrations of 0.1 and 1.0 mg/L. The calibration curve showed good linearity of plasma and urine with r = 0.9990 and r = 0.9992, respectively. This method is simple, amenable to automation and environmentally friendly. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Fluoride ion (F− ) is a ubiquitous element present in various environmental matrices [1]. Many communities add fluoride to their drinking water to promote dental health, and hydrogen fluoride is used in the production of aluminum and chlorofluorocarbons. Fluoride poisoning occurs by inorganic fluoride salts in public water systems [2,3], ingestion of fluoride-containing products [4–6] and hydrogen fluoride exposure [7,8]. Acute inhalation exposure to gaseous hydrogen fluoride can cause respiratory damage, such as severe irritation and pulmonary edema, in humans. Severe ocular irritation and dermal burns may occur following eye or skin exposure in humans [6]. Otherwise, chronic exposure to fluoride at low levels has the beneficial effect of dental cavity prevention and may also be useful for the treatment of osteoporosis [6–8]. Many methods have been proposed for the chromatographic determination of fluoride in various matrices. Ion chromatography [9–11] is commonly used for the determination of fluoride, but the sensitivity is not adequate to detect g/L levels in biological samples. High performance liquid chromatography (HPLC) methods are performed using derivatives of the F− -La3+ -alizarin
∗ Corresponding author. Tel.: +82 41 850 8811; fax: +82 41 850 8998. E-mail address: [email protected] (H.-S. Shin). http://dx.doi.org/10.1016/j.chroma.2015.06.066 0021-9673/© 2015 Elsevier B.V. All rights reserved.
complexone ternary complex [12], and gas chromatography (GC) methods are performed following the reaction with silylation reagents, such as trimethylchlorosilane and trimethylimidazolesilane [13–17]. When the derivatized fluoride is identified based on its retention time in the case of HPLC and GC, a high rate of false-identification may be introduced, especially in the analysis of complex samples, such as biological samples. GC-mass spectrometry (MS) methods were published for the determination of fluoride in human whole blood and urine after derivatization with pentafluorobenzyl bromide [18] and pentafluorobenzyl p-toluenesulphonate [19], but these methods do not have a sufficient detection limit to detect fluoride in real samples. Headspace (HS) GC–MS following ethylation with triethyloxoniumtetrachloroferrate (III) was recently developed for the sensitive analysis of fluoride in tap water, seawater and urine [20]. The derivatization reagent, however, cannot be obtained commercially, and it takes a long time to synthesize the reagent. Recently, a derivatization method of fluoride in biological samples with 2(bromomethyl)naphthalene was developed and validated, and it was favorable for sensitive detection by GC–MS [21]. For all cases, the derivatization product is volatile enough to use headspace solid-phase microextraction (HS-SPME). HS-SPME is a popular sample treatment technique used to analyze volatile compounds and requires no solvent for the sample extraction, and is environmentally friendly.
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
The present study aimed to develop a highly sensitive HS-SPME GC–tandem mass spectrometry (MS-MS) method to detect fluoride in biological samples and to apply the new method to urine and plasma samples. This is the first HS-SPME-GC-MS-MS method being reported for fluoride analysis in biological samples. 2. Experimental 2.1. Materials All organic solvents used were HPLC grade. 2(Bromomethyl)naphthalene(96%), 2-fluoronaphthalene (99%), 15-crown-5-ether (98%) and potassium fluoride (99.9%) were obtained from Sigma–Aldrich (St. Louis, MO, USA). Commercially available SPME fibers, 50/30 m-divinylbenzene-carboxenpolydimethylsiloxane (DVB-CAR-PDMS), 100 m-polydimethylsiloxane (PDMS), 65 m-polydimethylsiloxane-divinylbenzene (PDMS-DVB), 85 m-polyacrylate (PA) and 85 m-carboxenpolydimethylsiloxane (CAR-PDMS), were purchased from Supelco (Bellefonte, PA, USA). The fiber was initially conditioned according to the instructions of the manufacturer to remove contaminants and to stabilize the SPME fibers. Conditioning was performed in an extra split/splitless port with helium carrier gas prior to each adsorption. The pure water used in this study was purified by a Milli-Q Reagent-Grade water system (ZD20) and had a resistivity of over 17 M. 2.2. Biological samples Urine samples were collected from 10 volunteers and plasma samples were collected from three volunteers before and after the intake of mineral water. Blank urine and plasma, in which the analytes were detected in the low concentration with our method, were used. Spiked samples of human urine and plasma were prepared with the addition of standard solutions at concentrations of 0.02–2.0 mg/L and the internal standard solution at a concentration of 1.0 mg/L. 2.3. Extraction/derivatization procedure Sample preparation (extraction and derivatization) was conducted in 10 ml headspace vials with screw caps. To a solution containing 1.0 ml of plasma or urine sample, 200 l of 2-(bromomethyl)naphthalene solution (10.0%), 100 l of 15crown-5-ether (1.0%), and 25 l of 1.0 mg/L 2-fluoronaphthalene as an internal standard were added. The pH was adjusted to 7.0 with 1.0 ml of 0.1 M phosphate buffer and the total volume of the solution was adjusted to 4.0 ml with pure water. Derivatization and adsorption was conducted simultaneously in a headspace vial with continuous shaking, and the derivatives were desorbed in the injection port and were passed onto the column for analysis. Derivatization was performed for different SPME adsorption times (10, 15, 20, 25, 30 and 35 min), at different reaction temperatures (40, 50, 60, 70, 80, 90 and 95 ◦ C), at different desorption temperatures (230, 250, 270, 290 and 310 ◦ C), at different concentrations of 2-(bromomethyl)naphthalene (0.1, 0.5, 1.0, 1.5 and 2.0%), and at different pH (5.0, 6.0, 7.0, 8.0, 9.0. 10.0 and 11.0). The optimum conditions for the derivatization of fluoride were determined by the amounts of 2-(fluoromethyl)naphthalene formed. 2.4. Apparatus All mass spectra were obtained with an Agilent 7890/7000B instrument. The ion source was operated in the electron ionization mode (EI; 70 eV). Full-scan mass spectra (m/z 40–400) were recorded for analyte identification. An HP-5MS capillary column
217
(30 m × 0.25 mm I.D. × 0.25 m film thickness) was used. Samples were injected in the pulsed splitless mode. The flow rate of helium as the carrier gas was 1.0 ml/min. The injector temperature was set at 290 ◦ C. The oven temperature programs were set as follows. The initial temperature of 50 ◦ C was held for 1 min and then increased to the final temperature of 250 ◦ C at 15 ◦ C/min. The MS-MS detection was performed in a multiple reaction monitoring (MRM) mode. Nitrogen was used as collision gas and collision energy was optimized for the derivatives of 2-(fluoromethyl)naphthalene and 2-fluoronaphthalene. Quantification of 2-(fluoromethyl)naphthalene was carried out monitoring the transition m/z 160 → 159 using collision energy of 25 V, whereas identification was achieved with the transition m/z 160 → 139 and 160 → 133 using collision energy of 40 and 35 V, respectively. The 2-fluoronaphthalene, used as internal standard, was quantified monitoring the transition m/z 146 → 120 using collision energy of 25 V, whereas identification was achieved with the transition m/z 146 → 99 and 146 → 96 using collision energy of 40 and 30 V. 2.5. Calibration and Quantification The calibration curve for fluoride was established by derivatization after adding 0, 10, 25 and 250 l of fluoride standard (2.0 mg/L in reagent water), 50, 100 l of fluoride standard (20 mg L−1 ), 200 l of 2-(bromomethyl)naphthalene solution (10.0%), 100 l of 15-crown-5-ether (10.0 mg/ml), and 25 l of 1.0 mg/L 2-fluoronaphthalene as an internal standard in 1.0 ml of blank urine or plasma. The total volume of the solution was adjusted to 4.0 ml with pure water. The corresponding concentrations of the standards were 0.02, 0.05, 0.5, 1.0 and 2.0 mg/L in urine or plasma. The spiked samples were derivatized by the method in the above reaction procedures and were injected into the GS–MS–MS system. The ions selected for quantitation were m/z 159 for 2(fluoromethyl)naphthalene and m/z 120 for 2-fluoronaphthalene considering sensitivity and selectivity. The ratio of the peak area of the standard to that of the internal standard was used to quantify the compound. The limit of detection (LOD) and limit of quantitation (LOQ) of the fluoride coupled derivatization and extraction method were calculated as 3.14 and 10 times the standard deviation obtained from the data of seven replicate measurements. 3. Results and discussion 3.1. Optimization of SPME fibers and derivatization conditions 2-(Bromomethyl)naphthalene was used for the derivatization of fluoride in biological samples in our preceding study [21]. In this study, we proposed its application to HS-SPME GS–MS–MS detection of fluoride through derivatization in biological samples. The fluoride is substituted with the bromide of 2(bromomethyl)naphthalene to form 2-fluoromethylnaphthalene, which is vaporized and adsorbed in SPME fibers and to be used for fluoride quantification using GC–MS–MS. To study the optimal conditions for the HS-SPME, the SPME fiber selection, the effects of the reagent concentrations, acidity, reaction temperature, reaction time and the desorption temperature were examined. Every experiment was performed at the concentration of 1.0 mg/L fluoride. Five SPME fibers were evaluated to select a suitable fiber for detecting fluoride. The adsorption efficiencies on the SPME fibers were evaluated by comparing the peak areas of the fluoride derivative. The highest efficiency among the five fibers was obtained using 85 m-CAR-PDMS, as shown in Fig. 1; therefore, CAR-PDMS was selected for detecting fluoride in biological samples. In the
218
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
derivative was maximized at 290 ◦ C. When 15-crown-5-ether as a phase transfer catalyst was used in the biological samples, the signal from fluoride increased approximately two times [21]. Therefore, 15-crown-5-ether was added to all biological samples. In light of these results, the optimal HS-SPME conditions were achieved with the selection of a CAR/PDMS fiber, 0.5% 2-(bromomethyl)naphthalene, 250 mg/L 15-crown-5-ether as a phase transfer catalyst, extraction/derivatization temperature of 95 ◦ C, heating time of 20 min, and pH of 7.0. 3.2. Chromatography and mass spectrometry
Fig. 1. Extraction efficiencies of fluoride in relation to various solid-phase microextraction fibers (this experiment was performed at a reaction time of 20 min and at 95 ◦ C).
study of the reaction yield of fluoride in relation to the amount of 2-(bromomethyl)naphthalene added in the biological sample, a maximum area of the fluoride derivative was achieved by 0.5% 2(bromomethyl)naphthalene (Fig. 2a). The derivatization was tested at various pH values from 5.0 to 11.0. The reaction showed good response at a pH value of 7.0. The response declined rapidly above or below pH 7.0, as shown in Fig. 2b; therefore, the pH of 7.0 was selected. The pH condition coincided with our previous study [21], indicating that the derivatization reaction and vaporization to the headspace were maximized at the selected pH. The reactivity of fluoride with 2-(bromomethyl)naphthalene in relation to the reaction temperature and time was studied. The formation of the derivative was studied for 20 min over the reaction temperature range of 40, 50, 60, 70, 80, 90 and 95 ◦ C and then over the reaction time range of 2, 5, 8, 10, 20, 30, 40 and 50 min at 95 ◦ C. The results indicated that the reaction was maximized at approximately 20 min at 95 ◦ C (Fig. 3). Although the increase of the response continues at reaction temperatures over 95 ◦ C, it is desirable to keep the temperature under the boiling point of water. The desorption temperature was tested in the range of 270 to 290 ◦ C. The response of the
The optimum HS-SPME conditions were applied to the analysis of fluoride in biological samples. Fig. 4 shows a GS–MS–MS chromatogram after the derivatization-HS-SPME of fluoride. For the GC separation of the derivative, the use of a non-polar stationary phase (HP-5MS) was efficient. The derivative of fluoride showed a sharp peak, and the ratio of the peak area of the standard to that of the internal standard was used to quantify the analyte. The full scan mass spectra of 2-(fluoromethyl)naphthalene and 2-fluoronaphthalene by electron ionization at 70 eV are shown in Fig. 5. The molecular ion at m/z 160 and the diagnostic ions at m/z 141, 139, 133, 127 and 115 in Fig. 5a indicated 2(fluoromethyl)naphthalene. The molecular ion at m/z 146 and the diagnostic ions at m/z 126, 125 and 120 in Fig. 5b indicated 2fluoronaphthalene. 3.3. Validation of the assay The combination of a high derivatization yield and the high sensitivity of the derivative by HS-SPME EI–MS–MS (MRM) permitted the detection of fluoride at concentrations well below or similar to those reported previously. The LOD and LOQ in this study were calculated as 11 and 34 g/L in urine and 9 and 28 g/L in plasma using a 1.0 ml sample volume. Examination of the typical standard curve by computing a regression line of the peak area ratios of 2-(fluoromethyl)naphthalene to 2-fluoronaphthalene on the concentration using a least-squares fit demonstrated a linear
Fig. 2. Reaction yield of fluoride in relation to the concentration of 2- (bromomethyl)naphthalene and pH in 1.0 ml of urine (this experiment was performed at a reaction time of 20 min and at 95 ◦ C).
Fig. 3. Effect of the reaction temperature (A) and time (B) on the reaction of fluoride with 2-(bromomethyl)naphthalene.
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
219
Fig. 4. Gas chromatography–tandem mass spectrometry chromatograms of the blank biological samples (A), a sample spiked at the concentration of 0.05 mg/L fluoride (B), and real samples detected in the concentration of 0.48 mg/L in urine and 0.07 mg/L in plasma (C) (2-fluoronaphthalene (IS) = 5.9 min; 2-(fluoromethyl)naphthalene = 7.8 min).
relationship with a correlation coefficient of 0.999. The line of best fit for fluoride is y = 0.5355x–0.0035 over a range of 0.02–2.0 mg/L, where x is the analyte concentration (mg/L) and y is the peak area ratio of the analyte to the internal standard. The trueness and accuracy were assessed by determining the recovery in spiked samples. Intra-day accuracy was evaluated by five spiked samples at concentrations of 0.1 and 1.0 mg/L, and interday accuracy was determined by the recovery on five different days. The trueness and accuracy of the assay were good, as shown in Table 1. The trueness was in the range of 101.7–105.6%, and precision of the assay was less than 7.7%. To test the matrix influence on the determination of fluoride, the accuracy and precision of the method were assessed in real biological samples, in which the analyte was at low concentration. No interference peak was found near the retention time of the derivative. 3.4. Application To test the applicability of the proposed method in the analysis of real samples, the optimized approach was applied to urine and plasma samples. Whole-day urine samples were collected from
10 volunteers after intake of mineral water containing fluoride at 0.7 mg/L and from the same volunteers after intake of mineral water containing no fluoride for 10 days. Plasma samples were collected from three volunteers before and after intake of mineral water containing fluoride at 0.7 mg/L. Fluoride was detected in a concentration range of 0.44–1.08 mg/L (mean 0.83 ± 0.23 mg/L) in the ten urine samples after intake of mineral water containing fluoride and in a concentration range of 0.06–0.58 mg/L (mean 0.31 ± 0.16 mg/L) in the ten urine samples after intake of mineral water containing no fluoride. Otherwise, fluoride was detected in a concentration range of 0.033–0.037 mg/L (mean 0.035 ± 0.002 mg/L) and 0.041–0.075 mg/L (mean 0.060 ± 0.017 mg/L) in the three plasma samples before and after intake of mineral water containing fluoride, respectively. The fluoride concentrations in the urine and plasma collected after intake of the mineral water were approximately 2.7 and 1.7 times higher than those collected after intake of mineral water containing no fluoride, respectively. The proposed method sensitively determines the fluoride concentration in biological samples by HSSPME GS–MS–MS and permits reliable analysis of trace levels of fluoride in biological samples.
220
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
Fig. 5. The full mass spectra of 2-fluoromethylnaphthalene (A) and 2-fluoronaphthalene (B).
Table 1 Intra- and inter-day laboratory trueness and precision results for the analysis of fluoride in plasma or urine (n = 5). Matrix
Matrix blank (mg/L)
Spiked Conc. (mg/L)
Intra-day measured value
Inter-day measured value
Mean ± SD (mg/L)
Trueness (%)
Precision (%)
Mean ± SD (mg/L)
Trueness (%)
Precision (%)
Plasma
nd
0.1 1.0
0.095 ± 0.002 1.004 ± 0.043
95.0 100.4
1.8 4.3
0.102 ± 0.008 1.056 ± 0.047
101.7 105.6
7.7 4.4
Urine
nd
0.1 1.0
0.098 ± 0.007 0.992 ± 0.048
97.7 99.2
6.7 4.8
0.104 ± 0.006 1.039 ± 0.036
103.7 103.9
6.0 3.4
4. Conclusion
References
In this paper, we present a simple, amenable to automation and environmentally friendly to detect fluoride in biological samples. Derivatization was performed by the reaction of fluoride and 2-(bromomethyl)naphthalene in a headspace vial. The formed volatile 2-(fluoromethyl)naphthalene was vaporized and simultaneously adsorbed in an SPME fiber and then desorbed in GC–MS-MS. All reagents were spiked in the preparation step, and the analysis was performed automatically after capping. The proposed method can be used for the routine analysis of fluoride in biological samples. The major advantages of this method are as follows: (1) This method has high sensitivity and low interference due to use HS-SPME and GC–MS–MS. (2) This method offers easier manipulation of samples and shorter analysis duration and is amenable to automation. (3) This method requires no solvent for the sample extraction and is environmentally friendly. The developed method provides an important tool for evaluating trace levels of fluoride in biological samples.
[1] USEPA, Toxicological profile for fluorides, hydrogen fluoride, and fluorine (TP-91/1), US Department of Health and Human Services, Agency for toxic substances and diseases registry, Atlanta, Georgia (1993). [2] R.L. Vogt, L. Witherell, D. LaRue, D.N. Klaucke, Acute fluoride poisoning associated with an on-site fluoridator in a Vermont elementary school, Am. J. Public Health 72 (1982) 1168–1169. [3] B.D. Gessner, M. Beller, J.P. Middaugh, G.M. Whitford, Acute fluoride poisoning from a public water system, N. Engl. J. Med. 330 (1994) 95–99. [4] A. Poklis, M.A. Mackell, Disposition of fluoride in a fatal case of unsuspected sodium fluoride poisoning, Forensic Sci. Int. 41 (1989) 55–59. [5] P.A. Monsour, B.J. Kruger, A.F. Petrie, J.L. McNee, Acute fluoride poisoning after ingestion of sodium fluoride tablets, Med. J. Aust. 141 (1984) 503–505. [6] G.M. Whitford, Acute and chronic fluoride toxicity, J. Dent. Res. 71 (1992) 1249–1254. [7] J. Braun, H. Stöß, A. Zober, Intoxication following the inhalation of hydrogen fluoride, Arch. Toxicol. 56 (1984) 50–54. [8] M. Meldrum, Toxicology of hydrogen fluoride in relation to major accident hazards, Regul. Toxicol. Pharm. 30 (1999) 110–116. [9] A.L.H. Müller, C.C. Müller, F.G. Antes, J.S. Barin, V.L. Dressler, E.M.M. Flores, E.I. Müller, Determination of bromide, chloride, and fluoride in cigarette tobacco by ion chromatography after microwave-induced combustion, Anal. Lett. 45 (2012) 1004–1015. [10] V. Kontozova-Deutsch, F. Deutsch, L. Bencs, A. Krata, R. Van Grieken, K. De Wael, Optimization of the ion chromatographic quantification of airborne fluoride, acetate and formate in the Metropolitan Museum of Art, New York, Talanta 86 (2011) 372–376. [11] H. Yiping, W. Caiyun, Ion chromatography for rapid and sensitive determination of fluoride in milk after headspace single-drop microextraction with in situ generation of volatile hydrogen fluoride, Anal. Chim. Acta 661 (2010) 161–166. [12] X.R. Xu, H.B. Li, J.-D. Gu, K.J. Paeng, Determination of fluoride in water by reversed-phase high-performance liquid chromatography using
Acknowledgments This work was supported by the South Korean Ministry of Environment (MOE) as part of “the Environmental Health Action Program.”
S.-M. Kwon, H.-S. Shin / J. Chromatogr. A 1407 (2015) 216–221
[13]
[14]
[15]
[16]
F− -La3+ -alizarin complexone ternary complex, Chromatographia 59 (2004) 745–747. G. Yamamoto, K. Yoshitake, T. Kimura, T. Ando, Gas chromatographic determination of traces of fluoride with several alkylsilane extractants, Anal. Chim. Acta 222 (1989) 121–126. W. Tashkov, I. Benchev, N. Rizov, A. Kolarska, Fluoride determination in fluorinated milk by headspace gas chromatography, Chromatographia 29 (1990) 544–546. G. Wejnerowska, A. Karczmarek, J. Gaca, Determination of fluoride in toothpaste using headspace solid-phase microextraction and gas chromatographyflame ionization detection, J. Chromatogr. A 1150 (2007) 173–177. D.-M. Hui, M. Minami, Monitoring of fluorine in urine samples of patients involved in the Tokyo sarin disaster, in connection with the detection of other decomposition products of sarin and the by-products generated during sarin synthesis, Clin. Chim. Acta 302 (2000) 171–188.
221
[17] M. Kaykhaii, M. Ghalehno, Rapid and sensitive determination of fluoride in toothpaste and water samples using headspace single drop microextractiongas chromatography, Anal. Methods 5 (2013) 5622–5626. [18] S. Kage, K. Kudo, N. Nishida, H. Ikeda, N. Yoshioka, N. Ikeda, Determination of fluoride in human whole blood and urine by gas chromatography-mass spectrometry, Forensic Toxicol. 26 (2008) 23–26. [19] M. Sakayanagi, Y. Yamada, C. Sakabe, K. Watanabe, Y. Harigaya, Identification of inorganic anions by gas chromatography/mass spectrometry, Forensic Sci. Int. 157 (2006) 134–143. [20] E. Pagliano, J. Meija, J. Ding, R.E. Sturgeon, A. D’Ulivo, Z. Mester, Novel ethylderivatization approach for the determination of fluoride by headspace gas chromatography/mass spectrometry, Anal. Chem. 85 (2013) 877–881. [21] S.M. Kwon, H.S. Shin, Sensitive determination of fluoride in biological samples by gas chromatography-mass spectrometry after derivatization with 2-(bromomethyl) naphthalene, Anal. Chim. Acta 852 (2014) 162–167.
