1989
J. Sep. Sci. 2014, 37, 1989–1995
Vaida Smitiene Ivona Semasko Vida Vickackaite Department of Analytical and Environmental Chemistry, Vilnius University, Vilnius, Lithuania Received January 27, 2014 Revised April 8, 2014 Accepted April 25, 2014
Research Article
Speciation of methyltins by dispersive liquid–liquid microextraction and gas chromatography with mass spectrometry Dispersive liquid–liquid microextraction in combination with an in situ derivatization is suggested for methyltin compound sampling and preconcentration from water solutions. The derivatization was carried out with sodium tetraethylborate at pH 3. The effects of extraction and disperser solvents type, volume, and extraction time on the extraction efficiency were investigated. 1,2-Dichlorobenzene was used as an extraction solvent and ethanol was used as a disperser solvent. The calibration graphs for all the analytes were linear up to 2 g (Sn) L−1 , correlation coefficients were 0.998–0.999, LODs were 0.13, 0.05, and 0.06 ng (Sn) L−1 for trimethyltin, DMT, and monomethyltin, respectively. Repeatabilities of the results were acceptable with RSDs up to 12.1%. A possibility to apply the proposed method for methyltin compound determination in water samples was demonstrated. Keywords: Gas chromatography / Methyltin compounds / Microextraction / Water samples DOI 10.1002/jssc.201400074
1 Introduction Organometallic tin compounds have a wide industrial application. Due to their stability and transparency, organotin compounds are used as heat and light stabilizers in polyvinyl chloride (PVC) products. There are three major types of tin stabilizers—octyltin, butyltin, and methyltin. Among them, methyltin has gradually become the dominant type [1]. Monoand disubstituted methyltin compounds are used as stabilizing additives in PVC plastics [2]. Dimethyltin (DMT) chloride is also used to produce transparent conductive films for liquid crystal panels [3]. Trimethyltin (TMT) is a low-level by-product of DMT production [1]. In the environment methyltin compounds can also originate from the biomethylation of inorganic tin [2, 4, 5]. Methyltin chlorides are released from PVC products in landfills [2], they can leach from PVC water distribution lines and thus contaminate water, food, and various ecosystems [4]. Methylated tin is much more toxic than inorganic tin. Moreover, the toxicity of methyltin compounds is strongly dependent on the species. TMT is the most toxic, it is a well-known potent neurotoxin and causes neuropathalogy in rodents and humans [6]. The neurotoxic potential of monomethyltin (MMT) and DMT is lower, however, these compounds have also been found to exert neurotoxic efCorrespondence: Dr. Vida Vickackaite, Department of Analytical and Environmental Chemistry, Vilnius University, Naugarduko 24, LT-03225 Vilnius, Lithuania E-mail: [email protected] Fax: +370-5-233-0987
Abbreviations: DLLME, dispersive liquid–liquid microextraction; DMT, dimethyltin; MMT, monomethyltin; PVC, polyvinyl chloride; TMT, trimethyltin C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
fects on living organisms [4]. Thus, the development of accurate and sensitive analytical methods for determination of methyltin species is of special importance. For methyltin speciation analysis, GC with different detectors such as flame photometric detector [7–10], atomic emission spectrometric detector [2], MS [7, 11], inductively coupled plasma mass spectrometry detector [12] is the preferred technique. As MMT, DMT and TMT compounds present in the environment are in the ionized form, they need to be derivatized before GC analysis to obtain their volatile and thermostable forms. In the literature, several derivatization strategies for organotin compounds are described. The most commonly used derivatization reactions are the formation of hydrides by sodium borohydride and alkylation by alkylborates or by Grignard reagents [13–18]. Since trace levels of methyltins are present in environmental samples, prior to GC determination a preconcentration step is required. In recent years, preconcentration using miniaturized versions of extraction has gained a growing interest. The miniaturized version of SPE, SPME, has been applied for methyltin compounds quite extensively [7–11,19]. As a rule, derivatized methyltins are extracted from headspace. Recently methyltin compounds were extracted using another miniaturized SPE technique, headspace sorptive extraction, which in fact is a headspace mode of stir bar sorptive extraction [20]. Liquid-phase microextraction techniques have been developed as a miniaturized version of LLE. Only few articles deal with liquid-phase microextraction of organotins: single drop microextraction of butyltins and phenyltins followed by GC analysis [21, 22], organotins have been extracted into a single drop of an ionic liquid and analyzed by HPLC [23, 24]. www.jss-journal.com
1990
J. Sep. Sci. 2014, 37, 1989–1995
V. Smitiene et al.
Recently introduced dispersive liquid–liquid microextraction (DLLME) is based on a ternary solvent system [25]. A mixture of water-immiscible extraction solvent, which is dissolved in a water-miscible disperser solvent, is injected rapidly into the aqueous phase. The cloudy solution formed consists of fine droplets of extraction solvent that are dispersed into aqueous phase. Due to the considerably large surface area of the finely dispersed extraction solvent, the extraction of the analytes is achieved rapidly. The extraction solvent containing the analytes is separated by centrifugation and analyzed by an appropriate method. There are few studies on the application of DLLME for organotins extraction from water samples [26, 27]. However, only butyltin and phenyltin compounds were considered using DLLME but, as far as we know, for methyltin compounds, no application of this technique has been reported. This paper reports the results of the optimization of DLLME–GC–MS determination for the speciation analysis of methyltin compounds in aqueous solutions.
2 Materials and methods 2.1 Reagents and standards Monomethyltin (MMT) trichloride (95%), dimethyltin (DMT) dichloride (96%), trimethyltin (TMT) chloride (96%), sodium tetraethylborate (NaBEt4 ) (97%), acetone (99.9%), n-pentane (99%), methanol (99.95%), acetonitrile (99.9%), tetrachloromethane (99.5%), chlorobenzene (99%), 1,2dichlorobenzene (99%), methyl benzoate (99%), potassium dihydrogen citrate (KH7 C6 O7 ) (99%), sodium chloride (99.5%) were purchased from Sigma–Aldrich (Germany). Ethanol (96%) was purchased from Merck (Germany). Individual standard stock solutions each containing 10 mg/mL of monomethyltin trichloride, DMT dichloride, and TMT chloride were prepared in methanol. A combined standard solution of all the three methyltins was prepared in methanol from individual standard stock solutions. The solutions were stored at +4⬚C in the dark. Working standard solutions were prepared daily by diluting the combined standard solution with distilled water. The solution of internal standard chlorobenzene (1 g/mL) was prepared in ethanol. The buffer solution was prepared by dissolving the necessary amount of potassium dihydrogen citrate in distilled water to get 0.5 M concentration and then adding 0.1 M hydrochloric acid to adjust the pH to 3.
2.2 Instrumentation The chromatographic analysis was performed on a PerkinElmer Clarus 580 series gas chromatograph equipped with a programable temperature vaporizer injector and coupled to a PerkinElmer Clarus 560 S mass spectrometer (PerkinElmer, Shelton, USA). The GC system was equipped with Elite-5MS capillary column (30 m × 0.25 mm id, 0.25 m film thickness) coated with methylpolysiloxane (5% phenyl). C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Centrifugation was carried out with a Boeco S-8 centrifuge (Germany). 2.3 GC–MS conditions Helium was employed as a carrier gas with a constant flow of 1 mL/min. Injection was performed in the split mode (49:1). GC conditions were as follows: The oven temperature was programed: 50⬚C for 1 min, from 50 to 90⬚C at 15⬚C/min, from 90 to 250⬚C at 40⬚C/min and held at 250⬚C for 3 min. For methyltin solutions in n-pentane the injection port temperature was 250⬚C. For methyltin solutions obtained after DLLME the injection port temperature program was held at 160⬚C for 1 min and raised to 250⬚C at the heating rate of 200⬚C/min. The capillary column was connected to the ion source of the mass spectrometer by means of the transfer line maintained at 280⬚C. The electron ionization ion source conditions were as follows: electron energy 70 eV and temperature 180⬚C. GC–MS in full scan mode was used for the optimization of the DLLME method. The analyses were carried out with a filament multiplier delay of 2 min and the acquisition was performed in the range of m/z 45–400. In order to improve sensitivity and reduce interferences, selected ion monitoring mode was used for the quantitative analysis. The ion with the high abundance that was different from the ions of fragments of column bleed was chosen. The quantification ions (m/z values) were as follows: 163, 165, and 179 for TMT, 114, 151, and 179 for DMT, 121, 165, and 193 for MMT, 77 and 112 for internal standard chlorobenzene. 2.4 Derivatization and DLLME procedure Optimized derivatization and DLLME procedure was as follows: to a 10 mL centrifuge tube with a conical bottom 8 mL of methyltin compounds aqueous solution adjusted to pH 3, 20 L of 1 g/mL of chlorobenzene (internal standard) solution and 96 L of 5% of NaBEt4 (derivatization reagent) solution were added. The solution was left for 10 min for derivatization of methyltin compounds. Then 400 L of the mixture containing 380 L of ethanol (as a disperser solvent) and 20 L of 1,2-dichlorobenzene (as a extraction solvent) was rapidly injected to the solution using a 1 mL syringe. A cloudy solution was formed. The extraction solvent containing the analytes was separated by centrifugation for 3 min at 5000 rpm. The 1,2-dichlorobenzene phase with the analytes was sedimented in the bottom of the tube. One microliter of the extraction phase was injected into the GC–MS.
3 Results and discussion 3.1 Derivatization conditions Derivatization is one of the key factors in methyltin analysis. As previously described, three main derivatization strategies www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 37, 1989–1995
are applied for organotin derivatization. The literature review demonstrated that Grignard derivatization is tedious, time consuming, and requires dry conditions. Organotin hydrides obtained using sodium borohydride suffer from the lack of stability. Contrarily, alkylborates obtained by a derivatization procedure using sodium tetraalkylborate are stable in the water and the derivatization step can be accomplished in the aqueous phase [15]. Because of that, in this work a derivatization procedure using sodium tetraethylborate was preferred. The variables involved in the derivatization reaction, such as solution pH, reaction time, NaBEt4 concentration were optimized. To find out the optimum derivatization conditions, LLE was carried out prior to the GC–MS analysis: to 25 mL of 10 g/L aqueous methyltin solution, 100 L of 10% NaBEt4 solution was added (resulting in 0.04% NaBEt4 concentration in the solution of methyltins) and after 15 min the solution was vigorously extracted with 0.5 mL of n-pentane for 2 min. The extract was transferred into the sampling vial and automatically injected into the GC injection port. According to various reports, ethylation of organotin compounds by NaBEt4 is favorable at a broad pH range from 4 to 8 the commonly used pH for organotins determination being 4–6 [19, 26, 28–31]. However, most of these works deal with butyltins or phenyltins. Based on the results obtained for butyltins and phenyltins, derivatization of methyltins was also commonly performed at pH 4.8–6 [9,10,12]. On the other hand, organotins act as weak acids that favor the reaction with NaBEt4 [32], thus, pH values for derivatization should be as low as possible. In order to choose the optimum pH for methyltins, in this work derivatization efficiency was studied in the pH range from 1.5 to 5.5. The maximum efficiency was obtained at pH 3 (Fig. 1). This pH value was maintained using citrate buffer. The derivatization time was studied between 1 and 60 min. The results showed that the peak areas of the analytes
Figure 1. Effect of pH on the derivatization efficiency. Sample volume 25 mL, concentration of methyltins 10 g/L, derivatization with 0.04% NaBEt4 , derivatization time 15 min, extraction with 0.5 mL of n-pentane for 2 min.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1991
Figure 2. Effect of the derivatization time on the derivatization efficiency. Sample volume 25 mL, concentration of methyltins 10 g/L, derivatization with 0.04% NaBEt4 , extraction with 0.5 mL of n-pentane for 2 min.
increased up to 5–10 min (Fig. 2). Thus, 10 min derivatization time was chosen for further work.
3.2 Optimization of GC injection temperature The derivatization conditions described above were optimized using methyltin extracts in n-pentane. However, n-pentane was not suitable for DLLME. The preliminary investigations showed that the best extraction solvent for DLLME was 1,2-dichlorobenzene. However, using 1,2dichlorobenzene as an extraction solvent and applying sample injection into the hot injection port (250⬚C), the peaks of the analytes were asymmetric and broad (Fig. 3A). It seems that in the case of hot injection, 1,2-dichlorobenzene did not condense immediately and thus occupied a broad zone at the beginning of the column. Trapping of the analytes in this broad zone resulted in chromatographic peak broadening. In order to improve peak shapes, programed temperature sample introduction was applied. At the initial injection port temperature below the boiling point of the solvent 1,2-dichlorobenzene (180⬚C), the peaks were sharp and symmetric (Fig. 3B). In order to determine the optimal injection temperature, the efficiencies (expressed by the number of theoretical plates, N = 16 (tR /w)2 ) were calculated at different injection temperatures. The results presented in Fig. 4 demonstrate that 160–170⬚C is the optimal injection temperature. Thus, the following injection port temperature program was applied: the temperature was held at 160⬚C for the first 1 min and raised to 250⬚C at the heating rate of 200⬚C/min.
3.3 DLLME conditions Selection of an appropriate extraction solvent plays a main role for DLLME efficiency. An extraction solvent for www.jss-journal.com
1992
V. Smitiene et al.
J. Sep. Sci. 2014, 37, 1989–1995
Figure 5. Effect of the disperser solvent on the DLLME efficiency. Sample volume 8 mL, concentration of methyltins 5 g/L, derivatization with 0.04% NaBEt4 for 10 min, extraction solvent 1,2dichlorobenzene (20 L), centrifugation for 3 min at 5000 rpm.
Figure 3. Chromatograms of ethylated standard mixture of MMT, DMT, and TMT (5 g/L) and internal standard chlorobenzene after DLLME in 1,2-dichlorobenzene at (A) injection port temperature 250⬚C, (B) programed injection port temperature (held at 160⬚C for 1 min and raised to 250⬚C at the heating rate of 200⬚ C/min). The oven temperature was programed as follows: 50⬚C for 1 min, to 90⬚C at 15⬚C/min, to 250⬚C at 40⬚C/min and held for 3 min.
Figure 4. Effect of initial injection port temperature on the column efficiency.
traditional DLLME should have a higher density than water, should demonstrate a good extraction capability of the compounds of interest and its solubility in water should be low. In the case of subsequent GC analysis, the peak of an extraction solvent should be separated from analyte peaks. Tetrachloromethane, methyl benzoate, chlorobenzene, and 1,2-dichlorobenzene were tested as extraction solvents for derivatized methyltins. To investigate the effect of the ex C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
traction solvent, a mixture containing 500 L of acetone and 50 L of the extraction solvent was rapidly injected to 8 mL of the aqueous solution of derivatized methyltins. A cloudy solution formed was centrifuged at 5000 rpm and 1 L of the organic phase was manually injected into the GC injection port. Unfortunately, the peak of CCl4 overlapped with the peak of TMT and the peak of chlorobenzene overlapped with the peak of MMT. The peaks of the other two solvents were well separated from the peaks of the analytes with the retention times bigger than those of the analytes. Methyl benzoate showed 1.1–1.2 times higher extraction efficiency in comparison with 1,2-dichlorobenzene. However, the emulsion formed by methyl benzoate was very stable and 10 min of centrifugation was needed to separate the organic and water phases. Probably it could be explained by the fact that the density of methyl benzoate (1.08 g/mL) is quite close to that of water. In the case of 1,2-dichlorobenzene (density 1.30 g/mL) 3 min of centrifugation was sufficient for the separation of the phases. Moreover, because of the relatively high water solubility of methyl benzoate (2100 mg/L) the volume of the organic phase settled in the bottom of the centrifuge tube was 17–19 L, whereas in the case of 1,2-dichlorobenzene (water solubility 160 mg/L) it was 41–43 L. It means that the volume of 1,2-dichlorobenzene can be decreased twice or even more resulting in the increase of the analytes concentration in the extraction phase. Based on these considerations, 1,2-dichlorobenzene was chosen as an extraction solvent. The main requirement for disperser solvent is its miscibility with both the extraction solvent and the aqueous phase. Only a few solvents, namely, acetone, acetonitrile, methanol, and ethanol, fulfill this requirement and were studied. The mixture, containing 500 L of the disperser solvent and 50 L of 1,2-dichlorobenzene was used for DLLME. As demonstrated in Fig. 5, the highest extraction efficiency was achieved using ethanol, thus ethanol was selected as a disperser solvent. To investigate the effect of the extraction solvent volume, a solution containing 500 L of ethanol and 15–50 L of www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 37, 1989–1995
1993
Figure 6. Effect of the extraction solvent (1,2-dichlorobenzene) volume (A) and disperser solvent (ethanol) volume (B) on the DLLME efficiency. Sample volume 8 mL, concentration of methyltins 5 g/L, derivatization with 0.04% NaBEt4 for 10 min, centrifugation for 3 min at 5000 rpm.
1,2-dichlorobenzene was used. With the increase in extraction solvent volume, peak areas initially increased and reached the maximum at 20 L (Fig. 6A). Probably, because of a partial sedimentation of 1,2-dichlorobenzene on the centrifuge tube walls, in the case of 15 L of 1,2-dichlorobenzene, its volume in the bottom of the centrifuge tube was too small and some water phase instead of extraction phase was withdrawn into a microsyringe. On the other hand, when the extraction solvent volume exceeded 20 L, because of the bigger dilution of the analytes, peak areas of the analytes decreased. Thus, for the further work 20 L of 1,2-dichlorobenzene was used. To investigate the effect of the disperser solvent volume, different ethanol volumes (0.1–0.9 mL) and 20 L of extracting solvent were used. At low ethanol volume the cloudy state was not stable and probably this caused lower extraction efficiency. When the ethanol volume exceeded 0.3 mL, the changes in extraction efficiency were insignificant (Fig. 6B). However, with the increase of ethanol volume increased the stability of the emulsion. Thus, using more than 0.5 mL of ethanol, 3 min centrifugation time was insufficient. Because of that 0.3–0.5 mL ethanol volume was considered as the optimum. For the further work, in order to have a convenient ethanol/1,2-dichlorobenzene mixture volume for the injection (0.4 mL) and considering that the optimum 1,2dichlorobenzene volume is 20 L, 0.38 mL of ethanol volume was selected. As mentioned above, for preliminary studies, concentration of the derivatization reagent NaBEt4 in the solution of methyltins was 0.04%. At selected DLLME conditions, a concentration of the derivatization reagent was additionally assayed in the range 0.0025–0.1%. For all the methyltins, peak areas increased with the increase of NaBEt4 concentration up to 0.05–0.06% (Fig. 7). Based on the results, 0.06% concentration of NaBEt4 was selected. Addition of salt to an aqueous sample solution generally causes a decrease in the solubility of organic compounds in water, and this feature has been widely used to enhance the extraction of the analytes. In our case the aqueous solution contained salts used for the buffer preparation and for derivatization. Further increase of the salt concentration was accomplished by addition of NaCl which is commonly used for this purpose. The addition of up to 0.005 g/mL of NaCl C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 7. Effect of NaBEt4 quantity on the derivatization efficiency. Sample volume 8 mL, concentration of methyltins 5 g/L, extraction solvent 1,2-dichlorobenzene (20 L), disperser solvent ethanol (380 L), centrifugation for 3 min at 5000 rpm.
Table 1. Detection limits of the proposed method compared to former reports
Method
SPME-GC-FPD SPME-GC-PFPD SPME-GC-FPD SPME-GC-PFPD HSSE-TD-GC-MS DLLME-GC-MS
LOD, ng (Sn)/L
References
MMT
DMT
TMT
12 0.05 8.1 0.02 7.0 0.06
10 0.01 2.5 0.02 6.3 0.05
11 0.08 5.5 0.02 −a) 0.13
[33] [10] [9] [31] [20] This work
a) not determined.
slightly promoted the transport of the analytes to the extracting drop. However, with the further increase of NaCl, the density of the organic phase resulted lower than that of the aqueous phase. Because of that the organic phase formed the upper phase in two-phase system and did not sediment any more. In order to avoid this, in further experiments NaCl was not added to the samples. www.jss-journal.com
1994
J. Sep. Sci. 2014, 37, 1989–1995
V. Smitiene et al.
Table 2. Determination of methyltin compounds in spiked river water samples (concentrations in ng (Sn)/L, n = 5)
Spikes, ng (Sn)/L
Distilled water: Found values Nemunas: Found values Relative recovery, % Skirvyte: Found values Relative recovery, % Sventoji: Found values Relative recovery, % Akmena: Found values Relative recovery, %
MMT
DMT
TMT
10
20
10
20
10
20
9.8 ± 0.7 9.2 ± 0.9 94 10.4 ± 1.2 106 9.3 ± 1.2 95 9.6 ± 1.1 98
20.3 ± 1.3 19.6 ± 1.7 97 20.8 ± 1.9 102 18.9 ± 2.0 93 21.3 ± 2.1 105
10.1 ± 0.8 9.5 ± 1.0 94 10.8 ± 1.3 107 9.0 ± 1.1 89 10.6 ± 1.3 105
19.7 ± 1.6 20.8 ± 1.9 106 20.2 ± 1.5 103 20.4 ± 1.9 104 18.9 ± 2.1 96
9.7 ± 1.0 10.5 ± 1.1 108 9.3 ± 1.3 96 9.5 ± 1.0 98 9.0 ± 1.2 93
20.6 ± 1.7 20.3 ± 1.9 99 19.1 ± 1.3 93 19.9 ± 1.7 97 18.7 ± 1.8 91
3.4 Validation of the method The quality parameters of the suggested method such as linearity, LODs, and repeatabilities were determined under the optimized extraction conditions. Chlorobenzene was applied as an internal standard. For the determination of quality parameters GC–MS in SIM mode was used. The calibration curves were drawn with eight calibration points with three replicate injections of the extracts obtained after applying DLLME procedure. The linear ranges were from 0.43, 0.17, and 0.20 ng (Sn) L−1 up to 2 g (Sn) L−1 for TMT, DMT, and MMT, respectively. Correlation coefficients were 0.998– 0.999. The repeatabilities were determined by five repetitions analysis for 20 ng (Sn)/L of methyltin compounds. RSDs were 6.9–12.1%. LODs were defined as three times of baseline noise and were compared with those obtained by other methods on methyltin microextraction with GC determination (Table 1). The LODs of the proposed method are comparable with those presented in Refs. [10, 31] and significantly lower than those in Refs. [9, 33]. Additionally, DLLME has an advantage of the shortest extraction time and there is no need to use expensive SPME fibers.
3.5 Application The proposed method was applied for the determination of methyltins in river water samples. Samples from four rivers in Lithuania, namely, Nemunas near Rusne, Skirvyte near Rusne, Sventoji in the estuary, and Akmena in the estuary, were taken for the analysis. The derivatization, extraction and GC–MS analysis procedures were as described above. In all the four samples the studied methyltin compounds were not detected. In order to assess the matrix effect, the standard addition method was applied for the determination of methyltins. The water samples were spiked with 10 and 20 ng (Sn)/L of the studied methyltin compounds. The obtained results were compared with those obtained from spiked distilled water samples. Relative recoveries were determined as the ratio of the concentrations found in real and distilled water samples spiked at the same analyte concentrations. The C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
data (Table 2) indicate that the river water matrix has little effect on the extraction efficiency.
4 Concluding remarks This work demonstrated that dispersive liquid–liquid microextraction coupled to GC with mass spectrometric determination can be an adequate method for the speciation analysis of methyltin compounds in aqueous solutions. The proposed technique is fast, reliable, and environmentally friendly as it consumes only 20 L of the extraction solvent. Ethylation using sodium tetraethylborate allows the derivatization of methyltins directly in the aqueous phase. The real sample investigations demonstrated that the proposed method can be applied for river water analysis. The study was funded from the European Community’s social ˇ foundation under Grant Agreement No. VP1-3.1-SMM-08-K-01004/KS-120000-1756. The authors have declared no conflict of interest.
5 References [1] Tang, X. J., Wu, X., Dubois, A. M., Sui, G., Wu, B. H., Lai, G. C., Gong, Z. H., Gao, H. B., Liu, S. L., Zhong, Z. Y., Lin, Z. N., Olson, J., Ren, X. F., Bull. Environ. Contam. Toxicol. 2013, 90, 626–633. [2] Bjorn, A., Horsing, M., Ejlertsson, J., Svensson B. H., Waste Manag. Res. 2011, 29, 1327–1336. [3] Suzuki, Y., Endo, Y., Ogawa, M., Kim, Y., Onda, N., Yamanaka K., J. Chromatogr. B 2008, 868, 116–119. ¨ ¨ [4] Kruger, K., Hoing, T., Bensch, W., Diepgrond, V., Ahnefeld, M., Madeja, M., Binding, N., Musshoff, U., Neurotoxicology 2007, 28, 114–125. [5] Byrd, J. T., Andreae, M. O., Science 1982, 218, 565–569. [6] Concise International Chemical Assessment Document 73. Mono- and disubstituted methyltin, butyltin, and octyltin compounds, World Health Organization 2006.
www.jss-journal.com
J. Sep. Sci. 2014, 37, 1989–1995
Sample Preparation
1995
[7] Chen, B., Wang, T., Yin, Y., He, B., Jiang, G., Anal. Methods 2012, 4, 2109–2114.
[21] Colombini, V., Bancon-Montigny, C., Yang, L., Maxwell, P., Sturgeon, R. E., Mester, Z., Talanta 2004, 63, 555–560.
[8] El Ati-Hellal, M., Lespes, G., Dachraoui, M., Int. J. Environ. Anal. Chem. 2006, 86, 733–742.
[22] Shiojia, H., Tsunoia, S., Harinob, H., Tanakaa, M., J. Chromatogr. A 2004, 1048, 81–88.
[9] Cui, Z., Zhang, K., Zhou, Q., Liu, J., Jiang, G., Talanta 2011, 85, 1028–1033.
[23] Liu, J., Chi, Y., Jiang, G., J. Sep. Sci. 2005, 28, 87–91.
[10] Le Gac, M., Lespes, G., Potin-Gautier, M., J. Chromatogr. A 2003, 999, 123–134. [11] Tzillas, N. M., Zachariadis G. A., J. Sep. Sci. 2010, 33, 1610–1616. ˇ c, ˇ M., Milaci ˇ c, ˇ R., Sˇ can ˇ car, ˇ [12] Vahci J., Anal. Chim. Acta 2011, 694, 21–30.
[24] Sheikhloie, H., Saber-Trhrani, M., Abrumand-Azar, P., Waqif-Husein, S., Acta Chromatographica 2009, 21, 577– 589. [25] Rezaee, M., Assadi, Y., Milani Hosseini, M. R., Aghaee, E., Ahmadi, F., Berijani, S., J. Chromatogr. A 2006, 1116, 1–9.
[13] Morabito, R., Massanisso, P., Quevauviller, P., TRAC 2000, 19, 113–119.
[26] Pajand Birjandi, A., Bidari, A., Rezaei, F., Reza Milani Hosseini, M., Assadi, Y., J. Chromatogr. A 2008, 1193, 19–25.
[14] Magi, E., Liscio, C., Di Carro, M., J. Chromatogr. A 2008, 1210, 99–107.
[27] Smitiene, V., Baskirova, I., Vickackaite, V., Chemija 2013, 24, 210–216.
[15] Nsengimana, H., Cukrowska, E. M., Dinsmore, A., Tessier, E., Amouroux, D., J. Sep. Sci. 2009, 32, 2426– 2433.
[28] Vercauteren, J., Peres C., Devos, C., Sanda, P., Vanhaecke, F., Moens, L., Anal. Chem. 2001, 73, 1509–1514.
[16] Zachariadis, G.A., Rosenberg, E., J. Sep. Sci. 2012, 35, 1132–1137. [17] Delgado, A., Usobiaga, A., Prieto, A., Zuloaga, O., De Diego, A., Madariaga, J. M., J. Sep. Sci. 2008, 31, 768– 774. [18] Bergmann, K., Neidhart, B., J. Sep. Sci. 2001, 24, 221– 225.
[29] Zachariadis, G.A., Rosenberg, E., J. Chromatogr. B 2009, 877, 1140–1144. [30] Devos, C., Vliegen, M., Willaert, B., David, F., Moens, L., Sandra, P., J. Chromatogr. A 2005, 1079, 408–414. [31] Bravo, M. M., Valenzuela, A. S., Fuentes, E. P., Quiroz, W. V., J. Chromatogr. A 2012, 1223, 9–14.
[19] Kaur, V., Malik, A. K., Verma, N., J. Sep. Sci. 2006, 29, 333–345.
[32] Arnold, C. G., Weidenhaupt, A., Muller, S. R., Haderlein, S. B., Schwarzenbach R. P., Environ. Sci. Technol. 1997, 31, 2596–2602.
˜ ´ ´ [20] Cacho, J.I., Campillo, N., Vinas, P., Hernandez-C ordoba M., J. Chromatogr. A 2013, 1279, 1–6.
[33] Morcillo, Y., Cai, Y., Bayona, J.M., J. High Resolut. Chromatogr. 1995, 18, 767–770.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
J. Sep. Sci. 2014, 37, 1989–1995
Vaida Smitiene Ivona Semasko Vida Vickackaite Department of Analytical and Environmental Chemistry, Vilnius University, Vilnius, Lithuania Received January 27, 2014 Revised April 8, 2014 Accepted April 25, 2014
Research Article
Speciation of methyltins by dispersive liquid–liquid microextraction and gas chromatography with mass spectrometry Dispersive liquid–liquid microextraction in combination with an in situ derivatization is suggested for methyltin compound sampling and preconcentration from water solutions. The derivatization was carried out with sodium tetraethylborate at pH 3. The effects of extraction and disperser solvents type, volume, and extraction time on the extraction efficiency were investigated. 1,2-Dichlorobenzene was used as an extraction solvent and ethanol was used as a disperser solvent. The calibration graphs for all the analytes were linear up to 2 g (Sn) L−1 , correlation coefficients were 0.998–0.999, LODs were 0.13, 0.05, and 0.06 ng (Sn) L−1 for trimethyltin, DMT, and monomethyltin, respectively. Repeatabilities of the results were acceptable with RSDs up to 12.1%. A possibility to apply the proposed method for methyltin compound determination in water samples was demonstrated. Keywords: Gas chromatography / Methyltin compounds / Microextraction / Water samples DOI 10.1002/jssc.201400074
1 Introduction Organometallic tin compounds have a wide industrial application. Due to their stability and transparency, organotin compounds are used as heat and light stabilizers in polyvinyl chloride (PVC) products. There are three major types of tin stabilizers—octyltin, butyltin, and methyltin. Among them, methyltin has gradually become the dominant type [1]. Monoand disubstituted methyltin compounds are used as stabilizing additives in PVC plastics [2]. Dimethyltin (DMT) chloride is also used to produce transparent conductive films for liquid crystal panels [3]. Trimethyltin (TMT) is a low-level by-product of DMT production [1]. In the environment methyltin compounds can also originate from the biomethylation of inorganic tin [2, 4, 5]. Methyltin chlorides are released from PVC products in landfills [2], they can leach from PVC water distribution lines and thus contaminate water, food, and various ecosystems [4]. Methylated tin is much more toxic than inorganic tin. Moreover, the toxicity of methyltin compounds is strongly dependent on the species. TMT is the most toxic, it is a well-known potent neurotoxin and causes neuropathalogy in rodents and humans [6]. The neurotoxic potential of monomethyltin (MMT) and DMT is lower, however, these compounds have also been found to exert neurotoxic efCorrespondence: Dr. Vida Vickackaite, Department of Analytical and Environmental Chemistry, Vilnius University, Naugarduko 24, LT-03225 Vilnius, Lithuania E-mail: [email protected] Fax: +370-5-233-0987
Abbreviations: DLLME, dispersive liquid–liquid microextraction; DMT, dimethyltin; MMT, monomethyltin; PVC, polyvinyl chloride; TMT, trimethyltin C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
fects on living organisms [4]. Thus, the development of accurate and sensitive analytical methods for determination of methyltin species is of special importance. For methyltin speciation analysis, GC with different detectors such as flame photometric detector [7–10], atomic emission spectrometric detector [2], MS [7, 11], inductively coupled plasma mass spectrometry detector [12] is the preferred technique. As MMT, DMT and TMT compounds present in the environment are in the ionized form, they need to be derivatized before GC analysis to obtain their volatile and thermostable forms. In the literature, several derivatization strategies for organotin compounds are described. The most commonly used derivatization reactions are the formation of hydrides by sodium borohydride and alkylation by alkylborates or by Grignard reagents [13–18]. Since trace levels of methyltins are present in environmental samples, prior to GC determination a preconcentration step is required. In recent years, preconcentration using miniaturized versions of extraction has gained a growing interest. The miniaturized version of SPE, SPME, has been applied for methyltin compounds quite extensively [7–11,19]. As a rule, derivatized methyltins are extracted from headspace. Recently methyltin compounds were extracted using another miniaturized SPE technique, headspace sorptive extraction, which in fact is a headspace mode of stir bar sorptive extraction [20]. Liquid-phase microextraction techniques have been developed as a miniaturized version of LLE. Only few articles deal with liquid-phase microextraction of organotins: single drop microextraction of butyltins and phenyltins followed by GC analysis [21, 22], organotins have been extracted into a single drop of an ionic liquid and analyzed by HPLC [23, 24]. www.jss-journal.com
1990
J. Sep. Sci. 2014, 37, 1989–1995
V. Smitiene et al.
Recently introduced dispersive liquid–liquid microextraction (DLLME) is based on a ternary solvent system [25]. A mixture of water-immiscible extraction solvent, which is dissolved in a water-miscible disperser solvent, is injected rapidly into the aqueous phase. The cloudy solution formed consists of fine droplets of extraction solvent that are dispersed into aqueous phase. Due to the considerably large surface area of the finely dispersed extraction solvent, the extraction of the analytes is achieved rapidly. The extraction solvent containing the analytes is separated by centrifugation and analyzed by an appropriate method. There are few studies on the application of DLLME for organotins extraction from water samples [26, 27]. However, only butyltin and phenyltin compounds were considered using DLLME but, as far as we know, for methyltin compounds, no application of this technique has been reported. This paper reports the results of the optimization of DLLME–GC–MS determination for the speciation analysis of methyltin compounds in aqueous solutions.
2 Materials and methods 2.1 Reagents and standards Monomethyltin (MMT) trichloride (95%), dimethyltin (DMT) dichloride (96%), trimethyltin (TMT) chloride (96%), sodium tetraethylborate (NaBEt4 ) (97%), acetone (99.9%), n-pentane (99%), methanol (99.95%), acetonitrile (99.9%), tetrachloromethane (99.5%), chlorobenzene (99%), 1,2dichlorobenzene (99%), methyl benzoate (99%), potassium dihydrogen citrate (KH7 C6 O7 ) (99%), sodium chloride (99.5%) were purchased from Sigma–Aldrich (Germany). Ethanol (96%) was purchased from Merck (Germany). Individual standard stock solutions each containing 10 mg/mL of monomethyltin trichloride, DMT dichloride, and TMT chloride were prepared in methanol. A combined standard solution of all the three methyltins was prepared in methanol from individual standard stock solutions. The solutions were stored at +4⬚C in the dark. Working standard solutions were prepared daily by diluting the combined standard solution with distilled water. The solution of internal standard chlorobenzene (1 g/mL) was prepared in ethanol. The buffer solution was prepared by dissolving the necessary amount of potassium dihydrogen citrate in distilled water to get 0.5 M concentration and then adding 0.1 M hydrochloric acid to adjust the pH to 3.
2.2 Instrumentation The chromatographic analysis was performed on a PerkinElmer Clarus 580 series gas chromatograph equipped with a programable temperature vaporizer injector and coupled to a PerkinElmer Clarus 560 S mass spectrometer (PerkinElmer, Shelton, USA). The GC system was equipped with Elite-5MS capillary column (30 m × 0.25 mm id, 0.25 m film thickness) coated with methylpolysiloxane (5% phenyl). C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Centrifugation was carried out with a Boeco S-8 centrifuge (Germany). 2.3 GC–MS conditions Helium was employed as a carrier gas with a constant flow of 1 mL/min. Injection was performed in the split mode (49:1). GC conditions were as follows: The oven temperature was programed: 50⬚C for 1 min, from 50 to 90⬚C at 15⬚C/min, from 90 to 250⬚C at 40⬚C/min and held at 250⬚C for 3 min. For methyltin solutions in n-pentane the injection port temperature was 250⬚C. For methyltin solutions obtained after DLLME the injection port temperature program was held at 160⬚C for 1 min and raised to 250⬚C at the heating rate of 200⬚C/min. The capillary column was connected to the ion source of the mass spectrometer by means of the transfer line maintained at 280⬚C. The electron ionization ion source conditions were as follows: electron energy 70 eV and temperature 180⬚C. GC–MS in full scan mode was used for the optimization of the DLLME method. The analyses were carried out with a filament multiplier delay of 2 min and the acquisition was performed in the range of m/z 45–400. In order to improve sensitivity and reduce interferences, selected ion monitoring mode was used for the quantitative analysis. The ion with the high abundance that was different from the ions of fragments of column bleed was chosen. The quantification ions (m/z values) were as follows: 163, 165, and 179 for TMT, 114, 151, and 179 for DMT, 121, 165, and 193 for MMT, 77 and 112 for internal standard chlorobenzene. 2.4 Derivatization and DLLME procedure Optimized derivatization and DLLME procedure was as follows: to a 10 mL centrifuge tube with a conical bottom 8 mL of methyltin compounds aqueous solution adjusted to pH 3, 20 L of 1 g/mL of chlorobenzene (internal standard) solution and 96 L of 5% of NaBEt4 (derivatization reagent) solution were added. The solution was left for 10 min for derivatization of methyltin compounds. Then 400 L of the mixture containing 380 L of ethanol (as a disperser solvent) and 20 L of 1,2-dichlorobenzene (as a extraction solvent) was rapidly injected to the solution using a 1 mL syringe. A cloudy solution was formed. The extraction solvent containing the analytes was separated by centrifugation for 3 min at 5000 rpm. The 1,2-dichlorobenzene phase with the analytes was sedimented in the bottom of the tube. One microliter of the extraction phase was injected into the GC–MS.
3 Results and discussion 3.1 Derivatization conditions Derivatization is one of the key factors in methyltin analysis. As previously described, three main derivatization strategies www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 37, 1989–1995
are applied for organotin derivatization. The literature review demonstrated that Grignard derivatization is tedious, time consuming, and requires dry conditions. Organotin hydrides obtained using sodium borohydride suffer from the lack of stability. Contrarily, alkylborates obtained by a derivatization procedure using sodium tetraalkylborate are stable in the water and the derivatization step can be accomplished in the aqueous phase [15]. Because of that, in this work a derivatization procedure using sodium tetraethylborate was preferred. The variables involved in the derivatization reaction, such as solution pH, reaction time, NaBEt4 concentration were optimized. To find out the optimum derivatization conditions, LLE was carried out prior to the GC–MS analysis: to 25 mL of 10 g/L aqueous methyltin solution, 100 L of 10% NaBEt4 solution was added (resulting in 0.04% NaBEt4 concentration in the solution of methyltins) and after 15 min the solution was vigorously extracted with 0.5 mL of n-pentane for 2 min. The extract was transferred into the sampling vial and automatically injected into the GC injection port. According to various reports, ethylation of organotin compounds by NaBEt4 is favorable at a broad pH range from 4 to 8 the commonly used pH for organotins determination being 4–6 [19, 26, 28–31]. However, most of these works deal with butyltins or phenyltins. Based on the results obtained for butyltins and phenyltins, derivatization of methyltins was also commonly performed at pH 4.8–6 [9,10,12]. On the other hand, organotins act as weak acids that favor the reaction with NaBEt4 [32], thus, pH values for derivatization should be as low as possible. In order to choose the optimum pH for methyltins, in this work derivatization efficiency was studied in the pH range from 1.5 to 5.5. The maximum efficiency was obtained at pH 3 (Fig. 1). This pH value was maintained using citrate buffer. The derivatization time was studied between 1 and 60 min. The results showed that the peak areas of the analytes
Figure 1. Effect of pH on the derivatization efficiency. Sample volume 25 mL, concentration of methyltins 10 g/L, derivatization with 0.04% NaBEt4 , derivatization time 15 min, extraction with 0.5 mL of n-pentane for 2 min.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1991
Figure 2. Effect of the derivatization time on the derivatization efficiency. Sample volume 25 mL, concentration of methyltins 10 g/L, derivatization with 0.04% NaBEt4 , extraction with 0.5 mL of n-pentane for 2 min.
increased up to 5–10 min (Fig. 2). Thus, 10 min derivatization time was chosen for further work.
3.2 Optimization of GC injection temperature The derivatization conditions described above were optimized using methyltin extracts in n-pentane. However, n-pentane was not suitable for DLLME. The preliminary investigations showed that the best extraction solvent for DLLME was 1,2-dichlorobenzene. However, using 1,2dichlorobenzene as an extraction solvent and applying sample injection into the hot injection port (250⬚C), the peaks of the analytes were asymmetric and broad (Fig. 3A). It seems that in the case of hot injection, 1,2-dichlorobenzene did not condense immediately and thus occupied a broad zone at the beginning of the column. Trapping of the analytes in this broad zone resulted in chromatographic peak broadening. In order to improve peak shapes, programed temperature sample introduction was applied. At the initial injection port temperature below the boiling point of the solvent 1,2-dichlorobenzene (180⬚C), the peaks were sharp and symmetric (Fig. 3B). In order to determine the optimal injection temperature, the efficiencies (expressed by the number of theoretical plates, N = 16 (tR /w)2 ) were calculated at different injection temperatures. The results presented in Fig. 4 demonstrate that 160–170⬚C is the optimal injection temperature. Thus, the following injection port temperature program was applied: the temperature was held at 160⬚C for the first 1 min and raised to 250⬚C at the heating rate of 200⬚C/min.
3.3 DLLME conditions Selection of an appropriate extraction solvent plays a main role for DLLME efficiency. An extraction solvent for www.jss-journal.com
1992
V. Smitiene et al.
J. Sep. Sci. 2014, 37, 1989–1995
Figure 5. Effect of the disperser solvent on the DLLME efficiency. Sample volume 8 mL, concentration of methyltins 5 g/L, derivatization with 0.04% NaBEt4 for 10 min, extraction solvent 1,2dichlorobenzene (20 L), centrifugation for 3 min at 5000 rpm.
Figure 3. Chromatograms of ethylated standard mixture of MMT, DMT, and TMT (5 g/L) and internal standard chlorobenzene after DLLME in 1,2-dichlorobenzene at (A) injection port temperature 250⬚C, (B) programed injection port temperature (held at 160⬚C for 1 min and raised to 250⬚C at the heating rate of 200⬚ C/min). The oven temperature was programed as follows: 50⬚C for 1 min, to 90⬚C at 15⬚C/min, to 250⬚C at 40⬚C/min and held for 3 min.
Figure 4. Effect of initial injection port temperature on the column efficiency.
traditional DLLME should have a higher density than water, should demonstrate a good extraction capability of the compounds of interest and its solubility in water should be low. In the case of subsequent GC analysis, the peak of an extraction solvent should be separated from analyte peaks. Tetrachloromethane, methyl benzoate, chlorobenzene, and 1,2-dichlorobenzene were tested as extraction solvents for derivatized methyltins. To investigate the effect of the ex C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
traction solvent, a mixture containing 500 L of acetone and 50 L of the extraction solvent was rapidly injected to 8 mL of the aqueous solution of derivatized methyltins. A cloudy solution formed was centrifuged at 5000 rpm and 1 L of the organic phase was manually injected into the GC injection port. Unfortunately, the peak of CCl4 overlapped with the peak of TMT and the peak of chlorobenzene overlapped with the peak of MMT. The peaks of the other two solvents were well separated from the peaks of the analytes with the retention times bigger than those of the analytes. Methyl benzoate showed 1.1–1.2 times higher extraction efficiency in comparison with 1,2-dichlorobenzene. However, the emulsion formed by methyl benzoate was very stable and 10 min of centrifugation was needed to separate the organic and water phases. Probably it could be explained by the fact that the density of methyl benzoate (1.08 g/mL) is quite close to that of water. In the case of 1,2-dichlorobenzene (density 1.30 g/mL) 3 min of centrifugation was sufficient for the separation of the phases. Moreover, because of the relatively high water solubility of methyl benzoate (2100 mg/L) the volume of the organic phase settled in the bottom of the centrifuge tube was 17–19 L, whereas in the case of 1,2-dichlorobenzene (water solubility 160 mg/L) it was 41–43 L. It means that the volume of 1,2-dichlorobenzene can be decreased twice or even more resulting in the increase of the analytes concentration in the extraction phase. Based on these considerations, 1,2-dichlorobenzene was chosen as an extraction solvent. The main requirement for disperser solvent is its miscibility with both the extraction solvent and the aqueous phase. Only a few solvents, namely, acetone, acetonitrile, methanol, and ethanol, fulfill this requirement and were studied. The mixture, containing 500 L of the disperser solvent and 50 L of 1,2-dichlorobenzene was used for DLLME. As demonstrated in Fig. 5, the highest extraction efficiency was achieved using ethanol, thus ethanol was selected as a disperser solvent. To investigate the effect of the extraction solvent volume, a solution containing 500 L of ethanol and 15–50 L of www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 37, 1989–1995
1993
Figure 6. Effect of the extraction solvent (1,2-dichlorobenzene) volume (A) and disperser solvent (ethanol) volume (B) on the DLLME efficiency. Sample volume 8 mL, concentration of methyltins 5 g/L, derivatization with 0.04% NaBEt4 for 10 min, centrifugation for 3 min at 5000 rpm.
1,2-dichlorobenzene was used. With the increase in extraction solvent volume, peak areas initially increased and reached the maximum at 20 L (Fig. 6A). Probably, because of a partial sedimentation of 1,2-dichlorobenzene on the centrifuge tube walls, in the case of 15 L of 1,2-dichlorobenzene, its volume in the bottom of the centrifuge tube was too small and some water phase instead of extraction phase was withdrawn into a microsyringe. On the other hand, when the extraction solvent volume exceeded 20 L, because of the bigger dilution of the analytes, peak areas of the analytes decreased. Thus, for the further work 20 L of 1,2-dichlorobenzene was used. To investigate the effect of the disperser solvent volume, different ethanol volumes (0.1–0.9 mL) and 20 L of extracting solvent were used. At low ethanol volume the cloudy state was not stable and probably this caused lower extraction efficiency. When the ethanol volume exceeded 0.3 mL, the changes in extraction efficiency were insignificant (Fig. 6B). However, with the increase of ethanol volume increased the stability of the emulsion. Thus, using more than 0.5 mL of ethanol, 3 min centrifugation time was insufficient. Because of that 0.3–0.5 mL ethanol volume was considered as the optimum. For the further work, in order to have a convenient ethanol/1,2-dichlorobenzene mixture volume for the injection (0.4 mL) and considering that the optimum 1,2dichlorobenzene volume is 20 L, 0.38 mL of ethanol volume was selected. As mentioned above, for preliminary studies, concentration of the derivatization reagent NaBEt4 in the solution of methyltins was 0.04%. At selected DLLME conditions, a concentration of the derivatization reagent was additionally assayed in the range 0.0025–0.1%. For all the methyltins, peak areas increased with the increase of NaBEt4 concentration up to 0.05–0.06% (Fig. 7). Based on the results, 0.06% concentration of NaBEt4 was selected. Addition of salt to an aqueous sample solution generally causes a decrease in the solubility of organic compounds in water, and this feature has been widely used to enhance the extraction of the analytes. In our case the aqueous solution contained salts used for the buffer preparation and for derivatization. Further increase of the salt concentration was accomplished by addition of NaCl which is commonly used for this purpose. The addition of up to 0.005 g/mL of NaCl C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 7. Effect of NaBEt4 quantity on the derivatization efficiency. Sample volume 8 mL, concentration of methyltins 5 g/L, extraction solvent 1,2-dichlorobenzene (20 L), disperser solvent ethanol (380 L), centrifugation for 3 min at 5000 rpm.
Table 1. Detection limits of the proposed method compared to former reports
Method
SPME-GC-FPD SPME-GC-PFPD SPME-GC-FPD SPME-GC-PFPD HSSE-TD-GC-MS DLLME-GC-MS
LOD, ng (Sn)/L
References
MMT
DMT
TMT
12 0.05 8.1 0.02 7.0 0.06
10 0.01 2.5 0.02 6.3 0.05
11 0.08 5.5 0.02 −a) 0.13
[33] [10] [9] [31] [20] This work
a) not determined.
slightly promoted the transport of the analytes to the extracting drop. However, with the further increase of NaCl, the density of the organic phase resulted lower than that of the aqueous phase. Because of that the organic phase formed the upper phase in two-phase system and did not sediment any more. In order to avoid this, in further experiments NaCl was not added to the samples. www.jss-journal.com
1994
J. Sep. Sci. 2014, 37, 1989–1995
V. Smitiene et al.
Table 2. Determination of methyltin compounds in spiked river water samples (concentrations in ng (Sn)/L, n = 5)
Spikes, ng (Sn)/L
Distilled water: Found values Nemunas: Found values Relative recovery, % Skirvyte: Found values Relative recovery, % Sventoji: Found values Relative recovery, % Akmena: Found values Relative recovery, %
MMT
DMT
TMT
10
20
10
20
10
20
9.8 ± 0.7 9.2 ± 0.9 94 10.4 ± 1.2 106 9.3 ± 1.2 95 9.6 ± 1.1 98
20.3 ± 1.3 19.6 ± 1.7 97 20.8 ± 1.9 102 18.9 ± 2.0 93 21.3 ± 2.1 105
10.1 ± 0.8 9.5 ± 1.0 94 10.8 ± 1.3 107 9.0 ± 1.1 89 10.6 ± 1.3 105
19.7 ± 1.6 20.8 ± 1.9 106 20.2 ± 1.5 103 20.4 ± 1.9 104 18.9 ± 2.1 96
9.7 ± 1.0 10.5 ± 1.1 108 9.3 ± 1.3 96 9.5 ± 1.0 98 9.0 ± 1.2 93
20.6 ± 1.7 20.3 ± 1.9 99 19.1 ± 1.3 93 19.9 ± 1.7 97 18.7 ± 1.8 91
3.4 Validation of the method The quality parameters of the suggested method such as linearity, LODs, and repeatabilities were determined under the optimized extraction conditions. Chlorobenzene was applied as an internal standard. For the determination of quality parameters GC–MS in SIM mode was used. The calibration curves were drawn with eight calibration points with three replicate injections of the extracts obtained after applying DLLME procedure. The linear ranges were from 0.43, 0.17, and 0.20 ng (Sn) L−1 up to 2 g (Sn) L−1 for TMT, DMT, and MMT, respectively. Correlation coefficients were 0.998– 0.999. The repeatabilities were determined by five repetitions analysis for 20 ng (Sn)/L of methyltin compounds. RSDs were 6.9–12.1%. LODs were defined as three times of baseline noise and were compared with those obtained by other methods on methyltin microextraction with GC determination (Table 1). The LODs of the proposed method are comparable with those presented in Refs. [10, 31] and significantly lower than those in Refs. [9, 33]. Additionally, DLLME has an advantage of the shortest extraction time and there is no need to use expensive SPME fibers.
3.5 Application The proposed method was applied for the determination of methyltins in river water samples. Samples from four rivers in Lithuania, namely, Nemunas near Rusne, Skirvyte near Rusne, Sventoji in the estuary, and Akmena in the estuary, were taken for the analysis. The derivatization, extraction and GC–MS analysis procedures were as described above. In all the four samples the studied methyltin compounds were not detected. In order to assess the matrix effect, the standard addition method was applied for the determination of methyltins. The water samples were spiked with 10 and 20 ng (Sn)/L of the studied methyltin compounds. The obtained results were compared with those obtained from spiked distilled water samples. Relative recoveries were determined as the ratio of the concentrations found in real and distilled water samples spiked at the same analyte concentrations. The C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
data (Table 2) indicate that the river water matrix has little effect on the extraction efficiency.
4 Concluding remarks This work demonstrated that dispersive liquid–liquid microextraction coupled to GC with mass spectrometric determination can be an adequate method for the speciation analysis of methyltin compounds in aqueous solutions. The proposed technique is fast, reliable, and environmentally friendly as it consumes only 20 L of the extraction solvent. Ethylation using sodium tetraethylborate allows the derivatization of methyltins directly in the aqueous phase. The real sample investigations demonstrated that the proposed method can be applied for river water analysis. The study was funded from the European Community’s social ˇ foundation under Grant Agreement No. VP1-3.1-SMM-08-K-01004/KS-120000-1756. The authors have declared no conflict of interest.
5 References [1] Tang, X. J., Wu, X., Dubois, A. M., Sui, G., Wu, B. H., Lai, G. C., Gong, Z. H., Gao, H. B., Liu, S. L., Zhong, Z. Y., Lin, Z. N., Olson, J., Ren, X. F., Bull. Environ. Contam. Toxicol. 2013, 90, 626–633. [2] Bjorn, A., Horsing, M., Ejlertsson, J., Svensson B. H., Waste Manag. Res. 2011, 29, 1327–1336. [3] Suzuki, Y., Endo, Y., Ogawa, M., Kim, Y., Onda, N., Yamanaka K., J. Chromatogr. B 2008, 868, 116–119. ¨ ¨ [4] Kruger, K., Hoing, T., Bensch, W., Diepgrond, V., Ahnefeld, M., Madeja, M., Binding, N., Musshoff, U., Neurotoxicology 2007, 28, 114–125. [5] Byrd, J. T., Andreae, M. O., Science 1982, 218, 565–569. [6] Concise International Chemical Assessment Document 73. Mono- and disubstituted methyltin, butyltin, and octyltin compounds, World Health Organization 2006.
www.jss-journal.com
J. Sep. Sci. 2014, 37, 1989–1995
Sample Preparation
1995
[7] Chen, B., Wang, T., Yin, Y., He, B., Jiang, G., Anal. Methods 2012, 4, 2109–2114.
[21] Colombini, V., Bancon-Montigny, C., Yang, L., Maxwell, P., Sturgeon, R. E., Mester, Z., Talanta 2004, 63, 555–560.
[8] El Ati-Hellal, M., Lespes, G., Dachraoui, M., Int. J. Environ. Anal. Chem. 2006, 86, 733–742.
[22] Shiojia, H., Tsunoia, S., Harinob, H., Tanakaa, M., J. Chromatogr. A 2004, 1048, 81–88.
[9] Cui, Z., Zhang, K., Zhou, Q., Liu, J., Jiang, G., Talanta 2011, 85, 1028–1033.
[23] Liu, J., Chi, Y., Jiang, G., J. Sep. Sci. 2005, 28, 87–91.
[10] Le Gac, M., Lespes, G., Potin-Gautier, M., J. Chromatogr. A 2003, 999, 123–134. [11] Tzillas, N. M., Zachariadis G. A., J. Sep. Sci. 2010, 33, 1610–1616. ˇ c, ˇ M., Milaci ˇ c, ˇ R., Sˇ can ˇ car, ˇ [12] Vahci J., Anal. Chim. Acta 2011, 694, 21–30.
[24] Sheikhloie, H., Saber-Trhrani, M., Abrumand-Azar, P., Waqif-Husein, S., Acta Chromatographica 2009, 21, 577– 589. [25] Rezaee, M., Assadi, Y., Milani Hosseini, M. R., Aghaee, E., Ahmadi, F., Berijani, S., J. Chromatogr. A 2006, 1116, 1–9.
[13] Morabito, R., Massanisso, P., Quevauviller, P., TRAC 2000, 19, 113–119.
[26] Pajand Birjandi, A., Bidari, A., Rezaei, F., Reza Milani Hosseini, M., Assadi, Y., J. Chromatogr. A 2008, 1193, 19–25.
[14] Magi, E., Liscio, C., Di Carro, M., J. Chromatogr. A 2008, 1210, 99–107.
[27] Smitiene, V., Baskirova, I., Vickackaite, V., Chemija 2013, 24, 210–216.
[15] Nsengimana, H., Cukrowska, E. M., Dinsmore, A., Tessier, E., Amouroux, D., J. Sep. Sci. 2009, 32, 2426– 2433.
[28] Vercauteren, J., Peres C., Devos, C., Sanda, P., Vanhaecke, F., Moens, L., Anal. Chem. 2001, 73, 1509–1514.
[16] Zachariadis, G.A., Rosenberg, E., J. Sep. Sci. 2012, 35, 1132–1137. [17] Delgado, A., Usobiaga, A., Prieto, A., Zuloaga, O., De Diego, A., Madariaga, J. M., J. Sep. Sci. 2008, 31, 768– 774. [18] Bergmann, K., Neidhart, B., J. Sep. Sci. 2001, 24, 221– 225.
[29] Zachariadis, G.A., Rosenberg, E., J. Chromatogr. B 2009, 877, 1140–1144. [30] Devos, C., Vliegen, M., Willaert, B., David, F., Moens, L., Sandra, P., J. Chromatogr. A 2005, 1079, 408–414. [31] Bravo, M. M., Valenzuela, A. S., Fuentes, E. P., Quiroz, W. V., J. Chromatogr. A 2012, 1223, 9–14.
[19] Kaur, V., Malik, A. K., Verma, N., J. Sep. Sci. 2006, 29, 333–345.
[32] Arnold, C. G., Weidenhaupt, A., Muller, S. R., Haderlein, S. B., Schwarzenbach R. P., Environ. Sci. Technol. 1997, 31, 2596–2602.
˜ ´ ´ [20] Cacho, J.I., Campillo, N., Vinas, P., Hernandez-C ordoba M., J. Chromatogr. A 2013, 1279, 1–6.
[33] Morcillo, Y., Cai, Y., Bayona, J.M., J. High Resolut. Chromatogr. 1995, 18, 767–770.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
