Journal of Chromatography B, 947–948 (2014) 17–22
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
A sensitive and fast method for trihalomethanes in urine using gas chromatography–triple quadrupole mass spectrometry Pantelis Charisiadis, Konstantinos C. Makris ∗ Cyprus International Institute for Environmental and Public Health in Association with Harvard School of Public Health, Cyprus University of Technology, Limassol, Cyprus
a r t i c l e
i n f o
Article history: Received 24 July 2013 Accepted 30 November 2013 Available online 6 December 2013 Keywords: Chlorine Disinfection by products Exposure Mass spectrometry Trihalomethanes Urine
a b s t r a c t Because of the plethora of exposure sources and routes through which humans are exposed to trihalomethanes (THM), the limitation of their short half-lives could be overcome, if a highly sensitive method was available to quantify urinary THM concentrations at sub-ppb levels. The objective of this study was to develop a fast and reliable method for the determination of the four THM analytes in human urine. A sensitive methodology was developed for THM in urine samples using gas chromatography coupled with triple quadrupole mass spectrometry (GC–QqQ-MS/MS) promoting its use in epidemiological and biomonitoring studies. The proposed methodology enjoys limits of detection similar to those reported in the literature (11–80 ng L−1 ) and the advantages of small initial urine volumes (15 mL) and fast analysis per sample (12 min) when compared with other methods. This is the first report using GC–QqQ-MS/MS for the determination of THM in urine samples. Because of its simplicity and less time-consuming nature, the proposed method could be incorporated into detailed (hundreds of participants’ urine samples) exposure assessment protocols providing valuable insight into the dose–response relationship of THM and cancer or pregnancy anomalies. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Despite the recent market introduction of disinfectants void of chlorine, hypochlorite is still widely used as a common microbial disinfectant in potable water and solid surfaces in households (floors, toilets, etc.). Inevitably, every-day household activities, such as cleaning (mopping) the house and use of chlorinated tap water for washing dishes, showering, laundry, and swimming in chlorinated pools present us with common examples of daily exposure sources to chlorine and its disinfection by-products (DBP). The formation of DBP is triggered by reactions between residual chlorine and the natural organic matter in water or in dust attached to household solid surfaces or perhaps aerosols [1]. A suite of chlorinated DBP may be formed, including the
Abbreviations: THM, trihalomethanes; TCM, trichloromethane; BDCM, bromodichloromethane; DBCM, dibromochloromethane; TBM, tribromomethane; DBP, disinfection by-products; MTBE, t-butyl methyl ether; FMU, first morning void urine; GC–QqQ-MS/MS, gas chromatography coupled with triple quadrupole mass spectrometry; MRM, multi reaction monitoring; EI, electron impact ionization; SPME, solid phase micro extraction; HS, headspace. ∗ Corresponding author at: Cyprus International Institute for Environmental and Public Health in Association with Harvard School of Public Health, Cyprus University of Technology, Irenes 95, Limassol 3041, Cyprus. Tel.: +357 25002398; fax: +357 25002676. E-mail address: [email protected] (K.C. Makris). 1570-0232/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2013.11.060
trihalomethanes-THM, being the most abundant class of DBP in chlorinated tap water and currently regulated in both sides of the Atlantic Ocean (trichloromethane, bromodichloromethane, dibromochloromethane and tribromomethane) [2]. Numerous epidemiological studies showed the increased cancer risk for populations exposed to THM in potable water [3]. Increased risk of spontaneous abortion and preterm births among women who drink larger amounts of tap water was observed [4], suggesting a possible link between DBP and adverse pregnancy outcomes. Preterm birth is the primary cause of newborn deaths and the second major cause of death after pneumonia in children under five years [5]. Human absorption of THM may occur not only through ingestion of tap water, but also via inhalation and dermal uptake during showering, bathing, and swimming, with a different pattern of metabolism for the various routes of exposure and potentially large variation between-, and within-individuals [6]. The relatively high volatility of THM, especially for chloroform makes it difficult to accurately and precisely capture historic exposures to THM; when this is coupled with the very short biological half-life of THM in the order of approximately 1 h or less [7,8], it highlights the perplexity of accurately quantifying THM human exposure variability in urine samples [7,9–12]. Blood has been also considered as a biomarker of THM exposures, but quantified levels are of similar magnitude to those in urine (sub-ppb levels) [13–17]. We believe that the enhanced sensitivity of the analytical method required to quantify THM concentrations in urine or blood has hampered progress
18
P. Charisiadis, K.C. Makris / J. Chromatogr. B 947–948 (2014) 17–22
of epidemiological studies in gaining further insight into THM dose and disease process. The problem may be exacerbated by various THM exposure sources, making it difficult to capture exposure variability with only a few biomarker sample analyses, since there is typically an elevated cost and time-consuming methodology associated with urinary measurements. Because of the plethora of exposure sources and our daily exposures to THM via the aforementioned scenaria, the limitation of their short half-lives could be overcome, if a highly sensitive method was available to quantify urinary THM concentrations at parts per trillion levels. During the past couple of years, we have been systematically studying spatio-temporal THM exposure variability in the city of Nicosia, Cyprus, where first morning void urine samples were collected from recruited volunteers (n = ∼400). This has forced us to develop a sensitive and fast method to quantify low-level THM exposures in the participants. Herein, we described a simple liquid–liquid extraction procedure, using a small amount of urine volume and minimal total sample analysis time. The use of the sensitive gas chromatography coupled with triple quadrupole mass spectrometry (GC–QqQ-MS/MS) methodology in the multiple reactions monitoring (MRM) mode enabled us to reach detection limits close to those already reported in literature, but in a simpler and faster manner. To the best of our knowledge, there is no published study in the literature reporting measurements of urinary THM concentrations using GC–QqQ-MS/MS. The objective of this study was to develop an easy, fast and reliable method for the qualitative and quantitative determination of the four THM analytes in human urine. Our ultimate goal is to facilitate the inclusion of this method (because of its simplicity and less time consuming nature) into high-throughput exposure assessment measurements that could help epidemiological studies providing further insight into the dose–response relationship of DBP and cancer or pregnancy anomalies. 2. Materials and methods 2.1. Chemicals Decafluorobiphenyl, sodium sulphate, acetone and methanol GC grade were purchased from Sigma–Aldrich. t-Butyl methyl ether GC grade was purchased from Panreac (Barcelona, Spain). Trihalomethanes mix, containing trichloromethane (TCM), bromodichloromethane (BDCM), dibromochloromethane (DBCM), and tribromomethane (TBM), was purchased from Restek (Bellefonte, USA). 4-Bromofluorobenzene was purchased from Supelco (Bellefonte, USA). 2.2. Sampling, preservation and storage Urine samples (first morning void urine, FMU) were collected from 20 volunteers residing into two specific areas of the city of Nicosia, Cyprus. Our sampling team collected FMU urine samples in 60 mL polypropylene vials with minimum headspace; samples were temporarily stored at −20 ◦ C freezer until arrival in the laboratory, where they were either immediately analyzed, or stored in the −80 ◦ C freezer, until analysis. The study protocol was approved by the National Bioethics Committee of Cyprus (EEBK/EP/2012/17). 2.3. Standard solutions Stock solution of 200 mg L−1 was prepared from the initial concentrated trihalomethane mix of 2000 mg L−1 in methanol, and further diluted to prepare calibration and additive solutions. A surrogate solution of 1000 mg L−1 was prepared by adding 50 mg of decafluorobiphenyl into 50 mL of acetone. Stock solution of 200 mg L−1 was prepared from the initial and further diluted for the
working solution. Stock solution of 200 mg L−1 was prepared from the initial concentrated internal standard 4-bromofluorobenzene of 1000 mg L−1 in acetone and further diluted for the working solution. All solutions containing THM, internal standard and surrogate were kept away from light, prepared fresh, and stored always at −20 ◦ C in glass vials. 2.4. Sample preparation Our method development was based upon the principles of the EPA Method 551.1-1 for THM in drinking-water after several modifications which are described below. In detail, a liquid–liquid extraction protocol was optimized by mixing 15 mL urine sample (spiked with surrogate solution at a final concentration of 10 g L−1 ) with 2 mL of t-butyl methyl ether, adding 6.0 g of sodium sulfate to saturate the sample and shaking gently for 5 min in a lab shaker at 100 rpm. Following, samples were centrifuged for 1 min at 500 rpm for a clear phase separation. Half of mL of the organic phase was transferred into a GC autosampler screw top glass vial with blue PTFE/butyl rubber septa (Restek, USA), containing the internal standard solution at final concentration of 200 g L−1 . 2.5. GC–QqQ-MS/MS analysis Gas chromatography was performed in an Agilent 7890A gas chromatograph coupled with an Agilent 7000B triple quadrupole GC–QqQ-MS/MS system (Agilent Technologies, Waldbronn, Germany). Separation was performed using a 30 m × 0.25 mm × 0.25 m Rxi-5 ms (5% diphenyl/95% dimethyl polysiloxane) column from Restek. An Agilent 7693A Series automatic liquid sampler was used and the injection volume was 2 L with a 10.0 L ± 1% glass syringe (Agilent, Australia). Electron impact (EI) ionization was performed at electron energy of 70 eV. MS/MS was performed in the MRM mode. The quadrupole mass detector was operated at 150 ◦ C. The ion source temperature was set to 250 ◦ C, and the transfer line was set to 250 ◦ C. The column oven temperature programs were set to 30 ◦ C for 5 min, then to 100 ◦ C for 2.4 min with 50 ◦ C min−1 , and then to 300 ◦ C for 0.7 min with 80 ◦ C min−1 . Carrier gas was helium at 1 mL min−1 . Peaks were identified on the basis of their fragmentation patterns using the NIST Mass Spectral Search Program (NIST 08 version 2.0). The system was controlled by the software Mass Hunter Workstation (Agilent, rev. B.05.00). 2.6. Optimization of the GC–MS/MS MTBE extracts of the four THM in urine sample free of THM (concentrations of 1000 g L−1 ) were tested in full scan mode for the selection of the precursor ions and to optimize the MS/MS measurement of each analyte. In brief, MS spectra were obtained and analytes were identified by comparison to NIST mass spectra library (NIST 08 version 2.0). Mass spectra of the analytes in the ratio (m/z) 40–400 were recorded. Ions of the higher mass ratios and abundance were preferably chosen as the precursor ions in a way to increase the selectivity and sensitivity of the method. The chosen precursor ions were examined to identify their product ions in the daughter ion scan mode. The parent-daughter ion pairs used for multiple reaction monitoring (MRM) were chosen based on ion abundances in different collision energies while scanning cycles were kept constant to approximately 3 cycles per second. For each compound, molecular ion MS/MS spectra in different collision energies were obtained. Collision energies (5–60 eV) for the chosen parent-daughter ion pairs of each analyte were optimized to achieve maximum signal intensity, using nitrogen gas as the collision gas (Table 1).
P. Charisiadis, K.C. Makris / J. Chromatogr. B 947–948 (2014) 17–22
19
Table 1 Names, segments, retention time, dwell time, precursors and products ions of quantitative and confirmation transition, and collision energy for the trihalomethanes analysis. THMs
TCM BDCM DBCM TBM ISTD SR
Time segment (min)
4.800 5.900 6.800 7.800 9.800
Retention time (min)
5.561 6.325 7.324 8.371 8.856 10.307
Dwell time (ms)
75 75 75 37.5 37.5 75
Selective ions (m/z) – MRM transitions
Quantitative (Q)
Collision energy (eV)
Confirmation (q1)
Collision energy (eV)
Confirmation (q2)
Collision energy (eV)
Confirmation (q3)
Collision energy (eV)
83 → 47 83 → 47 129 → 48 173 → 94 174 → 95 334 → 265
35 33 43 48 15 33
83 → 48 85 → 47 127 → 48 171 → 92 174 → 75 334 → 315
42 37 46 48 55 23
85 → 47 129 → 48 208 → 129 252 → 173 176 → 95 334 → 165
38 43 2 2 38 60
85 → 49 127 → 48 131 → 50 175 → 94 176 → 75 334 → 234
37 45 46 55 17 45
2.7. Calibration curve and linearity As a means of eliminating any matrix effects, procedural calibration standards were prepared in pooled urine sample with negligible THM measured concentrations. An aliquot of 15 mL the pooled urine sample was placed in a 20 mL screw top glass vial with screw cap with hole and PTFE/silicone septum. The surrogate standard solution and the desired concentration dilution standard (in methanol) were added. The curve was established by measuring ten samples of urinary THM at final concentration from 39 to 20,240 ng L−1 (39, 79, 158, 316, 633, 1265, 2530, 5060, 10,120, and 20,240 ng L−1 ) using GC–QqQ-MS/MS in the MRM mode. Quantitative analysis was based on peak area measurements as ratios versus peak area of internal standard. All samples were treated as following the sample preparation procedure. The regression coefficient was always 0.9998 or better (Table 2). 2.8. Method detection limit The limit of detection (LOD) of the proposed methodology was determined at the lowest concentration of each analyte. The limit of quantification (LOQ) and limit of detection (LOD) were determined based on the standard deviation of eight measurements of the spiked blank urine samples [18]. The limit of detection (LOD) was calculated by adding 3× the standard deviation of the response lowest concentration of the calibration curve. Accordingly, the limit of quantification (LOQ) was equal to 3× the LOD. 2.9. Stability of THMs in urine samples—storage effect A stability experiment was performed to evaluate the stability of THM in stored urine samples during time of sampling and storage at −80 ◦ C. In an attempt to simulate as much as possible realistic sampling conditions, a pooled urine sample (with negligible THM measured concentrations) was placed in sampling vials. Samples were spiked with a concentration of THM in a final concentration of 10 g L−1 for each analyte (four analytes). They were kept in the −80 ◦ C freezer, and at specific time intervals, vials were sacrificed and samples were analyzed for THM. 3. Results 3.1. Calibration curve and recovery Ultrapure water was always used as the background matrix, because it did not contain residual chlorine or THM, being < always LOD. Selective ion – MRM optimized transitions for THM are depicted in Table 1. Mass spectra of the four trihalomethanes contain characteristic patterns caused by chlorine and bromine isotopes. For example, the trichloromethane mass spectrum contains an ion cluster at m/z 83, 85, and 87 with known relative abundances of 35Cl and 37Cl isotopes, while the bromine mass spectrum
Fig. 1. GC–MS/MS total ion chromatogram of the four THMs in urine sample spiked with an individual final concentration of 10 g L−1 .
contains an ion cluster at m/z 171, 173, and 175 with known relative abundances of 79Br and 81Br isotopes. Similar isotope clusters are produced by fragmentation of molecules containing chlorine and bromine. Total ion chromatogram showing each THM analyte, including the internal standard and surrogate ion are shown in Fig. 1. Linear calibration curves were observed for all THM in the range of 39–20,240 ng L−1 , with an r2 value of 0.9998 or better (Table 2). Accordingly, the LODs for all THM were in the range of 11–27 ng L−1 (Table 2). Intra-, and inter-day variability of analytical measurements was always 70%). Results indicated a wide variability in measured THM concentrations, ranging from < LOD to 781, 935, and 195 ng L−1 for TCM, BDCM and DBCM, respectively (Table 3).
Sample ID
TCM
BDCM
DBCM
TBM
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
488 N.Qa 546 655 420 455 678 767 N.Qa N.D 430 368 567 N.Q 781 564 N.Q 594 N.D N.Q
150 163 139 N.Q N.Q N.Q N.Q N.Q N.Q 258 N.Q 149 N.Q N.Q 156 245 N.Q 320 935 N.Q
195 155 145 132 124 116 114 N.Q 129 N.Q 119 143 136 122 N.Q N.Q 111 N.Q N.Q 115
N.Q N.Q N.D N.Q N.Q N.D N.D N.D N.Q N.D N.Q N.D N.D N.Q N.D N.Q N.D N.D N.D N.Q
a
N.D =
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
A sensitive and fast method for trihalomethanes in urine using gas chromatography–triple quadrupole mass spectrometry Pantelis Charisiadis, Konstantinos C. Makris ∗ Cyprus International Institute for Environmental and Public Health in Association with Harvard School of Public Health, Cyprus University of Technology, Limassol, Cyprus
a r t i c l e
i n f o
Article history: Received 24 July 2013 Accepted 30 November 2013 Available online 6 December 2013 Keywords: Chlorine Disinfection by products Exposure Mass spectrometry Trihalomethanes Urine
a b s t r a c t Because of the plethora of exposure sources and routes through which humans are exposed to trihalomethanes (THM), the limitation of their short half-lives could be overcome, if a highly sensitive method was available to quantify urinary THM concentrations at sub-ppb levels. The objective of this study was to develop a fast and reliable method for the determination of the four THM analytes in human urine. A sensitive methodology was developed for THM in urine samples using gas chromatography coupled with triple quadrupole mass spectrometry (GC–QqQ-MS/MS) promoting its use in epidemiological and biomonitoring studies. The proposed methodology enjoys limits of detection similar to those reported in the literature (11–80 ng L−1 ) and the advantages of small initial urine volumes (15 mL) and fast analysis per sample (12 min) when compared with other methods. This is the first report using GC–QqQ-MS/MS for the determination of THM in urine samples. Because of its simplicity and less time-consuming nature, the proposed method could be incorporated into detailed (hundreds of participants’ urine samples) exposure assessment protocols providing valuable insight into the dose–response relationship of THM and cancer or pregnancy anomalies. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Despite the recent market introduction of disinfectants void of chlorine, hypochlorite is still widely used as a common microbial disinfectant in potable water and solid surfaces in households (floors, toilets, etc.). Inevitably, every-day household activities, such as cleaning (mopping) the house and use of chlorinated tap water for washing dishes, showering, laundry, and swimming in chlorinated pools present us with common examples of daily exposure sources to chlorine and its disinfection by-products (DBP). The formation of DBP is triggered by reactions between residual chlorine and the natural organic matter in water or in dust attached to household solid surfaces or perhaps aerosols [1]. A suite of chlorinated DBP may be formed, including the
Abbreviations: THM, trihalomethanes; TCM, trichloromethane; BDCM, bromodichloromethane; DBCM, dibromochloromethane; TBM, tribromomethane; DBP, disinfection by-products; MTBE, t-butyl methyl ether; FMU, first morning void urine; GC–QqQ-MS/MS, gas chromatography coupled with triple quadrupole mass spectrometry; MRM, multi reaction monitoring; EI, electron impact ionization; SPME, solid phase micro extraction; HS, headspace. ∗ Corresponding author at: Cyprus International Institute for Environmental and Public Health in Association with Harvard School of Public Health, Cyprus University of Technology, Irenes 95, Limassol 3041, Cyprus. Tel.: +357 25002398; fax: +357 25002676. E-mail address: [email protected] (K.C. Makris). 1570-0232/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jchromb.2013.11.060
trihalomethanes-THM, being the most abundant class of DBP in chlorinated tap water and currently regulated in both sides of the Atlantic Ocean (trichloromethane, bromodichloromethane, dibromochloromethane and tribromomethane) [2]. Numerous epidemiological studies showed the increased cancer risk for populations exposed to THM in potable water [3]. Increased risk of spontaneous abortion and preterm births among women who drink larger amounts of tap water was observed [4], suggesting a possible link between DBP and adverse pregnancy outcomes. Preterm birth is the primary cause of newborn deaths and the second major cause of death after pneumonia in children under five years [5]. Human absorption of THM may occur not only through ingestion of tap water, but also via inhalation and dermal uptake during showering, bathing, and swimming, with a different pattern of metabolism for the various routes of exposure and potentially large variation between-, and within-individuals [6]. The relatively high volatility of THM, especially for chloroform makes it difficult to accurately and precisely capture historic exposures to THM; when this is coupled with the very short biological half-life of THM in the order of approximately 1 h or less [7,8], it highlights the perplexity of accurately quantifying THM human exposure variability in urine samples [7,9–12]. Blood has been also considered as a biomarker of THM exposures, but quantified levels are of similar magnitude to those in urine (sub-ppb levels) [13–17]. We believe that the enhanced sensitivity of the analytical method required to quantify THM concentrations in urine or blood has hampered progress
18
P. Charisiadis, K.C. Makris / J. Chromatogr. B 947–948 (2014) 17–22
of epidemiological studies in gaining further insight into THM dose and disease process. The problem may be exacerbated by various THM exposure sources, making it difficult to capture exposure variability with only a few biomarker sample analyses, since there is typically an elevated cost and time-consuming methodology associated with urinary measurements. Because of the plethora of exposure sources and our daily exposures to THM via the aforementioned scenaria, the limitation of their short half-lives could be overcome, if a highly sensitive method was available to quantify urinary THM concentrations at parts per trillion levels. During the past couple of years, we have been systematically studying spatio-temporal THM exposure variability in the city of Nicosia, Cyprus, where first morning void urine samples were collected from recruited volunteers (n = ∼400). This has forced us to develop a sensitive and fast method to quantify low-level THM exposures in the participants. Herein, we described a simple liquid–liquid extraction procedure, using a small amount of urine volume and minimal total sample analysis time. The use of the sensitive gas chromatography coupled with triple quadrupole mass spectrometry (GC–QqQ-MS/MS) methodology in the multiple reactions monitoring (MRM) mode enabled us to reach detection limits close to those already reported in literature, but in a simpler and faster manner. To the best of our knowledge, there is no published study in the literature reporting measurements of urinary THM concentrations using GC–QqQ-MS/MS. The objective of this study was to develop an easy, fast and reliable method for the qualitative and quantitative determination of the four THM analytes in human urine. Our ultimate goal is to facilitate the inclusion of this method (because of its simplicity and less time consuming nature) into high-throughput exposure assessment measurements that could help epidemiological studies providing further insight into the dose–response relationship of DBP and cancer or pregnancy anomalies. 2. Materials and methods 2.1. Chemicals Decafluorobiphenyl, sodium sulphate, acetone and methanol GC grade were purchased from Sigma–Aldrich. t-Butyl methyl ether GC grade was purchased from Panreac (Barcelona, Spain). Trihalomethanes mix, containing trichloromethane (TCM), bromodichloromethane (BDCM), dibromochloromethane (DBCM), and tribromomethane (TBM), was purchased from Restek (Bellefonte, USA). 4-Bromofluorobenzene was purchased from Supelco (Bellefonte, USA). 2.2. Sampling, preservation and storage Urine samples (first morning void urine, FMU) were collected from 20 volunteers residing into two specific areas of the city of Nicosia, Cyprus. Our sampling team collected FMU urine samples in 60 mL polypropylene vials with minimum headspace; samples were temporarily stored at −20 ◦ C freezer until arrival in the laboratory, where they were either immediately analyzed, or stored in the −80 ◦ C freezer, until analysis. The study protocol was approved by the National Bioethics Committee of Cyprus (EEBK/EP/2012/17). 2.3. Standard solutions Stock solution of 200 mg L−1 was prepared from the initial concentrated trihalomethane mix of 2000 mg L−1 in methanol, and further diluted to prepare calibration and additive solutions. A surrogate solution of 1000 mg L−1 was prepared by adding 50 mg of decafluorobiphenyl into 50 mL of acetone. Stock solution of 200 mg L−1 was prepared from the initial and further diluted for the
working solution. Stock solution of 200 mg L−1 was prepared from the initial concentrated internal standard 4-bromofluorobenzene of 1000 mg L−1 in acetone and further diluted for the working solution. All solutions containing THM, internal standard and surrogate were kept away from light, prepared fresh, and stored always at −20 ◦ C in glass vials. 2.4. Sample preparation Our method development was based upon the principles of the EPA Method 551.1-1 for THM in drinking-water after several modifications which are described below. In detail, a liquid–liquid extraction protocol was optimized by mixing 15 mL urine sample (spiked with surrogate solution at a final concentration of 10 g L−1 ) with 2 mL of t-butyl methyl ether, adding 6.0 g of sodium sulfate to saturate the sample and shaking gently for 5 min in a lab shaker at 100 rpm. Following, samples were centrifuged for 1 min at 500 rpm for a clear phase separation. Half of mL of the organic phase was transferred into a GC autosampler screw top glass vial with blue PTFE/butyl rubber septa (Restek, USA), containing the internal standard solution at final concentration of 200 g L−1 . 2.5. GC–QqQ-MS/MS analysis Gas chromatography was performed in an Agilent 7890A gas chromatograph coupled with an Agilent 7000B triple quadrupole GC–QqQ-MS/MS system (Agilent Technologies, Waldbronn, Germany). Separation was performed using a 30 m × 0.25 mm × 0.25 m Rxi-5 ms (5% diphenyl/95% dimethyl polysiloxane) column from Restek. An Agilent 7693A Series automatic liquid sampler was used and the injection volume was 2 L with a 10.0 L ± 1% glass syringe (Agilent, Australia). Electron impact (EI) ionization was performed at electron energy of 70 eV. MS/MS was performed in the MRM mode. The quadrupole mass detector was operated at 150 ◦ C. The ion source temperature was set to 250 ◦ C, and the transfer line was set to 250 ◦ C. The column oven temperature programs were set to 30 ◦ C for 5 min, then to 100 ◦ C for 2.4 min with 50 ◦ C min−1 , and then to 300 ◦ C for 0.7 min with 80 ◦ C min−1 . Carrier gas was helium at 1 mL min−1 . Peaks were identified on the basis of their fragmentation patterns using the NIST Mass Spectral Search Program (NIST 08 version 2.0). The system was controlled by the software Mass Hunter Workstation (Agilent, rev. B.05.00). 2.6. Optimization of the GC–MS/MS MTBE extracts of the four THM in urine sample free of THM (concentrations of 1000 g L−1 ) were tested in full scan mode for the selection of the precursor ions and to optimize the MS/MS measurement of each analyte. In brief, MS spectra were obtained and analytes were identified by comparison to NIST mass spectra library (NIST 08 version 2.0). Mass spectra of the analytes in the ratio (m/z) 40–400 were recorded. Ions of the higher mass ratios and abundance were preferably chosen as the precursor ions in a way to increase the selectivity and sensitivity of the method. The chosen precursor ions were examined to identify their product ions in the daughter ion scan mode. The parent-daughter ion pairs used for multiple reaction monitoring (MRM) were chosen based on ion abundances in different collision energies while scanning cycles were kept constant to approximately 3 cycles per second. For each compound, molecular ion MS/MS spectra in different collision energies were obtained. Collision energies (5–60 eV) for the chosen parent-daughter ion pairs of each analyte were optimized to achieve maximum signal intensity, using nitrogen gas as the collision gas (Table 1).
P. Charisiadis, K.C. Makris / J. Chromatogr. B 947–948 (2014) 17–22
19
Table 1 Names, segments, retention time, dwell time, precursors and products ions of quantitative and confirmation transition, and collision energy for the trihalomethanes analysis. THMs
TCM BDCM DBCM TBM ISTD SR
Time segment (min)
4.800 5.900 6.800 7.800 9.800
Retention time (min)
5.561 6.325 7.324 8.371 8.856 10.307
Dwell time (ms)
75 75 75 37.5 37.5 75
Selective ions (m/z) – MRM transitions
Quantitative (Q)
Collision energy (eV)
Confirmation (q1)
Collision energy (eV)
Confirmation (q2)
Collision energy (eV)
Confirmation (q3)
Collision energy (eV)
83 → 47 83 → 47 129 → 48 173 → 94 174 → 95 334 → 265
35 33 43 48 15 33
83 → 48 85 → 47 127 → 48 171 → 92 174 → 75 334 → 315
42 37 46 48 55 23
85 → 47 129 → 48 208 → 129 252 → 173 176 → 95 334 → 165
38 43 2 2 38 60
85 → 49 127 → 48 131 → 50 175 → 94 176 → 75 334 → 234
37 45 46 55 17 45
2.7. Calibration curve and linearity As a means of eliminating any matrix effects, procedural calibration standards were prepared in pooled urine sample with negligible THM measured concentrations. An aliquot of 15 mL the pooled urine sample was placed in a 20 mL screw top glass vial with screw cap with hole and PTFE/silicone septum. The surrogate standard solution and the desired concentration dilution standard (in methanol) were added. The curve was established by measuring ten samples of urinary THM at final concentration from 39 to 20,240 ng L−1 (39, 79, 158, 316, 633, 1265, 2530, 5060, 10,120, and 20,240 ng L−1 ) using GC–QqQ-MS/MS in the MRM mode. Quantitative analysis was based on peak area measurements as ratios versus peak area of internal standard. All samples were treated as following the sample preparation procedure. The regression coefficient was always 0.9998 or better (Table 2). 2.8. Method detection limit The limit of detection (LOD) of the proposed methodology was determined at the lowest concentration of each analyte. The limit of quantification (LOQ) and limit of detection (LOD) were determined based on the standard deviation of eight measurements of the spiked blank urine samples [18]. The limit of detection (LOD) was calculated by adding 3× the standard deviation of the response lowest concentration of the calibration curve. Accordingly, the limit of quantification (LOQ) was equal to 3× the LOD. 2.9. Stability of THMs in urine samples—storage effect A stability experiment was performed to evaluate the stability of THM in stored urine samples during time of sampling and storage at −80 ◦ C. In an attempt to simulate as much as possible realistic sampling conditions, a pooled urine sample (with negligible THM measured concentrations) was placed in sampling vials. Samples were spiked with a concentration of THM in a final concentration of 10 g L−1 for each analyte (four analytes). They were kept in the −80 ◦ C freezer, and at specific time intervals, vials were sacrificed and samples were analyzed for THM. 3. Results 3.1. Calibration curve and recovery Ultrapure water was always used as the background matrix, because it did not contain residual chlorine or THM, being < always LOD. Selective ion – MRM optimized transitions for THM are depicted in Table 1. Mass spectra of the four trihalomethanes contain characteristic patterns caused by chlorine and bromine isotopes. For example, the trichloromethane mass spectrum contains an ion cluster at m/z 83, 85, and 87 with known relative abundances of 35Cl and 37Cl isotopes, while the bromine mass spectrum
Fig. 1. GC–MS/MS total ion chromatogram of the four THMs in urine sample spiked with an individual final concentration of 10 g L−1 .
contains an ion cluster at m/z 171, 173, and 175 with known relative abundances of 79Br and 81Br isotopes. Similar isotope clusters are produced by fragmentation of molecules containing chlorine and bromine. Total ion chromatogram showing each THM analyte, including the internal standard and surrogate ion are shown in Fig. 1. Linear calibration curves were observed for all THM in the range of 39–20,240 ng L−1 , with an r2 value of 0.9998 or better (Table 2). Accordingly, the LODs for all THM were in the range of 11–27 ng L−1 (Table 2). Intra-, and inter-day variability of analytical measurements was always 70%). Results indicated a wide variability in measured THM concentrations, ranging from < LOD to 781, 935, and 195 ng L−1 for TCM, BDCM and DBCM, respectively (Table 3).
Sample ID
TCM
BDCM
DBCM
TBM
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
488 N.Qa 546 655 420 455 678 767 N.Qa N.D 430 368 567 N.Q 781 564 N.Q 594 N.D N.Q
150 163 139 N.Q N.Q N.Q N.Q N.Q N.Q 258 N.Q 149 N.Q N.Q 156 245 N.Q 320 935 N.Q
195 155 145 132 124 116 114 N.Q 129 N.Q 119 143 136 122 N.Q N.Q 111 N.Q N.Q 115
N.Q N.Q N.D N.Q N.Q N.D N.D N.D N.Q N.D N.Q N.D N.D N.Q N.D N.Q N.D N.D N.D N.Q
a
N.D =
