- Email: [email protected]
Determination of five 4-hydroxycoumarin rodenticides in animal liver tissues by ion chromatography with fluorescence detection
Journal of Chromatography A, 1155 (2007) 57–61
Determination of five 4-hydroxycoumarin rodenticides in animal liver tissues by ion chromatography with fluorescence detection Mi-Cong Jin a,b , Xiao-Hong Chen b , Yan Zhu a,∗ a
Department of Chemistry, Xixi Campus, Zhejiang University, Hangzhou 310028, China Ningbo Municipal Center for Disease Control and Prevention, Ningbo 315010, China
b
Available online 23 December 2006
Abstract A novel analytical method is proposed for rapid simultaneous determination of five 4-hydroxycoumarin rodenticides in animal liver tissues by eluent generator reagent free ion chromatography (RFIC) with fluorescence detection. Rodenticides were initially extracted from homogenized animal liver tissues with ethyl acetate and the extracts subjected to a solid-phase extraction process using Oasis HLB cartridges. The IC separation was carried out on an IonPac® AS11 analytical column (250 mm × 4.0 mm) using gradient KOH containing 10% acetonitrile as organic modifier at a constant flow rate of 1.0 mL/min. The analytes were detected by fluorescence at an excitation wavelength of 270 nm and an emission wavelength of 380 nm. The average recoveries of the objective compounds spiked in animal liver tissues were between 81% and 98%. The limits of quantification (LOQs) were 0.004–0.010 mg/kg for them. Within-day and day-to-day relative standard deviations (RSD) were less than 8.5% and 9.7%, respectively. It was confirmed that this method could be used in a toxicological analysis. © 2007 Elsevier B.V. All rights reserved. Keywords: Ion chromatography; 4-Hydroxycoumarin; Rodenticide; Liver
1. Introduction The 4-hydroxycoumarin rodenticides, mainly including warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum (Fig. 1), as the second-generation of anticoagulant rodenticides, were widely used for pest control in China. The increased commercial availability of these compounds has resulted in an increase in accidental and intentional ingestion for both animals and human beings. Analytical methods for rapid determination of these five rodenticides are required both for diagnosis and effective treatment of the intoxication and for forensic purposes. Although, a number of literatures have been described for the analysis of these rodenticides in biological matrices, the main methods for their determination were high-performance liquid chromatography (HPLC) coupled with various detectors such as ultraviolet [1,2], fluorescence [3–9], and mass spectrometer [10–13]. Unfortunately, no method was reported by ion chromatography (IC). IC with the advantage of no time-consuming derivatization procedures and a little or no toxic organic solvents offered an
∗
Corresponding author. Tel.: +86 571 88273637; fax: +86 571 88273637. E-mail address: [email protected] (Y. Zhu).
0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2006.12.074
attractive alternative to traditional HPLC methods. Our previous work [14,15] reported that chlorophenolic compounds were separated by IC and detected by atmospheric pressure chemical ionization mass spectrometry (APCI). We considered that both chlorophenolic compounds and 4-hydroxycoumarin rodenticides are weak organic acids owing to the phenolic hydroxyl in their chemical structures, and the 4-hydroxycoumarin rodenticides could be separated by IC. In this report, a rapid and convenient multicomponent method is presented for the determination of anticoagulants in animal liver tissues. Five rodenticides, warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, were simultaneously extracted from animal liver tissues. The extracts were cleaned-up by solid-phase extraction (SPE) and determined by IC with fluorescence detection. 2. Experimental 2.1. Instrumentation A Dionex Model ICS2000 ion chromatograph equipped with an isocratic pump, a column thermostat, an AS40 automated sampler, a 25 L sample loop and an ASRS-ULTRA-4 mm self-regenerating suppressor (recycle mode) (Dionex, Sunnyvale, CA, USA). A Dionex IonPac AG11 guard column
58
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
from Millipore (Molsheim, France). Warfarin (>98%), coumatetralyl (>99%), bromadiolone (>99%), flocoumafen (>98%), brodifacoum (>98%), mixed isomers of analytical standard grade, were purchased from Sigma (St. Louis, MO, USA). Drug-free liver samples were collected from The Best Supermarket in Ningbo (Zhejiang, China). Anhydrous sodium sulfate, dichloromethane, hydrochloric acid were obtained from Shanghai Reagent Company (Shanghai, China). 2.3. Preparation of standard stock solutions The five authentic standards were accurately weighed, transferred to volumetric flasks and dissolved in MeOH to make individual stock solutions of 1.0 mg/mL. These solutions were thoroughly mixed and stored at 4 ◦ C in tightly closed bottles until use. Secondary standard solutions of warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum were prepared with MeOH at the concentration of 10.0 g/mL, which were used for spiking in animal liver tissues. 2.4. Preparation of sample
Fig. 1. Chemical structures of warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum.
(50 mm × 4 mm, i.d.) and a Dionex IonPac AS11 separation column (250 mm × 4 mm, i.d.) were used throughout. The KOH eluent was achieved by a KOH eluent generator (Dionex, Sunnyvale, CA, USA) and the organic solvent was added by a Waters 510 pump. The ICS2000 system was controlled on a personal computer equipped with Dionex Chromeleon 6.5 Software and the flow of MeCN was controlled by Waters 680 model gradient system. The fluorescence detector (G1321A) (Agilent Technologies, Germany) was controlled, and data were analyzed, on a personal computer equipped with LC/MSD Trap Software 4.2 (Bruker). All PEEK (polyetheretherketone) tubing used for connection had an i.d. of 0.25 mm (Agilent Technologies, Germany). Samples were mixed in an IKA T25 basic ULTRA-TURRAX blender (IKA Works, Guangzhou, China). 2.2. Chemicals and solvents The acetonitrile (MeCN) and methanol (MeOH) (Merck, Darmstadt, Germany) used were HPLC grade. The water employed was supplied by a Milli-Q water purification system
For this study, liver tissues (2.0 g) and ethyl acetate (10.0 mL) were added into a 50 mL polypropylene centrifugal tube (Huadong Medicine, Hangzhou, China), and mixed by an IKA T25 basic ULTRA-TURRAX blender for 2 min. The extract was transferred to another 50 mL polypropylene centrifugal tube. The tissues were extracted with ethyl acetate (10.0 mL) again as mentioned above. The organic solvents combined and evaporated by a vacuum system. Then, the residues were made up to 50.0 mL with Milli-Q water, acidified to pH 2.5 by 0.10 mol/L hydrochloric acid, and uploaded on an Oasis HLB (200 mg, Waters, USA) cartridge that first was conditioned with 4.0 mL dichloromethane, 4.0 mL methanol and 5.0 mL water. The flow rate of the samples was 5.0 mL/min. The elution was made with 5.0 mL MeOH and 5.0 mL dichloromethane, then evaporated to dryness by a gentle stream of nitrogen, and the residues were reconstituted with 0.5 mL of the initial IC eluent and filtered through a 0.45 m nylon syringe filter (Agilent Technologies, Germany). Finally, a 25 L aliquot was injected into the IC system. For the extraction recovery experiments, the animal liver tissues were spiked at concentration levels of 0.05, 1.0 and 5.0 mg/kg for each objective compounds. 2.5. Ion chromatography analysis The IC method utilized an IonPac® AS11 analytical column (250 mm × 4.0 mm, Dionex) and an IonPac® AG11 guard column (50 mm × 4.0 mm, Dionex). The mobile phase was gradient KOH containing 10% MeCN modifier at a constant flow rate of 1.0 mL/min. The KOH gradient program was as follows: 0 → 2.0 min, 2.0 mmol/L KOH; 2.0 → 7.0 min, 2.0 → 20.0 mmol/L KOH (linear gradient), 7.0 → 10.0 min, 20.0 mmol/L KOH. Column temperature was held constant at 35 ◦ C. Detection was performed on a fluorescence detector at λex = 270 nm, λem = 380nm.
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
59
3. Results and discussion 3.1. Solid-phase extraction The first attempt at liver extraction was only based on homogenization in ethyl acetate, then the extract were vaporized and the residues were reconstituted to directly inject into the IC system for analyzing the objective compounds. Although the procedure was quick, the extracts were dirty. Column performance degraded after several hundred injections, evidenced by broad asymmetrical peaks and higher column pressure. This problem appeared to affect the whole IC column, as frequent guard column changes did not prevent deterioration of the analytical column. As a result, solid-phase extraction (SPE) was necessary for extract pretreatment. A solid-phase cartridge based on either C18 bonded silica or hydrophilic polymer was found to give acceptable results. The pretreatment method was validated using a Waters C18 Oasis HLB SPE cartridge. 3.2. Optimizing the separation Although the five rodenticides seem to have greater hydrophobic moieties, they still have the characteristics of the weakly organic acid, and can be deprotonated in basic medium. This property provides a possibility to retain 4-hydroxycoumarin rodenticides onto an anion-exchange column. Therefore, the diverse basic solutions of potassium hydroxide (KOH) eluents were chosen to optimize the separation and the peak shape for the five rodenticides studied. In the preliminary experiments, a peak broadening was observed for all the objective compounds with KOH eluent at lower concentration or without organic modifier in the mobile phase. Peak broadening could be caused by the interaction between the phenyl-group of 4-hydroxycoumarin rodenticides with the framework of the stationary phase. Therefore, MeCN was added to modify the polarity of the mobile phase. It was found that peak broadening was greatly decreased by adding 10% (v/v) of MeCN to the mobile phase, leading to sharp and symmetrical peaks for all the five rodenticdes. As a result, 10% (v/v) MeCN was added to the mobile phase. The effect of KOH concentrations and MeCN on retention times and resolutions of the five rodenticides at 0.5 mg/L each is shown in Fig. 2. As can be seen that the concentration of KOH at 5 mmol/L is too higher for the separation of warfarin and coumatetralyl, and is too lower for the separation of flocoumafen and brodifacoum. The best separation was obtained with a gradient mobile phase of KOH/MeCN. The mobile phase used in
Fig. 2. The effect of KOH concentrations and MeCN on retention times and resolutions of the five rodenticides at 0.5 mg/L each. Peaks identification: (1) warfarin; (2) coumatetralyl; (3) bromadiolone; (4) flocoumafen; (5) brodifacoum. Chromatographic conditions: IC column, an IonPac® AS11 analytical column (250 mm × 4.0 mm, Dionex) and an IonPac® AG11 guard column (50 mm × 4.0 mm, Dionex); flow rate, 1.0 mL/min; column temperature, 35 ◦ C; sample injection volume, 25 L; fluorescence detection, λex = 270 nm, λem = 380 nm; mobile phase, (A) 5 mmol/L KOH without MeCN; (B) 5 mmol/L KOH/10% MeCN; (C) gradient KOH/10% MeCN, the KOH gradient program: 0 → 2.0 min, 2.0 mmol/L KOH; 2.0 → 7.0 min, 2.0 → 20.0 mmol/L KOH (linear gradient), 7.0 → 10.0 min, 20.0 mmol/L KOH.
Table 1 Quality parameters of the IC-FL method for the determination of five rodenticides Compound
Regression equation
Linear range (mg/L)a
Coefficient of determination (r2 )
Limit of quantification for FL (mg/kg)b
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
A = 188.1C − 1.860 A = 357.4C − 1.362 A = 246.8C − 2.437 A = 281.2C − 2.312 A = 332.3C − 1.265
0.04–50.0 0.02–50.0 0.02–50.0 0.02–50.0 0.02–50.0
0.998 0.994 0.993 0.992 0.991
0.010 0.004 0.005 0.005 0.005
a b
Determined by the rodenticides standard solutions. 2.0 g of animal liver tissues.
60
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
the present study is a gradient KOH solution, containing 10% MeCN as modifier to obtain sharp and symmetrical peaks, as shown in Fig. 2C. Lowering the MeCN concentration caused broadening the rodenticide peaks and prolonging the analytes retention times. A concentration of 10% MeCN was applied to the subsequent study.
Day-to-day precision was also evaluated by performing replicates for three spiked samples each day on eight different days within a 2-week period (added 0.05, 1.0 and 5.0 mg/kg for each rodenticides). Day-to-day precision (RSD) on the basis of the
3.3. Validation of the IC analysis method Calibration curves for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum were tested in a concentration range of 0.02–50.0 mg/L. Calibration curve of each rodenticides showed good linearity in the range of 0.04–50.0 mg/L with coefficients of determination (r2 > 0.991). The limits of quantification (LOQs) for the five rodenticides were determined using a liver sample free of rodenticides spiked at low concentration of 0.05 mg/kg, extraction with ethyl acetate, detection with fluorescence and evaluation by the criterion that the signal to noise ratio (S/N) should be >10, for quantification purposes. The LOQs were 0.004–0.010 mg/kg for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, respectively. Table 1 shows the linear regression equation, and the LOQs for each analytes. The method was validated by standard addition to a animal liver sample with warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, at two different concentrations (0.05, 1.0 and 5.0 mg/kg). Recoveries for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum were achieved in the range of 81–98%, as shown in Table 2. Within-day precision was evaluated by continuously performing five replicates each day within 3 days for the determination of 3 spiked samples (added 0.05, 1.0 and 5.0 mg/kg for each rodenticides). Within-day precision (RSD) on the basis of the five rodenticides content was between 3.8% and 8.5%, as shown in Table 2.
Fig. 3. Chromatograms of a spiked liver sample (A) and a real poisoned dog liver sample (B). Peaks identification: (1) warfarin; (2) coumatetralyl; (3) bromadiolone; (4) flocoumafen; (5) brodifacoum. Chromatographic conditions: IC column, an IonPac® AS11 analytical column (250 mm × 4.0 mm, Dionex) and an IonPac® AG11 guard column (50 mm × 4.0 mm, Dionex); flow rate, 1.0 mL/min; column temperature, 35 ◦ C; sample injection volume, 25 L; fluorescence detection, λex = 270 nm, λem = 380 nm; mobile phase, gradient KOH/10% MeCN, the KOH gradient program: 0 → 2.0 min, 2.0 mmol/L KOH; 2.0 → 7.0 min, 2.0 → 20.0 mmol/L KOH (linear gradient), 7.0 → 10.0 min, 20.0 mmol/L KOH.
Table 2 The precision and accuracy for the IC-FL method Compound
Added (mg/kg)
Founda (mg/kg)
Recovery (%)
RSD (%) Within-day b
Day-to-dayc
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
0.05
0.049 0.045 0.048 0.047 0.043
± ± ± ± ±
0.003 0.003 0.004 0.004 0.002
98 90 96 94 86
6.1 6.7 8.3 8.5 4.7
8.4 7.4 9.1 9.7 7.3
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
1.0
0.97 0.85 0.85 0.81 0.83
± ± ± ± ±
0.06 0.06 0.07 0.05 0.07
97 85 85 81 83
6.2 7.1 8.2 6.2 8.4
8.2 7.4 9.4 7.2 8.5
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
5.0
4.64 4.70 4.62 4.31 4.23
± ± ± ± ±
0.33 0.22 0.17 0.21 0.25
93 94 92 86 85
7.1 4.7 3.8 4.9 5.9
7.3 6.3 5.6 5.8 6.3
a b c
The mean value was determined in 3 days (n = 5 replicates × 3 days, x¯ ± s). n = 5 replicates × 3 days. n = 2 replicates × 8 days within a 2-week period.
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
five rodenticides content was between 5.6% and 9.7%, as shown in Table 2. 3.4. Application of IC-FL to the real sample A spiked liver sample and a real poisoned liver sample were determined by IC with fluorescence detection. The concentrations in the spiked liver sample were 0.10, 0.010, 0.05, 0.05 and 0.05 mg/kg for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, respectively. The chromatograms of the two samples are shown in Fig. 3. As can be seen that the five objective compounds in the spiked sample are well separation, and the real poisoned sample contained brodifacoum of 0.015 mg/kg. 4. Conclusions A novel, sensitive, and accurate IC-FL method, with the proposed SPE procedures using an Oasis HLB cartridge has been developed for the simultaneous determination of warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum in animal liver tissues with recoveries above 81% for all the target compounds. The advantages of the described method are of lower amount of organic solvent and more compatible with the environment than the reported high-performance liquid chromatographic method. It has been applied to the determination of the 4-hydroxycoumarin rodenticides in animal liver tissues.
61
Acknowledgements This work was supported by a grant from the Medicine Science Foundation of Ningbo, Zhejiang Province, China (2004056) and by National Natural Science Foundation of China (Nos. 20375035, 20527005). References [1] F.Y. Guan, L. Liu, Y. Luo, Chinese J. Anal. Chem. 23 (1995) 159. [2] F.Y. Guan, I. Akira, S. Hiroshi, W. Kanako, K. Takeshi, S. Osamu, J. Pharm. Biomed. Anal. 21 (1999) 179. [3] L.J. Felice, T. Chalermchaikit, M.J. Murphy, J. Anal. Toxicol. 15 (1991) 126. [4] J.F. Lawrence, C. Thongchai, J.M. Michael, J. Anal. Toxicol. 15 (1991) 126. [5] K. Hunter, J. Chromatogr. 321 (1985) 255. [6] K. Hunter, E.A. Sharp, A. Newton, J. Chromatogr. 435 (1988) 83. [7] A.P. Eugenie, H. Jan den, F.S. Jean, A.W. Frederik, J. Anal. Toxicol. 19 (1995) 557. [8] V. Fauconnet, H. Pouliquen, L. Pinault, J. Anal. Toxicol. 21 (1997) 548. [9] C. Thongchai, J.F. Lawrence, J.M. Michanel, J. Anal. Toxicol. 17 (1993) 56. [10] F.Y. Guan, A. Ishii, H. Seno, K. Watanabe-Suzuki, T. Kumazawa, O. Suzuki, J. Chromatogr. B 731 (1999) 155. [11] M.C. Jin, X.H. Chen, Rapid Commun. Mass Spectrom. 20 (2006) 2741. [12] M.C. Jin, X.H. Chen, H.P. Chen, J. Liq. Chrom. Relat. Tech. 29 (2006) 2641. [13] T. Grobosch, B. Angelow, L. Sch¨onberg, D. Lampe, J. Anal. Toxicol. 30 (2006) 281. [14] M.C. Jin, Y.W. Yang, Anal. Chim. Acta 566 (2006) 193. [15] M.C. Jin, Y. Zhu, J. Chromatogr. A 1118 (2006) 111.
Determination of five 4-hydroxycoumarin rodenticides in animal liver tissues by ion chromatography with fluorescence detection Mi-Cong Jin a,b , Xiao-Hong Chen b , Yan Zhu a,∗ a
Department of Chemistry, Xixi Campus, Zhejiang University, Hangzhou 310028, China Ningbo Municipal Center for Disease Control and Prevention, Ningbo 315010, China
b
Available online 23 December 2006
Abstract A novel analytical method is proposed for rapid simultaneous determination of five 4-hydroxycoumarin rodenticides in animal liver tissues by eluent generator reagent free ion chromatography (RFIC) with fluorescence detection. Rodenticides were initially extracted from homogenized animal liver tissues with ethyl acetate and the extracts subjected to a solid-phase extraction process using Oasis HLB cartridges. The IC separation was carried out on an IonPac® AS11 analytical column (250 mm × 4.0 mm) using gradient KOH containing 10% acetonitrile as organic modifier at a constant flow rate of 1.0 mL/min. The analytes were detected by fluorescence at an excitation wavelength of 270 nm and an emission wavelength of 380 nm. The average recoveries of the objective compounds spiked in animal liver tissues were between 81% and 98%. The limits of quantification (LOQs) were 0.004–0.010 mg/kg for them. Within-day and day-to-day relative standard deviations (RSD) were less than 8.5% and 9.7%, respectively. It was confirmed that this method could be used in a toxicological analysis. © 2007 Elsevier B.V. All rights reserved. Keywords: Ion chromatography; 4-Hydroxycoumarin; Rodenticide; Liver
1. Introduction The 4-hydroxycoumarin rodenticides, mainly including warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum (Fig. 1), as the second-generation of anticoagulant rodenticides, were widely used for pest control in China. The increased commercial availability of these compounds has resulted in an increase in accidental and intentional ingestion for both animals and human beings. Analytical methods for rapid determination of these five rodenticides are required both for diagnosis and effective treatment of the intoxication and for forensic purposes. Although, a number of literatures have been described for the analysis of these rodenticides in biological matrices, the main methods for their determination were high-performance liquid chromatography (HPLC) coupled with various detectors such as ultraviolet [1,2], fluorescence [3–9], and mass spectrometer [10–13]. Unfortunately, no method was reported by ion chromatography (IC). IC with the advantage of no time-consuming derivatization procedures and a little or no toxic organic solvents offered an
∗
Corresponding author. Tel.: +86 571 88273637; fax: +86 571 88273637. E-mail address: [email protected] (Y. Zhu).
0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2006.12.074
attractive alternative to traditional HPLC methods. Our previous work [14,15] reported that chlorophenolic compounds were separated by IC and detected by atmospheric pressure chemical ionization mass spectrometry (APCI). We considered that both chlorophenolic compounds and 4-hydroxycoumarin rodenticides are weak organic acids owing to the phenolic hydroxyl in their chemical structures, and the 4-hydroxycoumarin rodenticides could be separated by IC. In this report, a rapid and convenient multicomponent method is presented for the determination of anticoagulants in animal liver tissues. Five rodenticides, warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, were simultaneously extracted from animal liver tissues. The extracts were cleaned-up by solid-phase extraction (SPE) and determined by IC with fluorescence detection. 2. Experimental 2.1. Instrumentation A Dionex Model ICS2000 ion chromatograph equipped with an isocratic pump, a column thermostat, an AS40 automated sampler, a 25 L sample loop and an ASRS-ULTRA-4 mm self-regenerating suppressor (recycle mode) (Dionex, Sunnyvale, CA, USA). A Dionex IonPac AG11 guard column
58
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
from Millipore (Molsheim, France). Warfarin (>98%), coumatetralyl (>99%), bromadiolone (>99%), flocoumafen (>98%), brodifacoum (>98%), mixed isomers of analytical standard grade, were purchased from Sigma (St. Louis, MO, USA). Drug-free liver samples were collected from The Best Supermarket in Ningbo (Zhejiang, China). Anhydrous sodium sulfate, dichloromethane, hydrochloric acid were obtained from Shanghai Reagent Company (Shanghai, China). 2.3. Preparation of standard stock solutions The five authentic standards were accurately weighed, transferred to volumetric flasks and dissolved in MeOH to make individual stock solutions of 1.0 mg/mL. These solutions were thoroughly mixed and stored at 4 ◦ C in tightly closed bottles until use. Secondary standard solutions of warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum were prepared with MeOH at the concentration of 10.0 g/mL, which were used for spiking in animal liver tissues. 2.4. Preparation of sample
Fig. 1. Chemical structures of warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum.
(50 mm × 4 mm, i.d.) and a Dionex IonPac AS11 separation column (250 mm × 4 mm, i.d.) were used throughout. The KOH eluent was achieved by a KOH eluent generator (Dionex, Sunnyvale, CA, USA) and the organic solvent was added by a Waters 510 pump. The ICS2000 system was controlled on a personal computer equipped with Dionex Chromeleon 6.5 Software and the flow of MeCN was controlled by Waters 680 model gradient system. The fluorescence detector (G1321A) (Agilent Technologies, Germany) was controlled, and data were analyzed, on a personal computer equipped with LC/MSD Trap Software 4.2 (Bruker). All PEEK (polyetheretherketone) tubing used for connection had an i.d. of 0.25 mm (Agilent Technologies, Germany). Samples were mixed in an IKA T25 basic ULTRA-TURRAX blender (IKA Works, Guangzhou, China). 2.2. Chemicals and solvents The acetonitrile (MeCN) and methanol (MeOH) (Merck, Darmstadt, Germany) used were HPLC grade. The water employed was supplied by a Milli-Q water purification system
For this study, liver tissues (2.0 g) and ethyl acetate (10.0 mL) were added into a 50 mL polypropylene centrifugal tube (Huadong Medicine, Hangzhou, China), and mixed by an IKA T25 basic ULTRA-TURRAX blender for 2 min. The extract was transferred to another 50 mL polypropylene centrifugal tube. The tissues were extracted with ethyl acetate (10.0 mL) again as mentioned above. The organic solvents combined and evaporated by a vacuum system. Then, the residues were made up to 50.0 mL with Milli-Q water, acidified to pH 2.5 by 0.10 mol/L hydrochloric acid, and uploaded on an Oasis HLB (200 mg, Waters, USA) cartridge that first was conditioned with 4.0 mL dichloromethane, 4.0 mL methanol and 5.0 mL water. The flow rate of the samples was 5.0 mL/min. The elution was made with 5.0 mL MeOH and 5.0 mL dichloromethane, then evaporated to dryness by a gentle stream of nitrogen, and the residues were reconstituted with 0.5 mL of the initial IC eluent and filtered through a 0.45 m nylon syringe filter (Agilent Technologies, Germany). Finally, a 25 L aliquot was injected into the IC system. For the extraction recovery experiments, the animal liver tissues were spiked at concentration levels of 0.05, 1.0 and 5.0 mg/kg for each objective compounds. 2.5. Ion chromatography analysis The IC method utilized an IonPac® AS11 analytical column (250 mm × 4.0 mm, Dionex) and an IonPac® AG11 guard column (50 mm × 4.0 mm, Dionex). The mobile phase was gradient KOH containing 10% MeCN modifier at a constant flow rate of 1.0 mL/min. The KOH gradient program was as follows: 0 → 2.0 min, 2.0 mmol/L KOH; 2.0 → 7.0 min, 2.0 → 20.0 mmol/L KOH (linear gradient), 7.0 → 10.0 min, 20.0 mmol/L KOH. Column temperature was held constant at 35 ◦ C. Detection was performed on a fluorescence detector at λex = 270 nm, λem = 380nm.
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
59
3. Results and discussion 3.1. Solid-phase extraction The first attempt at liver extraction was only based on homogenization in ethyl acetate, then the extract were vaporized and the residues were reconstituted to directly inject into the IC system for analyzing the objective compounds. Although the procedure was quick, the extracts were dirty. Column performance degraded after several hundred injections, evidenced by broad asymmetrical peaks and higher column pressure. This problem appeared to affect the whole IC column, as frequent guard column changes did not prevent deterioration of the analytical column. As a result, solid-phase extraction (SPE) was necessary for extract pretreatment. A solid-phase cartridge based on either C18 bonded silica or hydrophilic polymer was found to give acceptable results. The pretreatment method was validated using a Waters C18 Oasis HLB SPE cartridge. 3.2. Optimizing the separation Although the five rodenticides seem to have greater hydrophobic moieties, they still have the characteristics of the weakly organic acid, and can be deprotonated in basic medium. This property provides a possibility to retain 4-hydroxycoumarin rodenticides onto an anion-exchange column. Therefore, the diverse basic solutions of potassium hydroxide (KOH) eluents were chosen to optimize the separation and the peak shape for the five rodenticides studied. In the preliminary experiments, a peak broadening was observed for all the objective compounds with KOH eluent at lower concentration or without organic modifier in the mobile phase. Peak broadening could be caused by the interaction between the phenyl-group of 4-hydroxycoumarin rodenticides with the framework of the stationary phase. Therefore, MeCN was added to modify the polarity of the mobile phase. It was found that peak broadening was greatly decreased by adding 10% (v/v) of MeCN to the mobile phase, leading to sharp and symmetrical peaks for all the five rodenticdes. As a result, 10% (v/v) MeCN was added to the mobile phase. The effect of KOH concentrations and MeCN on retention times and resolutions of the five rodenticides at 0.5 mg/L each is shown in Fig. 2. As can be seen that the concentration of KOH at 5 mmol/L is too higher for the separation of warfarin and coumatetralyl, and is too lower for the separation of flocoumafen and brodifacoum. The best separation was obtained with a gradient mobile phase of KOH/MeCN. The mobile phase used in
Fig. 2. The effect of KOH concentrations and MeCN on retention times and resolutions of the five rodenticides at 0.5 mg/L each. Peaks identification: (1) warfarin; (2) coumatetralyl; (3) bromadiolone; (4) flocoumafen; (5) brodifacoum. Chromatographic conditions: IC column, an IonPac® AS11 analytical column (250 mm × 4.0 mm, Dionex) and an IonPac® AG11 guard column (50 mm × 4.0 mm, Dionex); flow rate, 1.0 mL/min; column temperature, 35 ◦ C; sample injection volume, 25 L; fluorescence detection, λex = 270 nm, λem = 380 nm; mobile phase, (A) 5 mmol/L KOH without MeCN; (B) 5 mmol/L KOH/10% MeCN; (C) gradient KOH/10% MeCN, the KOH gradient program: 0 → 2.0 min, 2.0 mmol/L KOH; 2.0 → 7.0 min, 2.0 → 20.0 mmol/L KOH (linear gradient), 7.0 → 10.0 min, 20.0 mmol/L KOH.
Table 1 Quality parameters of the IC-FL method for the determination of five rodenticides Compound
Regression equation
Linear range (mg/L)a
Coefficient of determination (r2 )
Limit of quantification for FL (mg/kg)b
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
A = 188.1C − 1.860 A = 357.4C − 1.362 A = 246.8C − 2.437 A = 281.2C − 2.312 A = 332.3C − 1.265
0.04–50.0 0.02–50.0 0.02–50.0 0.02–50.0 0.02–50.0
0.998 0.994 0.993 0.992 0.991
0.010 0.004 0.005 0.005 0.005
a b
Determined by the rodenticides standard solutions. 2.0 g of animal liver tissues.
60
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
the present study is a gradient KOH solution, containing 10% MeCN as modifier to obtain sharp and symmetrical peaks, as shown in Fig. 2C. Lowering the MeCN concentration caused broadening the rodenticide peaks and prolonging the analytes retention times. A concentration of 10% MeCN was applied to the subsequent study.
Day-to-day precision was also evaluated by performing replicates for three spiked samples each day on eight different days within a 2-week period (added 0.05, 1.0 and 5.0 mg/kg for each rodenticides). Day-to-day precision (RSD) on the basis of the
3.3. Validation of the IC analysis method Calibration curves for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum were tested in a concentration range of 0.02–50.0 mg/L. Calibration curve of each rodenticides showed good linearity in the range of 0.04–50.0 mg/L with coefficients of determination (r2 > 0.991). The limits of quantification (LOQs) for the five rodenticides were determined using a liver sample free of rodenticides spiked at low concentration of 0.05 mg/kg, extraction with ethyl acetate, detection with fluorescence and evaluation by the criterion that the signal to noise ratio (S/N) should be >10, for quantification purposes. The LOQs were 0.004–0.010 mg/kg for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, respectively. Table 1 shows the linear regression equation, and the LOQs for each analytes. The method was validated by standard addition to a animal liver sample with warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, at two different concentrations (0.05, 1.0 and 5.0 mg/kg). Recoveries for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum were achieved in the range of 81–98%, as shown in Table 2. Within-day precision was evaluated by continuously performing five replicates each day within 3 days for the determination of 3 spiked samples (added 0.05, 1.0 and 5.0 mg/kg for each rodenticides). Within-day precision (RSD) on the basis of the five rodenticides content was between 3.8% and 8.5%, as shown in Table 2.
Fig. 3. Chromatograms of a spiked liver sample (A) and a real poisoned dog liver sample (B). Peaks identification: (1) warfarin; (2) coumatetralyl; (3) bromadiolone; (4) flocoumafen; (5) brodifacoum. Chromatographic conditions: IC column, an IonPac® AS11 analytical column (250 mm × 4.0 mm, Dionex) and an IonPac® AG11 guard column (50 mm × 4.0 mm, Dionex); flow rate, 1.0 mL/min; column temperature, 35 ◦ C; sample injection volume, 25 L; fluorescence detection, λex = 270 nm, λem = 380 nm; mobile phase, gradient KOH/10% MeCN, the KOH gradient program: 0 → 2.0 min, 2.0 mmol/L KOH; 2.0 → 7.0 min, 2.0 → 20.0 mmol/L KOH (linear gradient), 7.0 → 10.0 min, 20.0 mmol/L KOH.
Table 2 The precision and accuracy for the IC-FL method Compound
Added (mg/kg)
Founda (mg/kg)
Recovery (%)
RSD (%) Within-day b
Day-to-dayc
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
0.05
0.049 0.045 0.048 0.047 0.043
± ± ± ± ±
0.003 0.003 0.004 0.004 0.002
98 90 96 94 86
6.1 6.7 8.3 8.5 4.7
8.4 7.4 9.1 9.7 7.3
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
1.0
0.97 0.85 0.85 0.81 0.83
± ± ± ± ±
0.06 0.06 0.07 0.05 0.07
97 85 85 81 83
6.2 7.1 8.2 6.2 8.4
8.2 7.4 9.4 7.2 8.5
Warfarin Coumatetralyl Bromadiolone Flocoumafen Brodifacoum
5.0
4.64 4.70 4.62 4.31 4.23
± ± ± ± ±
0.33 0.22 0.17 0.21 0.25
93 94 92 86 85
7.1 4.7 3.8 4.9 5.9
7.3 6.3 5.6 5.8 6.3
a b c
The mean value was determined in 3 days (n = 5 replicates × 3 days, x¯ ± s). n = 5 replicates × 3 days. n = 2 replicates × 8 days within a 2-week period.
M.-C. Jin et al. / J. Chromatogr. A 1155 (2007) 57–61
five rodenticides content was between 5.6% and 9.7%, as shown in Table 2. 3.4. Application of IC-FL to the real sample A spiked liver sample and a real poisoned liver sample were determined by IC with fluorescence detection. The concentrations in the spiked liver sample were 0.10, 0.010, 0.05, 0.05 and 0.05 mg/kg for warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum, respectively. The chromatograms of the two samples are shown in Fig. 3. As can be seen that the five objective compounds in the spiked sample are well separation, and the real poisoned sample contained brodifacoum of 0.015 mg/kg. 4. Conclusions A novel, sensitive, and accurate IC-FL method, with the proposed SPE procedures using an Oasis HLB cartridge has been developed for the simultaneous determination of warfarin, coumatetralyl, bromadiolone, flocoumafen and brodifacoum in animal liver tissues with recoveries above 81% for all the target compounds. The advantages of the described method are of lower amount of organic solvent and more compatible with the environment than the reported high-performance liquid chromatographic method. It has been applied to the determination of the 4-hydroxycoumarin rodenticides in animal liver tissues.
61
Acknowledgements This work was supported by a grant from the Medicine Science Foundation of Ningbo, Zhejiang Province, China (2004056) and by National Natural Science Foundation of China (Nos. 20375035, 20527005). References [1] F.Y. Guan, L. Liu, Y. Luo, Chinese J. Anal. Chem. 23 (1995) 159. [2] F.Y. Guan, I. Akira, S. Hiroshi, W. Kanako, K. Takeshi, S. Osamu, J. Pharm. Biomed. Anal. 21 (1999) 179. [3] L.J. Felice, T. Chalermchaikit, M.J. Murphy, J. Anal. Toxicol. 15 (1991) 126. [4] J.F. Lawrence, C. Thongchai, J.M. Michael, J. Anal. Toxicol. 15 (1991) 126. [5] K. Hunter, J. Chromatogr. 321 (1985) 255. [6] K. Hunter, E.A. Sharp, A. Newton, J. Chromatogr. 435 (1988) 83. [7] A.P. Eugenie, H. Jan den, F.S. Jean, A.W. Frederik, J. Anal. Toxicol. 19 (1995) 557. [8] V. Fauconnet, H. Pouliquen, L. Pinault, J. Anal. Toxicol. 21 (1997) 548. [9] C. Thongchai, J.F. Lawrence, J.M. Michanel, J. Anal. Toxicol. 17 (1993) 56. [10] F.Y. Guan, A. Ishii, H. Seno, K. Watanabe-Suzuki, T. Kumazawa, O. Suzuki, J. Chromatogr. B 731 (1999) 155. [11] M.C. Jin, X.H. Chen, Rapid Commun. Mass Spectrom. 20 (2006) 2741. [12] M.C. Jin, X.H. Chen, H.P. Chen, J. Liq. Chrom. Relat. Tech. 29 (2006) 2641. [13] T. Grobosch, B. Angelow, L. Sch¨onberg, D. Lampe, J. Anal. Toxicol. 30 (2006) 281. [14] M.C. Jin, Y.W. Yang, Anal. Chim. Acta 566 (2006) 193. [15] M.C. Jin, Y. Zhu, J. Chromatogr. A 1118 (2006) 111.