ANALYST, FEBRUARY 1986, VOL. 111 145 Gas Chromatographic Determination of Triclopyr in Environmental Waters Tadashi Tsukioka Nagano Research Institute for Health and Pollution, 1978, Komemura, Amori, Nagano-shi, Nagano, Japan Ryuzo Takeshita School of Pharmaceutical Science, Toho University, 22 I , Miyama, Funabashi-shi, Chiba, Japan and Tetsuro Murakami Department of Chemical Engineering, Kogakuin University, 1-24-2, Nish ish inju ku, Shinjuku-ku, Tokyo, Japan The reaction of BF3-trifluoroethanol with an extract of triclopyr from an acidified sample solution to form the trifluoroethyl ester has been applied to the determination of triclopyr in environmental waters. The product is cleaned up by silica-gel column chromatography and determined by gas chromatography with electron- capture detection.The detection and determination limits were 0.005 ng and 0.00025 pg ml-1, respectively. The recovery and coefficient of variation were found to be 90-93% and less than 4%, respectively ( n = 71, for recovery experiments on river waters. Keywords: Triclopyr determination; gas chromatography; herbicide residue; river water; halogenated alk yla tion Triclopyr (3,5,6-trichloro-2-pyridyloxyacetic acid) is a hor- mone-type herbicide that is effective for the destruction of arrowroots (Pueraria thunbergianal) and deciduous shrubs. It is on the market as the triethylammonium salt (triclopyr- TEA) or butoxyethyl ester (triclopyr-BE). This herbicide, singly or mixed with Frenock (sodium 2,2,3,3- tetrafluoropropionate) , is used extensively in woods and forests.2 It is therefore necessary to determine the pollution of natural waters by this herbicide, for which purpose a simple and highly precise microanalytical method is required.A microanalytical method for the determination of triclopyr has not been reported, although there are many reports of the determination of phenoxy herbicides,sl6 which have similar properties to triclopyr. Phenoxy herbicides have a high polarity and low volatility, preventing the use of a direct GC method for their determination. Thus, they are subjected to GC or GC - MS after conversion into more volatile com- pounds, i.e. , alkyl esters,sg halogenated alkyl esters10Jl or halogenated aromatic esters. 12-16 In the environment triclopyr- BE is hydrolysed gradually into triclopyr-BE is more easily total amount of triclopyr and triclopyr-BE is more easily determined than are the two compounds separately.A method of determination has been developed in which triclopyr and triclopyr-BE in acidic, aqueous solution are extracted with diethyl ether, converted into the trifluoroethyl (TFE) ester, cleaned up by silica-gel column chromatography and determined by gas chromatography with electron-capture detection. (ECD - GC). This method has sufficient sensitivity, precision and manageability to be applicable to environmental waters. Experimental Reagents 3,5,6-Trichloro-2-pyridyloxyacetic acid (triclopyr) , triethyl- ammonium 3,5,6-trichloro-2-pyridyloxyacetate (triclopyr- TEA) and butoxyethyl 3,5,6-trichloro-2-pyridyloxyacetate (triclopyr-BE) were obtained from Dow Chemical Japan.Standard solutions of each of these compounds were prepared from a concentrated solution of 1000 pg ml-1 in acetone by diluting with acetone to 10, 1 or 0.1 yg ml-1. Polychlorinated biphenyls (PCB) , dibutyl phthalate, 2,4-dichloro- phenoxyacetic acid (2,4-D) , 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) and 2-methyl-4-chlorophenoxyacetic acid (MCP) were obtained from Wako Pure Chemical Co., Japan. Silica gel, Wako-gel S-1 from Wako Pure Chemical Co. , Japan, was activated by heating for 12 h at 130 "C before use. A solution of 15% mlV boron trifluoride in 2,2,2-trifluoroethanol (BF3- TFE) was obtained from Tokyo Kasei Co., Japan. Hexane, benzene, acetone and diethyl ether were of the grade suitable for detection of pesticide residues.All the other reagents were of guaranteed grade. Apparatus The gas chromatograph was a Shimadzu Model GC-3BE equipped with an electron-capture detector (63Ni) , the Reacti- Therm was from Pierce Chemical Co., USA, the gas chromatograph - mass spectrometer was a Model JMS-D300 from Japan Electron Optics Laboratory Co., Japan, and the chromatographic column was 10 mm in diameter and 300 mm in length. GC Conditions The stationary phase was 5% XE-60 on Chromosorb W (60-80 mesh) , packed into a glass column (3 mm in diameter and 200 cm in length). The carrier gas was nitrogen with a flow-rate of 28 ml min-1. The temperature was 155 "C for both the column and the detector and 200 "C for the injection port.146 > .- c ANALYST, FEBRUARY 1986, VOL. 111 I I t a c 0 a a a L I 1 5 70 Tim e/m i n Fig.1. Gas chromatogram of triclopyr-TFE. Column, 2 m, 5% XE-60; column and detector temperature, 155 "C; injection-port temperature, 200 "C; and carrier gas, N2 at a flow-rate of 28 ml min-* looo t 210 11 - 1000 c t M' 337 0 250 300 350 400 m/z Fig. 2. EI mass spectrum of triclopyr-WE. Column, 2 m, 5% XE-60; column t e y erature, 160 "C; injection-port and enricher temperatures, 200 8; ion-source temperature, 250 "C; ionisation voltage, 70 eV; and carrier gas, He at a flow-rate of 40 ml min-1 Table 1. Effects of reaction temperature and reaction time on the esterification of triclopyr Esterification at reaction temperature, % Reaction time/ min 50°C 60°C 70°C 80°C 90°C ' 10 44 63 90 100 100 20 70 82 100 100 100 40 94 100 100 100 100 60 100 100 100 100 100 80 100 100 100 100 100 Standard Procedure The standard procedure consists of four steps: extraction, esterification, clean-up and determination.Table 2. Effects of reaction temperature and reaction time on the ester-group exchange reaction Exchange at reaction temperature, "/o Reaction time/ min 50°C 60°C 70°C 80°C 90°C 10 22 46 61 76 93 20 51 68 87 95 99 100 100 40 73 94 99 100 100 60 85 100 100 100 100 80 91 100 100 q , , g 75 .- 0.5 1 .o .+I - a Volume of reagent/ml LK Fig. 3. Effect of amount of reagent on the esterification 8 2 100 2{ s,' a, 75 Fig. 4. Effect of amount of reagent on the exchange reaction of ester groups Extraction A 200-ml portion of the sample water is placed in a 300-ml separating funnel, to which 6 g of NaCl and 1 ml of 9 M H2SO4 are added and shaken to 10 min with each of two 50-ml portions of diethyl ether.The combined diethyl ether extracts are washed with 20 ml of 10% mlV NaCl solution, dried with anhydrous Na2S04 and concentrated to less than 5 ml in a Kuderna - Danish (KD) concentrator. Esterification The concentrate is transferred into a 5-ml vial and the solvent is removed with a gentle stream of Nz. A 0.25-ml portion of BF3-TFE is added to this vial, which is covered by a Teflon cap and placed on a Reacti-Therm at 80 "C and allowed to react for 1 h. After cooling, the contents are transferred into a 100-ml separating funnel with 30 ml of hexane, washed twice with 20 ml of 10% mlV NaCl solution, dried with anhydrous Na2S04 and concentrated to less than 5 ml in the KD concentrator.Clean-up The concentrate is transferred on to a 3-g silica-gel column (10 mm i.d.) that has been slurry-packed in hexane. The column is washed with 100 ml of benzene - hexane (10 + 90), and the adsorbed compound is eluted with 100 ml of benzene - hexane (35 + 65). Determination The eluate is concentrated in the KD concentrator to less than 5 ml, diluted to the appropriate volume and subjected to ECD - GC.ANALYST, FEBRUARY 1986, VOL. 111 50 23 d c 3 -Z 25 r 3 E" a 147 - - H2S04 concentrationh Fig. 5. Effect of acid concentration on the recovery I I 1 I I I 0 50 Volume of solvent/mI t a, t 0 Q a n Fig. 6. Elution pattern of triclopyr-TFE 0 10 Timeim in 20 Fig. 7. Gas chromatogram of river water before clean-up. Condi- tions as in Fig.1 A blank test is conducted by the same procedure with 200 ml of distilled water. Results and Discussion Formation and Identification of Triclopyr-TFE The followingexperiment was conducted in order to prevent tailing during gas chromatography and to produce ECD - GC analysis of high sensitivity. A 1-mg mass of triclopyr and 0.5 ml of BF3-TFE were placed in a vial, heated at 80 "C for 1 h, and extracted with hexane. To select the column conditions for the GC of the reaction product, a test was conducted in the range 150-200 "C on 2% OV-17, 2% OV-101, 5% SE-52, 5% DEGS and 5% XE-60. Fig. 1 shows the chromatogram obtained with XE-60, which was the best with respect to the peak shape and separation from coexisting substances. A mass spectral measurement gave a peak at mlz = 337 corresponding to the molecular ion (M+) of triclopyr-TFE (Fig.2). Investigation of the Conditions for TFE Esterification Esterification To ascertain the optimum reaction temperature and time for the esterification, 2.5 ml (containing 2.5 pg) of triclopyr standard solution were placed in a 5-ml vial and the solvent was removed in a gentle stream of N2. A 0.25-ml portion of BF3-TFE was added to this vial, which was covered by a Teflon cap. Esterification was conducted at 50-90 "C for 10-80 min. As Table 1 shows, the ester was obtained quantitatively under the following reaction conditions: 60 min at 50 "C, 40 rnin at 60 "C, 20 min at 70 "C and 10 rnin at 80 or 90 "C. The amount of triclopyr, the temperature and the reaction time were kept constant at 20 pg, 60 "C and 60 min, respectively, and the amount of the reagent was changed in the range 0.05-1 ml.Fig. 3 shows that the esterification is constant with a minimum amount of 0.1 ml of BF3-TFE, but 0.25 ml of BF3-TFE was used for actual samples in the event that they contain other substances that consume BF3-TFE. Ester-group exchange reaction The effects of reaction temperature, reaction time and exchange ratio on the ester-group exchange reaction were investigated with 2.5 ml (containing 2.5 pg) of triclopyr-BE standard solution in a vial, under the same conditions as the esterification. Table 2 shows that the ester-group exchange reaction does not occur as easily as the esterification; even an 80-min reaction time at 50 "C did not yield a 100% exchange.Quantitative exchange requires 60 rnin at 60 "C, ca. 40 rnin at 70 "C, 40 min at 80 "C and 20 min at 90 "C. The amount of reagent required for a quantitative reaction was found by using 20 pg of triclopyr-BE under the same conditions as the esterification. Although 0.1 ml of reagent was sufficient, 0.25 ml was used for the same reason as in the esterification (Fig. 4). Hence the optimum conditions for esterification are as follows: addition of 0.25 ml of BF3-TFE and reaction for 1 h at 80 "C. Investigation of the Extraction Conditions Extraction of triclopyr Diethyl ether was selected as the extraction solvent because it is easily removed after extraction. The optimum acid concentration for extraction was deter- mined in the following manner. A 2.5-ml volume of triclopyr- TEA standard solution (containing 2.5 pg) and different amounts of 9 M H2S04 were added to 100 ml of distilled water to produce the final solution with H2S04 concentrations of 0-0.5 M.Each was extracted with 50 ml of diethyl ether and then submitted to the standard analysis procedure. The relationship between the acid concentration and the recovery148 ANALYST, FEBRUARY 1986, VOL. 111 L 0 5 10 Timeimin Fig. 8. Gas chromatogram of river water after clean-up. Conditions as in Fig. 1 0 4 1 I 10 Time/m in 20 Fig. 9. Gas chromatogram of TFE esters of acid herbicides. Conditions as in Fig. 1. Peaks: 1, MCP, 2 ng; 2, triclopyr, 0.2 ng; 3, W D , 0.4 ng; 4, 2,4,5-T, 0.3 ng t a rn C 0 P a a 0 5 10 Time/min Fig. 10. Gas chromatogram of a river water. Conditions as in Fig.1 was found and the results are shown in Fig. 5. The recovery of triclopyr-TEA was ca. 55% with no acid added and ca. 100% for an acid concentration above 0.025 M. To determine the optimum salt concentration for ex- traction, the relationship between the salt concentration and the extraction ratio was investigated by changing the NaCl concentration in the range &30% mlV, keeping the acid concentration at 0.05 M. The recovery did not change. As environmental samples sometimes contain suspended sub- stances that form emulsions which prolong the time required for separation, the addition of up to 3% NaCl solution was adopted to prevent such an emulsion formation. Extraction of triclopyr-BE Triclopyr-BE can be extracted by ordinary hydrophobic solvents and hence was extracted with diethyl ether, as was triclopyr-TEA.The dependence on the recovery with acid and salt concentrations was similarly investigated and was found to be independent of both acid and salt concentration over the range examined. Hence 6 g of NaCl and 1 ml of 9 M H2S04 were added to 200 ml of sample before extraction. Clean-up by Silica-gel Column Chromatography A variety of substances in natural waters can be simul- taneously extracted and esterified by TFE and therefore can cause interference in the determination of triclopyr-TFE. Hence the esterification products were purified by silica-gel column chromatography. Triclopyr-TFE (10 pg) was loaded on to a column (10 mm i.d.) containing 3 g of silica gel and eluted with 100 ml of benzene - hexane (10 + 90); no ester species were found in the eluate.The ester was completely eluted with a mixture (100 ml) of benzene - hexane (35 + 65) (Fig. 6).ANALYST, FEBRUARY 1986, VOL. 111 149 Table 3. Results of the addition and recovery experiments for triclopyr-TEA and triclopyr-BE Coefficient of Amount variation Sample Compound added/yg Recovery, YO (n = 7), Yo Distilled water . . . . Triclopyr-TEA 0.5 92 3.1 Triclop yr-BE 0.5 93 2.7 Riverwater . . . . . . Triclopyr-TEA 0.5 90 3.5 Triclopyr-BE 0.5 92 3.4 The clean-up procedure was as follows: the column was washed with 100 ml of benzene - hexane (10 + 90) and the ester was eluted with 100 ml of benzene - hexane (35 + 65). The usefulness of this procedure was demonstrated with 0.5 ml (containing 0.5 pg) of triclopyr standard solution in 200 ml of an actual river water (Figs.7 and 8). Effect of Interfering Substances The influence of phthalate esters, PCB, HCB, MCP, 2,4-D and 2,4,5-T, which can be extracted with diethyl ether, was examined by taking 5 pg of each substance in distilled water. The phthalate esters were extracted and esterified with TFE but eluted slightly from the silica-gel column. Their GC retention time (ca. 2.5 min) was considerably different from that of triclopyr-TFE. PCB and HCB were extracted but eluted by the washing solvent from the silica-gel column. MCP, 2,4-D and 2,4,5-T were extracted, esterified with TFE and eluted from the silica-gel column. However, they caused no interference as they have different GC retention times (Fig. 9).Calibration Graphs and Recovery Experiments A calibration graph was prepared as follows: standard solutions containing 0.2, 0.4, 0.6, 0.8 and 1.0 pg of triclopyr were placed in 5-ml vials, the solvent was removed in a gentle stream of N2, the residue was esterified with TFE and the amount of the ester measured by ECD - GC. The calibration graph was linear over the range examined. The detection limit was 0.005 ng with a 5-pl sample injection and the determina- tion limit was 0.00025 pg ml-1 using 200 ml of sample solution. The recovery was examined by the standard procedure after adding a given amount of triclopyr to 200 ml of distilled water or river water without triclopyr. Table 3 shows that the proposed method is satisfactory, with recoveries of more than 90% and a coefficient of variation of less than 4%.Application to Real Samples The effectiveness of the proposed procedure was tested by analysing actual samples of river water (n = 12). It was found that 0.25 ml of BF3-TFE was sufficient, its use in excess causing no interference, and that the coexistence of pollutants or natural organics did not interfere with the determination. A replicate experiment on a river water containing 0.001 pg ml-1 of triclopyr gave a standard deviation of less than 4% (n = 5 ) . Fig. 10 shows the detection of 0.0015 pg ml-1 of triclopyr in a river water sampled near a site where triclopyr had been released 1 week before. Conclusion Triclopyr can be successfully extracted from acidified aqueous samples with diethyl ether and esterified with BF3-TFE.Interfering substances can be eliminated by a clean-up method with a silica-gel column and using benzene - hexane as an eluting solvent. The recovery from actual river waters is 9&93%, with coefficients of variation of less than 4%. The detection and determination limits are 0.005 ng and 0.00025 pg ml-1, respectively. The proposed procedure is also useful for the simultaneous determination of MCP, 2,4-D and 2,4,5-T if the clean-up method is modified. The authors thank Dow Chemical Japan for the gifts of triclopyr, triclopyr-BE and triclopyr-TEA and Professor I. Matsuzaki and Messrs. S. Shimizu and H. Ozawa for many helpful suggestions. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. References Makino, T., “Makino’s Illustrated Flora in Color,” Hokuyokan, Tokyo, 1982, p. 263. Noyaku Yoran Henshu Iinkai, “Noyaku Yoran,” Nippon Shokubutsu Boeki Kyokai, Tokyo, 1983, p. 159. Howard, S. F., and Yip, G., J. Assoc. Off. Anal. Chem., 1971, 51, 970. Khan, S . U., J. Assoc. Off. Anal. Chem., 1975, 58, 1027. Thio, A. P., Kornet, M. J., Tan, H. S. I., and Tompkins, D. H.,Anal. Lett., 1979, 12, 1009. Cotterill, E. G., Analyst, 1982, 107, 76. Goetz, R., Fresenius 2. Anal. Chem., 1983, 314, 131. Sekita, H., Takeda, M., Saito, Y., and Uchiyama, M., Eisei Kagaku, 1982, 28, 219. Chmil, V. D., Zh. Anal. Khim., 1981,36, 1121. Gutermann, W. H., and Lisk, D. J., J. Assoc. Off. Anal. Chem., 1977, 60, 1070. Woodham, D. W., Mitchell, W. G., Loftis, C. D., and Collier, C. W., J. Agric. Food Chem., 1971, 19, 186. Agemian, H., and Chau, A. S . Y., Analyst, 1976, 101, 732. Johnson, L. G., J. Assoc. Off. Anal. Chem., 1973, 56, 1503. Cotterill, E. G., J. Chromatogr., 1979, 171, 478. Draper, W. M., J. Agric. Food Chem., 1982, 30,227. Lee, H. B., and Chau, A. S . Y., J. Assoc. Off. Anal. Chem., 1983, 66, 1023. Paper A51259 Received July 16th, 1985 Accepted August 2nd, 1985