Jump to main content.


Trace Organic Analysis

Determination of Toxaphene in Soil by Electron-capture Negative Ion Mass Spectrometry after Fractionation by High-performance Gel Permeation Chromatography

William C. Brumley* and Cynthia M. Brownrigg
U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory, PO Box 93478, Las Vegas, NV 89193-3478 (USA)

Andrew H. Grange
Lockheed Engineering and Sciences Company, Environmental Program Office, 980 Kelly Johnson Drive, Las Vegas, NV 89119 (USA)

Original citation:
W. C. Brumley, C. M. Brownrigg, and A. H. Grange, "Determination of Toxaphene in Soil by Electron-capture Negative Ion Mass Spectrometry after Fractionation by High-performance Gel Permeation Chromatography," J. Chromatogr., 633, 177-183 (1993).

ABSTRACT

Toxaphene is extracted from soil by standard procedures using Soxhlet or sonication methods. The extract is fractionated by high-performance gel permeation chromatography (HPGPC), which separates toxaphene from the bulk of co-extractives including polychlorinated biphenyls. This HPGPC fractionation has broad application to many problems of environmental analysis. A solid-phase extraction cleanup with silica gel further removes any polar components present in the collected fraction. Determination of toxaphene is accomplished by electron-capture negative ion mass spectrometry (ECNI-MS) after introduction by capillary gas chromatography. Levels down to 100 µg/kg in soil are obtainable. Brief mention is made of high-resolution ECNI-MS carried out at a resolution of 10 000.

INTRODUCTION

Toxaphene is a multi-component mixture of chlorinated terpenes, primarily based on the C10 structure of bornane [1-5]. Hundreds of individual compounds make up the commercial pesticide. Although toxaphene is now limited in use, interest in its analysis continues because of its persistence and widespread transport in the environment [4, 6-10]. Current US Environmental Protection Agency (EPA) methodology (SW-846) [11] consists of electron ionization mass spectrometry (MS) and gas chromatography (GC)--electron-capture detection methods (8270 and 8080) [11], both of which are greatly affected by interferences such as polychlorinated biphenyls (PCBs), chlordane and other pesticides. Swackhamer et al. [4] published an electron capture negative ion (ECNI) method that was selective and lowered detection limits for toxaphene in fish. Their work was based, in part, on previous studies indicating the advantages of negative ion approaches [2, 3, 12]. Currently, no negative ion methods are recommended in SW-846, despite recent examinations that suggest the technique is useful [13-15] when due caution is exercised [16]. In our work, the negative ion approach is extended to soil matrices with the use of high performance gel permeation chromatography (HPGPC) [17, 18] and solid-phase extraction cartridges [19, 20]. Results obtained with high-resolution ECNI-MS [21, 22] at a resolution of 10,000 are briefly discussed in the context of resolving potentially interfering ions that contain oxygen [4] from those of the same nominal mass that come from toxaphene.

EXPERIMENTAL

Chemicals

Toxaphene, chlordane, other pesticides and PCB reference standards were obtained from the EPA Repository (Research Triangle Park, NC, USA). The following solutions from the Repository are defined: Pesticides II: aldrin, a-benzene hexachloride (BHC), b-BHC, -BHC, 4,4'-DDD, 4,4'-DDE, 4,4'-DDT, dieldrin, a-endosulfan, -endosulfan, endosulfan sulfate, endrin, endrin aldehyde, heptachlor and heptachlor epoxide: Acid Extractables II: benzoic acid, p-chloro-m-cresol, 2-chlorophenol, o-cresol, p-cresol, 2,4-dichlorophenol, 2,4-dimethylphenol, 4,6-dinitro-o-cresol, 3,4-dinitrophenol, 2-nitrophenol, 4-nitrophenol, pentachlorophenol, phenol, 2,4,5-trichlorophenol and 2,4,6-trichlorophenol; Basic Extractables: 4-chloroaniline, 2-nitroaniline, 3-nitroaniline, and 4-nitroaniline. Additional pesticide standards were purchased from Supelco (Bellefonte, PA, USA), Chem Service (West Chester, PA, USA) was the source of 2,2',3,4,4',5,6,6'-octachlorobiphenyl (OCB), [13C1]Chlordane (CHL) was obtained from Cambridge Isotope Laboratories (Woburn, MA. USA). Methylene chloride, methanol, diethyl ether, and hexane were obtained from Burdick & Jackson (Muskegon, Ml, USA).

Extraction/cleanup

Soils consisted or Nevada soil obtained locally, soil from Eagle Harbor, Puget Sound, Washington; organic potting soil purchased locally, and clay of unknown origin. The soils were weighed out and then spiked with the appropriate amount or toxaphene standard in hexane at a concentration of 50 ng/L. The spiked soil was thoroughly mixed after spiking.

Extraction was accomplished according to standard methods (3540 and 3550) [11]. Micro-Soxhlet extractions of 2-g samples were performed with a Wheaton micro-Soxhlet apparatus using methylene chloride. Sonication extraction of 20-30-g soil samples used a methylene chloride/acetone (1:1, v/v) solvent with a Heat Systems-Ultrasonic sonicator, Model XL2020 (Farmingdale, NY, USA).

Extracts were concentrated to about 0.5 mL for injection into a GPC system consisting of a guard column (50 X 7.8 mm) and two standard columns (250 x 22.5 mm) in tandem packed with Phenogel 10-m particles of 100 pore size. The system was equipped with a Valco (Houston, TX, USA) injector, Isco (Lincoln, NE, USA) UA-5 detector, Isco 260D syringe pump, and Isco Foxy 200 fraction collector. The flow was 7 mL/min [17] and fractions 29-31 (14:00-15:30 min:s) were collected for toxaphene (12-mm diameter tubes, 30 s per tube) with a retention time of about 15 min using methylene chloride as solvent.

Solid-phase extraction cleanup used Supelco 3-g silica cartridges that were pre-rinsed with 6 mL of hexane, 6 mL of hexane-diethyl ether (50:50, v/v), and 6 mL of hexane. The sample was applied in 0.5 mL of hexane followed by elution with 0.4 mL of hexane, and 4 mL of diethyl ether-hexane (5:95). The eluent was concentrated to a 1.00-0.1-mL volume as appropriate and spiked with internal standards at 0.5 mg/kg level for unknowns or at a ratio of approximately 80 pg OCB and 1.2 ng CHL to 50 ng toxaphene standard. The CHL internal standard was primarily used for retention time reference.

GC-MS

A J&W DB-5 column (Folsom, CA, USA) 30 m x 0.25 mm I.D. (0.25-m film thickness) was used with a flow-rate of 38 cm/s at 60C. The initial temperature was 60C for 3 min followed by a program rate of 20C/min to 300C. A Finnigan-MAT (San Jose, CA, USA) 4021 was operated in the negative ion mode at a source temperature setting of 170C and 0.50 mA filament emission current with an electron energy of 70 eV; methane was used as the moderator gas for electron capture (0.40 Torr source pressure reading; 1 Torr = 133.322 Pa).

The following ions were monitored: m/z 309,311, 326, 341,342, 343, 345, 377, 379, 381, 410, 411, 413, 415, 430, 444, 447 and 449, dwell time was 0.05 s/ion with a total cycle time of 1.025 s. Only the areas of the ions at m/z 341, 343, 345, 377, 379, 381, 411, 413, 415, 447 and 449 with appropriate scan ranges were summed (with an area threshold) to determine response factors and to quantify toxaphene in extracts.

The standard glass GC injector insert of the Finnigan 9610 CC system was modified by a glass-blower (Supelco) to have a restriction that passed a 26-gauge syringe needle for injection but not a 0.53 mm 1.D. capillary inserted from the oven side as a retention gap. A standard 10-L Hamilton syringe (Reno, NV, USA) fitted with 13-cm needle was then used for on-column injections with the modified insert acting as a needle guide.

High-resolution ECNI-MS

A DB-5 column 30m X 0.25-mm ID (0.25-m film thickness) was used with on-column injections using a retention gap on a Hewlett-Packard 5890A gas chromatograph. The temperature program was that of Swackhamer et al. [4]: 80C for 1 min; 10C/min to 200C; 1.5C/min to 230C; 10C/min to 250C. The mass spectrometer was a Fisons/VG 70-250SE operated in the negative ion mode with filament current of 0.200 A at an electron energy of 50eV, source temperature of 110, -8 kv accelerating voltage, and methane as moderator gas at a source housing pressure of about 1x 10-4 mbar. Fomblin (Ultrarnark 1600; PCR, Gainesville, FL) was used as calibrant. Resolutions were determined using peak widths displayed by the selected ion recording software.

RESULTS AND DISCUSSION

Quantitation Procedure

The ions monitored for toxaphene consist of the (M - Cl.)- ions resulting from C10H12Cl6 through C10H8Cl10 elemental compositions (bornanes) and overlapping contributions from (M - CL)- ions resulting from C10H10C6 through C10H6Cl10 (bornenes). We have found that ion pairs at m/z 309 and 311 and m/z 447 and 449 contribute little to the overall toxaphene response and were eliminated from consideration. In addition, the m/z 309-311 range was also subject to matrix contributions. The areas of the 11 ions given in the experimental section were integrated by automated procedures and summed to obtain a response for toxaphene. The response of the internal standard was obtained from the area of m/z 430 of OCB. An average response factor based on at least five runs of toxaphene standards spiked with OCB was used in quantitating toxaphene in soil extracts. As a control measure, the response factor was checked daily to confirm that it fell within 15% of the average response factor.

Previous work [4] included correction factors due to the presence of coextractives in quantitating toxaphene. In our work, no corrections were necessary for the presence of chlordane. Nevertheless, its presence (along with heptachlor and nonachlor) was monitored by the responses at m/z 342, 410, and 444 in order to assess potential contributions to ions indicative of toxaphene.

Responses of PCBs which are also potential interferences, were monitored at m/z 326 (pentachlorobiphenyls) and 430 (octachlorobiphenyls). PCBs constitute an interference due to the oxygen reaction producing (M - CL + O)- ions that are of the same nominal mass as ions monitored for toxaphene. We suspect that it is common for operators not to be able to completely eliminate the oxygen reaction in certain instruments. Our HPGPC cleanup, however, eliminates PCBs prior to determination of toxaphene by ECNI-MS and no correction for their presence is needed.

Results of the determination of toxaphene in soil

Table 1 provides results of analyses of unspiked (blank) soils (samples 4, 6, 11, 13, 15, 17 and 19) and spiked soils (samples 1-3, 5, 7-10, 12, 14, 16, 18, and 20) for toxaphene ranging from spiking levels of 100 g/kg (samples 3 and 16) to 10.0 mg/kg (sample 10). Samples spiked with chlordane and other pesticides, PCBs, anilines and phenols were analyzed as well as soils already contaminated with polynuclear aromatics (PNAs) (samples 3 and 16). Figures 1A and 1B give example chromatograms of m/z 377 from a standard and from a cleaned-up extract (sample 18) at a spiked level of 0.5 g/kg, respectively. Figure 1C shows an example of a toxaphene standard taken through the cleanup.

Table 1. Determination of Toxaphene in Soils by ECNI-MS

Soil origin
Sample No
Spike level toxaphene (mg/kg)
Spike level chlordane (mg/kg)
Spike level PCBs (mg/kg)a
Spike level phenols/anilines (mg/kg)b
Toxaphene level (mg/kg) ECNI-MS
Nevada
1
1.0
-
-
-
0.823
Nevada
-
1.0
-
-
-
0.862
Nevada
2
0.5
-
-
-
0.394
Nevada
-
0.5
-
-
0.396
Nevada
3
0.1
-
-
-
0.111
Nevada
-
0.1
-
-
0.137
-
-
0.1
-
-
-
0.103
Nevada
4
Blank
-
-
-
0.046
Nevada
5
0.5
1.0
-
-
0.467
Nevada
6
Blank
-
1.0
-
0.017c
Nevada
7
0.5
-
1.0
-
0.423
Nevada
8
0.5
-
1.0
-
0.360
Nevada
9
0.5
-
-
1.0
0.474
Nevada
-
0.5
-
-
-
0.549
Nevada
-
0.5
-
-
-
0.421
Nevada
10
10.0
-
-
-
1.83d
Eagle Harbor
11
Blank
-
-
-
0.032c
Eagle Harbor
12
0.125
-
-
-
0.083
Clay
13
Blank
-
-
-
0.069
Clay
14
0.5
-
-
-
0.703
Organic
15
Blank
-
-
-
0.025c
Organic
16
0.1
-
-
-
0.082
CH2Cl2
17
Blank
-
-
-
0.021c
CH2Cl2
18
0.5
-
-
-
0.459
Nevada
19
Blank
-
-
-
0.016c
Nevada
20
0.5
1.0
1.0
1.0e
0.450

a PCB 1242, 1248, and 1254 each 1 mg/kg.
b 1 ppm of Basic Extractables and Acid Extractables II.
c Not confirmable as toxaphene.
d Weathered 6 months.
e 1 ppm of Pesticides II.

Image of chromatograph - contact brumley.william@epa.gov for further information

Image of chromatograph - contact brumley.william@epa.gov for further information

Image of chromatograph - contact brumley.william@epa.gov for further information

Figure 1. Ion chromatograms of m/z 377 of
(a) a standard of toxaphene;
(b) sample 18, 0.5 mg/kg toxaphene taken through the cleanup; and
(c) a standard of toxaphene taken through the cleanup.

Chromatographic patterns and relative ion abundances remain largely consistent between standards and extracts. There are, however, some changes in the pattern caused by the cleanup.

For spiked samples, one could assess comparisons of ion chromatographic patterns using a standard taken through the cleanup. For real samples, however, weathering [23] could alter relative contributions of specific components of toxaphene that could vary on a case-by-case basis. This possibility and the results discussed later led us to rely solely on quantitations obtained using the toxaphene standard directly. We found, for example, that for a soil spiked at 10 mg/kg (sample 10) and weathered for 6 months, only slight changes in ion chromatographic patterns were evident compared with standards.

The detection limit for toxaphene is actually lower than 100 g/kg as evidenced, for example, by results for sample 16 (80 g/kg). Quantitative results for blanks reflect reagent and matrix contributions integrated by the automated procedure and do not indicate the presence toxaphene-like compounds.

A reagent blank (sample 17) taken through the cleanup gave a result of 0.021 mg/kg (not confirmable as toxaphene), so that laboratory contamination involving toxaphene is not at issue. The higher-level blanks (samples 4 and 13) were a result of carryover from calibration of the HPGPC with toxaphene standards. Once proper precautions were taken in rinsing the injector a background level of about 0.02 mg/kg resulted. This response was not confirmable as toxaphene. Hence, a 0.100 mg/kg limit seems reasonable in view or apparent unidentified background contributions at this level of cleanup.

The effects of PCBs chlordane and other pesticides are considered in the spiking studies reported in Table 1 (samples 5-9 and 20). In general, quantitation of 0.5 mg/kg levels are unaffected by these compounds. This substantiates the improvement of this approach over existing methodology [11]. As an indication of the presence of PCBs, ions monitored at m/z 326 and 430 correspond to molecular anions of pentachlorobiphenyls and octachlorobiphenyls and serve as a check on the efficiency of the HPGPC cleanup. Without the HPGPC cleanup, a 1.0 mg/kg spike of PCBs would result in a "toxaphene-like" response quantitated as 3-6 mg/kg toxaphene in our instrument.

Because PCBs create an interference when oxygen is present in the ion source, the question arises as to the contribution of the internal standard OCB in producing a toxaphene-like response due to the oxygen reaction. No corrections were found necessary at the levels studied and the retention time of this compound.

A blank soil spiked at 1.0 mg/kg PCBs (sample 6) establishes the efficiency of the HPGPC cleanup in removing PCBs (compare 0.017 mg/kg to other blank levels in samples 4, 11, 15, and 19). The 1.0 mg/kg spikes of PCBs and chlordane and other pesticides (e.g. sample 20) do not generate interferences either because they are not collected in the fraction with toxaphene or because they do not afford ions that can interfere. Chlordane is collected in the fraction with toxaphene but does not interfere at levels comparable to those of toxaphene5 as evidenced by the results in Table 1 (samples 5 and 20). By monitoring chlordane response at m/z 342 one can assess whether the chlordane level is of such magnitude as to require any or the corrections suggested by Swackhamer et al. [4].

Selective fractionation by HPGPC has broad implications for cleanup of environmental samples. Figure 2 illustrates the chromatography and separations obtained with HPGPC for standards (Figure 2A) and for a typical soil extract (Figure 2B).

Image of chromatograph - contact brumley.william@epa.gov for further information

Image of chromatograph - contact brumley.william@epa.gov for further information

Figure 2. HPGPC separations.
(a)
Standards of (1) bis(2-ethylhexyl)phthalate, retention time 12 min 20 s; (2) toxaphene, retention time 14 min 50 s; (Arochlor, retention time 15 min 40 s; and (4) perylene, retention time 17 min 30s.
(b)
Soil extract with (1) main matrix components centered at retention time 11 min 30 s, and (2) sulfur at retention time 21.0 min.

HPGPC unlike most adsorption chromatography separates toxaphene, a multi-component mixture as one peak rather than separated into multi-components over a broad retention range. Since HPGPC is operated in an automatic fractionation mode, other analytes such as PCBs are also collected in individual fractions as a first cleanup step. This approach seems to be an obvious advantage in PCB analysis since the usual subsequent cleanup steps are spent isolating PCBs from interferences. Other target analytes found in the window from bis(2-ethylhexyl)phthalate to beyond perylene but before sulfur can be collected separately or taken as a whole by combining fractions if desired [17, 18].

Finally, the solid-phase extraction cleanup step with silica removes polar components (sample 9) that might be present. The main benefit is an extract free from relatively non-volatile components that could cause problems in the injection port or retention gap.

The precision of determination for a given sample extract is about 10% as seen, for example, at the 0.5 mg/kg level with sample 9. Precision of recovery (reproducibility) at the same spiking level is also about 10% (samples 2, 5, and 7-9). The relative response factor for toxaphene versus the internal standard (OCR) exhibited a 13% R.S.D. over a 3-month period. Linearity of the relative response of toxaphene to the internal standard was demonstrated for levels of toxaphene from 0.1 to 1.0 mg/kg with OCB levels at 0.5 mg/kg. The recovery of toxaphene from spiked samples is summarized as follows from the data in Table 1. At the 0.5 mg/kg level for spiked soils for which extensive data are available, the average recovery was 85% 10% (7 determinations). At the 1.0 and 0.1 mg/kg levels the recoveries were similar but are based on fewer determinations.

High-resolution ECNI-MS

Although a variety of soils were analyzed in developing the ECNI-MS methodology, potential exists in environmental analysis for unexpected difficulties or interferences. A high-resolution ECNIMS determination could present an alternative or tiered approach that would give the analyst a more selective determinative technique without having to resort immediately to further sample cleanup.

In the early stages of this work, high-resolution ECNI-MS was examined to explore its ability to eliminate interferences caused by the reaction of oxygen with PCBs and the potential interference posed by polychlorinated diphenyl ethers (PCDPEs) [4]. High-resolution ECNI-MS at 10 000 resolution can eliminate responses from oxygen-containing ions that are isobaric in nominal mass with toxaphene ions regardless of whether they arise from PCDPEs or from oxygen reactions with PCBs (e.g., m/z 342.8646 for C12H4OCl337Cl2 and 342.89062 for C10H11C1537Cl).

Potential also exists for the presence of m/z 342.8776 from C10H9Cl437Cl in toxaphene, which would require a resolution of about 26 000 to resolve from the interfering ion. Studies of ratios of toxaphene standards to PCBs of 1, 1:10, 1:100 and 1:1000 at resolutions of 1000, 5000, 10 000 and 15 000 were made. Results indicated that a resolution of 10 000 was sufficient to give reliable quantitations in the presence of PCBs.

Despite these apparent advantages, high-resolution ECNI-MS has not been widely adopted, even though reports of its applications [21, 22] and calibration procedures [22] have appeared. Some additional effort by instrument manufacturers may be necessary to deal with arcing and beam instability problems.

NOTICE

Although the research described in this article has been funded wholly or in part by the US Environmental Protection Agency through contract No. 68-C0-0049, it has not been subjected to Agency review. Therefore, it does not necessarily reflect the views of the Agency, and no official endorsement should be inferred. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

REFERENCES

  1. R. L. Holmstead, S. Khalifa and J. E. Casida, J. Agric. Food Chem., 22 (1974) 939.
  2. M. A. Saleh, J. Agric. Food Chem., 31 (1983) 748.
  3. B. Jansson and V. Widvqist, J. Environ. Anal. Chem., 13 (1983) 309.
  4. D. L. Swackhamer, M. J. Charles and R. A. Hites, Anal. Chem., 59 (1987) 913.
  5. M. A. Saleh and J. E. Casida, J. Agric. Food Chem., 26 (l978) 583.
  6. C P. Fuerst, C. Fuerst and W. Groebel, Dtsch. Lebensm.-Raundsch., 85 (1989) 273.
  7. F. Becker, G. Lach and H. Parlar, Toxicol. Environ. Chem., 20 (1989) 203.
  8. G. Lach and H. Parlar, Toxicol. Environ. Chem., 31-32 (1991) 209.
  9. H. Parlar, F. Becker, R. Mueller and G. Lach, Fresenius' Z. Anal. Chem., 331 (1988) 804.
  10. F. I. Onuska and K. A. Terry, J. Chromatogr., 471 (1989) 161.
  11. Test Methods for Evaluating Solid Waste (SW-846), Vol. B, US Environmental Protection Agency, Washington, DC., 3rd ed., November 1986.
  12. M. A. Ribick, G. R. Dubay, J. D. Petty, D. L. Stalling and C. J. Schmit, Environ. Sci. Technol., 16 (1982) 310.
  13. M. Oehme, D. Stöckl and H. Knöoppel, Anal. Chem., 58 (1986) 554.
  14. E. A. Stemmler and R. A. Hites, Biomed. Environ. Mass Spectrom., 15 (l989) 659.
  15. E. A. Stemmler, R. A. Hites, B. Arbogast, W. L. Budde, M. L. Deinzer, R. C. Dougherty, J. W. Eichelberger, R. L. Fultz, C. Brimm, E. P. Grimsrud, C. Sakashita and L. J. Sears, Anal. Chem., 60 (1988) 781.
  16. S. Erhardt-Zabik, J. T. Watson and M. J. Zabik, Biomed. Environ. Mass Spectrom., 19 (1990) 101.
  17. M. M. Krahn, C. A. Wigren, R. W. Pearce, L. K. Moore, R. G. Bogar, W. D. Macteod, Jr., S.-L. Chan and D. W. Brown, NOAA Technical Memorandum NMFS FINWC-153, US Department of Commerce, October 1988.
  18. G. J. Fallick, R. Cotter, R. Foster and R. L. Wellman, Proceedings of the 7th Annual Waste Testing and Quality Assurance Symposium, Washington DC, July 8-12, 1991, p. II-232.
  19. V. Lopez-Avila, J. Milanes N. S. Dodhiwala, and W. F. Beckert, J. Chromatogr. SCI, 27 (1989) 209.
  20. P. R. Loconto, LC.GC, 9 (1991) 460.
  21. A. H. Grange and W. C. Brumley, Proceedings of the 39th ASMS Conference on Mass Spectrometry and Allied Topics, Nashville, TN, May 19-24, 1991.
  22. G. M. Brilis and W. C. Brumley, Anal. Chim. Acta, 229 (1990) 163.
  23. J. N. Seiber, S. C. Madden, M. M. McChesney and W. L. Winterlin, J. Agric. Food Chem., 27 (1979) 284.

Trace Organic Analysis Home Page
Analytical Environmental Chemistry
[ Environmental Sciences ]
[ Office of Research Development]
[ National Exposure Research Laboratory]
Author: William C. Brumley


Local Navigation


Jump to main content.