Determination of Chemical Warfare Agent Degradation Products at

Degradation Products at Low-Part-per-Billion. Levels in Aqueous Samples and. Sub-Part-per-Million Levels in Soils Using Capillary. Electrophoresis. Al...
0 downloads 0 Views 99KB Size
Anal. Chem. 1999, 71, 1285-1292

Determination of Chemical Warfare Agent Degradation Products at Low-Part-per-Billion Levels in Aqueous Samples and Sub-Part-per-Million Levels in Soils Using Capillary Electrophoresis Alaa-Eldin F. Nassar* and Samuel V. Lucas

Battelle Memorial Institute, 2012 Tollgate Road, Bel Air, Maryland 21015 Lynn D. Hoffland

U.S. Army Material Command Treaty Laboratory, Aberdeen Proving Ground, Maryland 21010

A significant enhancement in the method detection limits is observed in the analysis of chemical warfare agent (CWA) degradation products in environmental samples by capillary electrophoresis (CE) using electrokinetic injection. The CE method uses indirect UV detection of the nonderivitized acidic analyte and a cationic surfactant, didodecyldimethylammonium hydroxide, for reversal of the electroosmotic flow. Analytes studied include the dibasic acid methylphosphonic acid (MPA) and its monoacid/monoalkyl esters, RMPA, where R ) ethyl, isopropyl, and pinacolyl (2-(3,3-dimethylbutyl)). The CE method uses an attractive buffer system which is highly stable and inexpensive, and, in addition to reversing the electroosmotic flow, provides excellent separation efficiencies within a 3-min run. This CE method is also free from interference caused by carbonate, humic acids, and fluoride. Compared to pressure injection, electrokinetic injection with this CE buffer system provided substantially lower detection limits, up to 100-fold lower for samples in reagent water. However, to best realize the benefits of the electrokinetic injection enhancement for environmental samples, a prior cleanup of the sample using standard ion-exchange cartridges is necessary. This cleanup step uses sequential cartridges to remove sulfate (barium cartridge), chloride (silver cartridge), and cations (H+ cartridge). Using this approach, detection limits for these four acids were as low as 1-2 µg/L for water samples and 25-50 µg/L for aqueous leachates of soil samples (10 mL of leachate/1.5 g of soil). The utility of this method for separation of CWA degradation products by CE is discussed in terms of pressure injection versus electokinetic injection. The effects of voltage and time of injection on the separation were investigated. Results from three types of soils and four types of water (groundwater, artificial seawater, tap water, bay water) indicated that the method has potential for environmental monitoring. Quantitative CE analysis with electrokinetic injection enhance10.1021/ac980886d CCC: $18.00 Published on Web 02/25/1999

© 1999 American Chemical Society

ment of detection limits of these types of environmental samples requires the use of an appropriate internal standard approach. The data presented here indicate that an internal standard-based approach could be expected to give analysis results in the sub-part-per-million concentration range of 90-110% of the true value. Capillary electrophoresis (CE) was first introduced by Jorgenson, and it has proven to be a powerful separation technique that is capable of rapid and efficient separation of many compounds within complex matrixes.1 Because CE can be used successfully with small samples, it is useful for the investigation of both chemical and biological processes. The exceptional separation efficiencies of CE place a major emphasis on the injection, detection, and other processes used in analyzing samples. Samples can be injected only in limited volumes, which leads to difficulties with low concentration sensitivity. Several articles have reported efforts to address this concern.2-8 Isotachophoresis allows preconcentration of samples into discrete, tightly focused bands which can improve the method detection limits of CE.7,8 This phenomenon occurs when there is a difference in conductivity between the buffer solution and the sample matrix. Electrokinetic injection is a related technique which has shown a great deal of promise (1) For example see: (a) Jorgenson, J. W.; Lukacs, K. D. Anal. Chem. 1981, 53, 1298-1302. (b) Jorgenson, J. W.; Lukacs, K. D. Science 1983, 222, 181-189. (c) Jorgenson, J. W. ACS Symp. Ser. 1987, No. 335, 182-198. (d) Walbroehl, Y.; Jorgenson, J. W. J. Microcolumn Sep. 1989, 1, 41. (e) Kennedy, R. T.; Oates, M. D.; Cooper, B. R.; Nickerson, B.; Jorgenson, J. W. Science 1989, 246, 57-63. (2) Reinhoud, N. J.; Tjaden, U. R.; Greef, J. van der. J. Chromatogr., A 1993, 653, 303-312. (3) Wolf, S. M.; Vouros, P. Anal. Chem. 1995, 67, 891-900. (4) Thompson, T. J.; Foret, F.; Vouros, P.; Karger, B. Anal. Chem. 1993, 65, 900-906. (5) Sjorgen, A.; Dasgupta, P. K. Anal. Chem. 1996, 68, 1933-1940. (6) Dasgupta, P. K.; Surowiec, K. Anal. Chem. 1996, 68, 4291-4299. (7) Landers, J. P., Ed. Handbook of Capillary Electrophoresis, 2nd ed., CRC Press: Boca Raton, FL, 1996. (8) Righetti, P. G., Ed. Capillary Electrophoresis in Analytical Biotechnology, 1st ed.; CRC Press: Boca Raton, FL, 1996.

Analytical Chemistry, Vol. 71, No. 7, April 1, 1999 1285

in electrophoretic analysis because it does not require extra equipment nor does it not add to analysis times.7,8 One example of the value of CE involves chemical warfare (CW) agent degradation products. CW agents have been of concern to governments, military leaders, and scientists since their introduction.9 As these highly toxic agents have evolved, so too have efforts to identify, quantify, and neutralize them as well as identify and quantify their degradation products in a wide variety of matrixes. In emergency situations such as an explosion or a chemical warfare attack, the mobile laboratory would be the first line of detection and analysis. Because space, weight, and waste storage are of primary concern for mobile laboratories, CE offers the advantages of compact, lightweight equipment, smaller samples, and reduced waste. Sensitivity, speed, and high resolution are valuable assets for analysis of CW agent degradation products in agent destruction trails. In addition, the Chemical Warfare Convention (CWC) has created a demand for verifiable compliance measurements, which only adds to the need for reliable, rapid, and easily transportable methods for the determination of chemical warfare agents and their degradation products. In electrokinetic injection, analytes enter the capillary by migration under the influence of the applied electric field. The moles of each type of ion taken into the capillary in t seconds can be described by eq 1, developed by Rose and Jorgenson,10 where

detection for the determination of a series of organophosphoric and organophosphonic acids in environmental samples.22,23 Also, ion chromatography (IC) can be used for the analysis of alkylphosphonic acids and their monoesters.24-27 The IC separation requires gradient elution, which adds substantially to the analysis time, makes retention times less reproducible, produces more waste solvent, and requires more sophisticated instrumentation. CE has also been used as an alternative approach, as it offers short analysis times, little sample preparation other than dilution and filtration, and flexibility in formulating electrolytes to minimize matrix interference.28-30 Also, because the required daily solvent/ buffer usage is measured in milliliters as compared to liters with HPLC and IC, the portability of instrumentation dramatically improves, as well as allowing analysis of much smaller samples. Recently,31-33 we reported on CE methods as alternatives to IC in the analysis of anionic and cationic chemical warfare agent degradation products in environmental samples and byproducts from the chemical degradation of CW agents. Analysis time is rapid with the analytes of interest obtained within 3 min (anions) or 6 min (cations) from the time of injection. Analyte detection limits are in the high-microgram per liter range using pressure injections, and separations are robust with migration time variations less than 5.0% RSD on a day-to-day basis. The methods are versatile, permitting determination of G- or V-type nerve agent

moles injected ) UappE(kb/ks)πr2Ct

(11) Purdon, J. G.; Pagotto, J. P.; Miller, R. K. J. Chromatogr. 1989, 475, 261272. (12) Harvey, D. J.; Horning, M. G. J. Chromatogr. 1973, 79, 65-74. (13) Smith, R.; Schager, J. J. J. High Resolut. Chromatogr. 1996, 19, 151-154. (14) Verweij, A.; Boter, H. L. Pestic. Sci. 1976, 7, 355-362. (15) Shih, M. L.; Smith, J. R.; McMonagle, J. D.; Dolzine, T. W.; Gresham, V. C. Biol. Mass Spectrom. 1991, 20, 717-723. (16) Tornes, J. Aa.; Johnsen. B. A. J. Chromatogr. 1989, 467, 129-138. (17) Daughton, C. G.; Cook A. M.; Alexander, M. Anal. Chem. 1979, 51, 19491953. (18) Ruppel, M. L.; Suba, L. A.; Marvel. J. T. Biomed. Mass Spectrom. 1976, 3, 28-31. (19) Bossle, P. C.; Martin, J. J.; Sarver, E. W.; Sommer, H. Z. J. Chromatogr. 1983, 267, 209-212. (20) Wills, E. R. J.; Hulst, A. G. J. Chromatogr. 1988, 454, 261-272. (21) Roach, M. C.; Ungar, L. W.; Zare, R. N.; Reimer, L. M.; Pompliano D. L.; Frost, J. W. Anal. Chem. 1987, 59, 1056-1059. (22) (a) Kientz, C. E.; Verweij, A.; de Jong, G. J.; Brinkman, U. A. T. J. Microcolumn Sep. 1992, 4, 465-475. (b) Kientz, C. E.; Verweij, A.; de Jong, G. J.; Brinkman, U. A. T. J. Microcolumn Sep. 1992, 4, 477-483. (23) Kientz, C. E.; Hooijschuur, E. W. J.; Brinkman, U. A. T. J. Microcolumn Sep. 1997, 9, 253-259. (24) Oehrle, S. A.; Bossle, P. C. J. Chromatogr., A 1995, 692, 247-252. (25) Kingery, A. F.; Allen, H. E. Anal. Chem. 1994, 66, 155-159. (26) Bossle, P. C.; Reutter, D. J.; Sarver, E. W. J. Chromatogr. 1987, 407, 399404. (27) Schiff, L. J.; Pleva, S. G.; Sarver, E. W. In Ion Chromatographic Analysis of Environmental Pollutants; Mulick, J. D., Sawicki, E., Eds.; Ann Arbor Science: Ann Arbor, MI, 1979; Vol. 2, pp 329-344. (28) de Griend C. E. S.-v.; Kientz, C. E.; Brinkman, U. A. T. J. Chromatogr., A 1994, 673, 299-302. (29) (a) Mercier, J.-P.; Morin, P.; Dreux, M.; Tambute, A. J. Chromatogr., A 1996, 741, 279-285. (b) Mercier, J.-P.; Morin, P.; Dreux, M.; Tambute, A. J. Chromatogr., A 1997, 779, 245-252. (30) Pianetti, G. A.; Taverna, M.; Baillet, Aa.; Mahuzier, G.; Baylocq-Ferrier, D. J. Chromatogr. 1993, 630, 371-377. (31) Nassar, A.-E. F.; Lucas, S. V.; Jones, W. R.; Hoffland, L. D. Anal. Chem. 1998, 70, 1085-1091. (32) (a) Nassar, A.-E. F.; Emery, A.; Hoffland, L. D. Proc. Am. Chem. Soc., Div. Environ. Chem. 1997, 37 (1), 39-42. (b) Hoffland, L. D.; Calloway, R.; Emery, A.; Nassar, A.-E. F. Proc. Am. Chem. Soc., Div. Environ. Chem. 1997, 37 (2), 41-43. (33) (a) Nassar, A.-E. F.; Lucas, S. V.; Myler, C. A.; Jones, W. R.; Campisano, M.; Hoffland, L. D. Anal. Chem. 1998, 70, 3598-3604. (b) Battelle Memorial Institute, Analytical Chemistry, Columbus, OH 43201, unpublished results.

(1)

Uapp is the apparent mobility of analyte (Uapp ) Uep + UEOF, Uep is the electrophoretic mobility of the analyte and UEOF is the mobility due to the electroosmotic flow, EOF), E is the applied electric field (V/m), r is the capillary radius, C is the sample concentration (mol/m3), and kb/ks is the ratio of conductivities of the buffer in the capillary and the sample. Although the theoretical and, to a limited degree, the practical advantages of electrokinetic injection have been demonstrated,10 the broader development of applications has lagged due to its quantitative limitations. These limitations exist because analyte transfer into the capillary during electrokinetic injection depends on several variables which may be either unknown or poorly controlled (for example, the EOF in the sample near the capillary inlet, sample total ionic content, and nonanalyte electrophoretic mobilities). Despite these quantitative limitations, electrokinetic injection is very simple and requires no additional instrumentation, and to a large extent, these quantification difficulties can be overcome by the use of sample pretreatment and the use of internal standards or the standard addition approach for quantitation. Several papers have been published on the analysis of chemical warfare agent degradation products using either gas chromatography/mass spectrometry (GC/MS) or high-performance liquid chromatography (HPLC) with fluorescence detection.18-21 The limitations of the methods used in these studies include the following: insufficient sensitivity, interference from naturally occurring components in the matrixes, requirement for derivatization, or long analysis times. Others have reported on microcolumn liquid chromatography and CE with flame photometric (9) Conference on Disarmament Special Committee on Chemical Weapons. CD/ CW/WP. 367 Geneva, October 7, 1991. (10) Rose, D. J.; Jorgenson, J. W. Anal. Chem. 1991, 60, 642-648.

1286 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999

Chart 1. Structure of the Hydrolysis Products of Nerve Agents

degradation products in any aqueous solution at three-digit microgram per liter levels. In addition, we have shown that the interday and intraday reproducibility for the anions were in the range of about 1-3% RSD with accuracy reflected by spike recoveries in actual CW agent degradation byproduct samples of not less than 86%.33 In this paper, we report our work on the electrokinetic injection method that lowers the detection limits of CE analysis (using reversed EOF) of alkylphosphonic acids and their monoesters. When we applied our CE method to environmental samples, no interferences were observed for these analytes in the electropherogram. Using electrokinetic injection, the detection limits were as low as 1-2 µg/L in environmental water samples and as low as 25-50 µg/L in aqueous leachates of soil samples. These detection limits represent up to a 100-fold improvement over the limits we found for pressure injection. We also investigated the effect of electrokinetic injection voltage and time on these separations. EXPERIMENTAL SECTION Apparatus. All experiments were performed on either a Beckman P/ACE System 5500 with a UV detector using a 75 µm (i.d.) × 60 cm (effective length) fused-silica capillary (Beckman Instruments, Inc., Fullerton, CA) or a Hewlett-Packard (HP) capillary electrophoresis system with UV detector using a 75 µm (i.d.) × 56 cm (effective length) fused-silica capillary. Both instruments have the capability to control the temperature of the fused-silica capillary. The electropherogram signals were recorded using P/ACE Station software for the Beckman system and HP Chem-Station software for the HP system. Reagents. Chart 1 shows the structures of the analytes that have been studied in this work. Alkylphosphonic acids and their monoesters were synthesized in our laboratory (U.S. Army Material Command Treaty Laboratory, Aberdeen Proving Ground, MD). Commercial sources were as follows: didodecyldimethylammonium bromide (DDAB, >99%), Eastman Kodak (Rochester, NY); phenylphosphonic acid and boric acid, Sigma Chemical Co. (St. Louis, MO); and isooctylphenoxypoly(ethoxy)ethanol (Triton X-100), Union Carbide. All reagent solutions and diluted samples were prepared using deionized water. Solutions were stored in Nalgene (polypropylene) plastic bottles. All buffers were degassed and filtered through a 0.45-µm cellulose nitrate membrane filter prior to use. GHP Acrodisc 0.45-µm pore size (Gelman Sciences, Ann Arbor, MI) Dionex On-Guard-A, Ba, H, and Ag cartridges (Dionex Corp., Sunnyvale, CA) were used for sample pretreatment prior to the CE analysis. All other chemicals were reagent grade.

The didodecyldimethylammonium hydroxide (DDAOH) was prepared from the commercially available bromide salt by replacing the bromide with hydroxide using Dionex On-Guard A cartridges (Dionex Corp.).7 The anion-exchange cartridges are supplied in the carbonate form and are converted to the hydroxide form by passing 10 mL of 1.0 M NaOH through the cartridge followed by 10 mL of deionized water. The cartridge is then used to convert a 10-mL aliquot of 25 mM DDAB to DDAOH. This conversion step eliminates a large interfering peak in the electropherogram due to bromide and was performed as needed. The purpose of each component of the CE run buffer is as follows: Boric Acidsideally suited to control the pH at 4.0 (pK ∼4) and is also fully transparent at the detection wavelength, 210 nm. Also, because it is only partially ionized, it can be used at a relatively high concentration (200 mM) which gives a sample stacking assist to the injection. DDABsproduces the double layer at the capillary surface which reverses the EOF. Also, because it is a surfactant, it has an additional benefit to buffer storage stability due to its bacteriostatic properties. Phenylphosphonic Acidsthis chromophore for indirect UV detection is ideally suited for the target analytes due to it very high extinction coefficient, the similar electrophoretic mobility of the phenylphosphonate anion, and its high acidity producing full ionization at the buffer pH. Triton X-100saids buffer storage stability due to its bacteriostatic properties and also provides a less-noisy detector signal. Procedures. New capillaries were pretreated with a 10 min rinse of deionized water, followed by 10 min of 0.5 M sodium hydroxide, then 10 min of deionized water, and finally a 10-min rinse of the analysis buffer. Constant voltage (negative polarity, detector side anodic) was used throughout to drive the separations. The wavelength used for indirect UV detection was 210 nm and the capillary temperature was 40 °C. The electrolytes were 200 mM boric acid/10 mM phenylphosphonic acid/0.03 wt % Triton X-100/0.35 mM DDAOH at pH 4.0 for indirect UV detection. Sodium hydroxide at 0.2 N was used to adjust the pH. To obtain reproducible injections, the same injection sequence was followed for each sample: the capillary was rinsed with deionized water for 3 min and then with run buffer for 5 min prior to each injection. For the HP system, the capillary inlet and outlet levels must be equal to ensure that there is no siphoning in either direction. Environmental Samples. The environmental samples used in this work were four types of water (in addition to reagent water) and three types of soil. The environmental water samples were tap water, groundwater (untreated), Chesapeake Bay water (taken near Treaty Laboratory at Aberdeen Proving Ground, so it is mostly fresh rather than mostly saline), and artificial seawater. The artificial seawater is prepared from reagent water and a salt concentrate mixture designed for use in saltwater aquariums (Synthetic Sea Salt (Instant Ocean) supplied by Aquarium Systems Inc., Mentor, OH). The artificial seawater was prepared to give a specific gravity of 1.022 at 25 °C. The three standard soils were supplied under contract to Dugway Proving Ground/DNA by R. T. Corp., Laramie, NY 83070. These soils had been previously extensively characterized by the supplier, and the characteristics Analytical Chemistry, Vol. 71, No. 7, April 1, 1999

1287

Table 1. Selected Characteristic Data of Soil Matrixes sandy loam

sandy clay loam

loam

3540 50.0 20.5 13.9 nda 3163 97.3 228.0 5.72 38.7 21.3

11033 204.3 11.0 14.7 9.7 27000 577.3 3.04 0.92 0.48 112.3

5617 2.5 0.67 9.9 0.48 99.6 130

1773 4.6 14.9 11.3 1.85 86.0 16700

45067 0.04 1.27 19.1 5.96 96.4 526.7

3.3 93.3 3.3 S

10.0 50.0 40.0 L

10.0 65.0 25.0 SL

39 nd nd

38.7 13.7 14.3

Metal Analysis element aluminum, mg/kg barium, mg/kg calcium soluble (satd paste), mg/kg cation exchange capacity (CEC), mequiv/100 g copper, mg/kg iron, mg/kg manganese, mg/kg magnesium, mequiv/L potassium, soluble (satd paste), mequiv/L sodium, soluble (satd paste), mequiv/L zinc, mg/kg

1400 23.1 5.8 2.0 2.0 2173 90.8 0.74 0.65 0.26 10.0 Soil Analysis

parameter carbon, total, µg/g carbonate, total (CaCO3), % conductivity at 25 °C, mΩ-1/cm exchangeable acidity, mequiv/100 g organic matter, % solids, % sulfate, soluble (water), mg/kg texture by hydrometer clay, % sand, % silt, % texture classifn acetone, µg/kg toluene, µg/kg xylene, µg/kg a

Organics 39 nd nd

nd, not detected.

that are most relevant to the present work are listed, as received, in Table 1. No further or duplicate characterization of these soils or their leachates were performed in our laboratories. Environmental Sample Preparation. Soil samples were leached with reagent water by sonication for 5.0 min (1.5 g of soil and 10.0 mL of water). After centrifugation (15 min using a benchtop centrifuge), the soil leachate was decanted. Spiked leachates of soils were prepared by adding 1.0 mL of an aqueous solution of analytes directly into the 1.5-g soil sample and, once mixed, immediately leached with an additional 9 mL of reagent water. This experimental design was selected to create a bona fide soil leachate matrix with analytes at known levels, not to test the effectiveness of the leaching procedure for removing analytes from environmentally aged soils. In work in progress at Battelle for more than seven years, extraction of phosphonic acids from all types of soils using reagent water has been proved effective, so demonstration of extraction effectiveness was not an issue in the present work. The spiking solution contained analytes at the following levels: 2.00, 2.50, 3.00, and 4.00 µg/mL (ppm) for methylphosphonic acid (MPA), ethyl methylphosphonate (EMPA), isopropyl methylphosphonate (IMPA), and pinacolyl methylphosphonate (PMPA), respectively. Water samples were spiked by combining 9.00 mL of reagent water with 1.00 mL of spiking solution. Thus, both spiked water samples and soil leachates contained the same levels of analytes (assuming 100% soil leaching recovery into the aqueous phase and a 1.0-mL soil hydration demand): 200, 250, 300, and 400 µg/L (ppb) of MPA, EMPA, IMPA, and PMPA, respectively. Water samples and soil leachates 1288 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999

were prepared for CE analysis by passage through four filter/ treatment cartridges stacked in a series in the following order (inlet to outlet): 0.45-µm membrane filter (25-mm i.d.) and then three pretreatment cartridges, On-Guard-Ba (to remove the sulfate), On-Guard-Ag (to remove the chloride), and On-Guard-H (to remove cations). The On-Guard cartridges are from Dionex Corp. and were used in accordance with the supplier’s recommended flow rate of ∼1.0 mL/min. Filter/On-Guard cartridge stacks were pretreated by first aspirating 10 mL of reagent water. Then, the sample was aspirated through the stack, and the first 3 mL of filtrate was discarded prior to filling of CE analysis vials. An Alltech 16-port vacuum manifold was used to prepare (prewash) the cartridge stacks and process the samples. No analyses were performed on the aqueous field samples and soil leachates to demonstrate the amount of reduction of nonanalyte anions or cations by the stacked On-Guard cartridges. However, electropherograms were obtained of pretreatment and posttreatment (of the entire three-cartridge stack) water samples and leachates. In some cases, the pretreatment sample was fully nonanalyzable (i.e., spiked analytes were not detectable). In other cases, spiked analytes could be seen in the noncleaned up sample, but the stacked cartridge cleanup dramatically improved their detectability. Electropherograms illustrating the effect of the cleanup on sandy loan leachate soil are shown in Figure 1. RESULTS AND DISCUSSION Comparison of Electrokinetic and Hydrodynamic Injection. In our application of electrokinetic injection, the detector

Figure 1. Spiked and nonspiked electropherograms illustrating the effect of the sample cleanup cartridge stack: precleanup (upper), postcleanup (lower). The sample is from soil, sandy loam leachate. Experimental conditions: buffer, 200 mM borate, 10 mM phenylphosphonic acid, 0.03 wt % Triton X-100, pH 4.0 and 75 µm (i.d.) × 56 cm fused-silica capillary; peaks detected by indirect UV detection at 210 nm; injection, voltage, 30 kV (negative polarity); temperature, 40 °C; electrokinetic injection, -10 kV for 10 s.

end of the capillary is immersed in the run buffer so that the reversed EOF (flowing from detection end to injection end) does not deplete the capillary of run buffer during the injection. Thus, migration of analyte anions into the capillary inlet during electrokinetic injection is upstream against the exiting capillary EOF. This migration into a zone of higher ionic strength accounts for the sample injection enhancement (versus hydrodynamic injection). Equation 1 clearly shows that electrokinetic injection is increased with lower ionic strength (conductivity) in the sample compared with the run buffer at pH 4.0, so the greatest injection enhancement should be observed with samples containing only analyte and reagent water (i.e., containing no anions contributed by the sample matrix.) Figure 2 compares the CE detection response for MPA in reagent water using hydrodynamic and electrokinetic injections. Both electropherograms were generated under identical conditions (except for injection method) and are plotted on identical vertical scales. The upper trace used electrokinetic injection (10 s at -10 kV) and the MPA concentration was 0.1 µg/mL. The lower trace used hydrodynamic injection (10 s at 50 mbar) and a 100-fold higher concentration of MPA, 10 µg/mL. The signal-tonoise ratios for these two analyses were about the same, 60 for electrokinetic injection and 65 for hydrodynamic injection. Thus, Figure 2 shows greater than a 100-fold improvement in CE analysis detection for electrokinetic injection compared to hydrodynamic injection. The detection limits for methylphosphonic acid and its monoesters in reagent water using electrokinetic injection is about

Figure 2. Comparison of the electropherograms of the MPA. Experimental conditions: buffer, 200 mM borate, 10 mM phenylphosphonic acid, 0.03 wt % Triton X-100, pH 4.0 and 75 µm (i.d.) × 56 cm fused-silica capillary; peaks detected by indirect UV detection at 210 nm; injection, voltage, 30 kV (negative polarity); temperature, 40 °C. (upper electropherogram) Electrokinetic injection (-10 kV for 10 s). (lower electropherogram) Hydrodynamic injection (50 mbar for 10 s).

1-2 µg/L, based on data from artificial seawater spiked near this level. Effect of Electrokinetic Injection Voltage and Injection Time on Separation. We investigated the effect of electrokinetic injection voltage on separation over the range -5 to -30 kV. As expected, increased voltage initially produced increased peak heights but eventually resulted in reduced resolution. Also, we studied the effect of increasing the injection time from 5 to 30 s at -10 kV. As the time of injection increases, peak heights and Analytical Chemistry, Vol. 71, No. 7, April 1, 1999

1289

Table 2. Comparison of Electrokinetic Analyte Injection Efficiency for Environmental Matrixes peak area, arbitrary units MPA EMPA IMPA PMPA spike

level,a

µg/L

water samples tap water groundwater bay water artificial seawater reagent water soil leachates sandy loam sandy clay loam loam ratio, groundwater/sandy clay loam

200

250

300

400

12.4 31.8 16.7 14.7 30.7

8.0 21.6 9.5 9.9 19.0

8.2 20.7 9.8 17.0b 17.2

12.9 31.3 15.2 15.5 24.5

5.5 3.0 4.3 10.6

3.5 2.4 2.4 9.1

3.7 2.2 3.1 9.4

6.9 3.4 4.0 9.1

a Soil leachates; the level reflects 100% recovery into the aqueous phase. b Thought to be an outlier; not included in ratio calculations (see Internal Standard discussion).

Figure 3. Analysis of phosphonic compounds at 200-400 µg/L extracted from soil, sandy loam (upper trace), and waters (lower trace) by CE. Experimental conditions were as the electrokinetic injection in Figure 1.

peak areas increase until, at 15 s of injection time, the peaks start to lose resolution. On the basis of these results, we conclude that optimal electrokinetic injection conditions in this system are -10 kV for 10 s. However, one must keep in mind that these optimums are for samples containing only analytes in reagent water over the analytical range of interest (5-500 µg/L). In this scenario, sample loading has apparently reached a steady state at 15 s so that any additional increase in the loading time only degrades the separation as space charge and diffusion limitations start to dominate. It is expected that samples with higher ionic strength would reach this steady-state point at a lower injection time. CE Analysis of Spiked Environmental Samples. MPA, EMPA, IMPA, and PMPA are indicator compounds for the presence, or past use, of methylphosphonate-based CW agents. IMPA, PMPA, and EMPA are the first-step hydrolysis products of GB (2-propyl methylphosphonofluoridate), GD (pinacolyl methylphosphonofluoridate) and VX (O-ethyl S-[2-(diisopropylamino)ethyl] methylphosphonothiolate), respectively, and MPA is the further hydrolysis product to the diacid of all three of these. Sandy loam, loam, and sandy clay loam have been chosen as three different types of soils containing a wide range of levels and types of both organic and inorganic constituents, some of which are known to form an ion pair with the phosphonic acids. Figure 3 (upper trace) illustrates a mixture of these four phosphonic acids in soil (sandy loam) spiked at a level that would give 200-400 µg/L of the acids in the aqueous leachate (1.3-2.6 µg/g of soil). Sandy loam gives the largest peak areas, loam is intermediate, and sandy clay loam is the smallest (Table 2). Data 1290 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999

Figure 4. Analysis of phosphonic compounds extracted from soil, sandy loam (upper trace) at 50-100 µg/L, and seawater (lower trace) at 5-10 µg/L by CE. Experimental conditions were as the electrokinetic injection in Figure 1.

produced with blank soils prepared the same way showed no peaks that would interfere with the acids. Figure 3 (lower trace) shows the electrophergrams for similarly spiked groundwater, tap water, artificial seawater, and bay water. As for the soils, correponding electropherograms from nonspikes of these water matrixes produced no peaks that would interfere with these four phosphonic acids. Figure 4 (upper trace) shows an electropherogram of the phosphonic acids extracted from sandy loam soil. The spike levels for the four analytes give about 50-100 µg/L in the leachate, and this level gives a detection ∼2-fold above the

detection limit and seawater (lower trace) at 5-10 µg/L by CE. These results indicate that this electrokinetic injection-CE method could be successfully applied for the separation of these phosphonic acids from soils and environmental water samples at low parts-per-billion (µg/L), making it potentially useful in analyzing environmental samples. Effect of Sample Ionic Strength. Variations in the total ionic content of the sample can have a great impact on the amount of an analyte transported into the capillary by electrokinetic injection. Indeed, spiked samples of the artificial seawater and two of the soil leachates (sandy clay loam and loam) were not even analyzable by this CE-electrokinetic injection method without prior treatment with the Dionex On-Guard cartridges to reduce the ionic strength. For the other environmental samples, the ratio of the analyte peak area after On-Guard cartridge treatment versus without this cleanup step (with identical electrokinetic injection conditions) ranged from nearly 30 for the bay water to a little less than 2 for the sandy loam soil. The On-Guard-Ag and OnGuard-Ba cartridges are designed to specifically remove halides (especially choride) and sulfate, respectively. Thus, the degree of improvement in analyte transfer with cartridge cleanup reflects the degree to which the total ionic content is represented by these anions. In tests to determine whether the cartridge stack absorbed any of the phosphonate analytes during this cleanup step, it was found that reagent water spiked either before or after cartridge stack cleanup gave identical CE analysis results when hydrodynamic injection was used. On this basis, it was assumed that analytes were not absorbed by the stack when the environmental matrixes were processed. Even after cleanup, there is still an extremely large difference in efficiency of analyte transfer across these very diverse environmental samples. Table 2 shows the peak areas obtained from cleaned up samples spiked at the same level (prior to cleanup) with the four alkylphosphonic esters. For all four analytes, the amount of analyte transferred to the capillary by electrokinetic injection ranged over a factor of ∼10 with matrix. It is clear that analyte injection efficiency is highest when the cleanup step removes a high percentage of the nonanalyte anions (artificial seawater, groundwater) and lowest when anions other than halide and sulfate (removed by the On-Guard cartridges) predominate (i.e., soil leachates). The differences in analyte injection efficiency among the soil leachates probably have at least a partial explanation in a comparison of the characteristics shown in Table 1. The soil with the least efficient analyte transfer (sandy clay loam) clearly has the highest leachable ionic content, and conversely, the soil leachate with the greatest analyte-transfer efficiency (sandy loam) is the lowest in nearly every category of ionic content indicators (Table 1). Possibly, the extremely high leachable sulfate from the sandy clay loam exceeded the capacity of the On-GuardBa cleanup cartridge. Thus, the detection limit enhancement achievable with electrokinetic injection for these alkylphosphonates is highly dependent on the ability to selectively remove nonanalyte anions from the aqueous matrix prior to CE analysis. This was especially apparent when the positive control samples (postcleanup spiked reagent water) showed analyte peak areas that were ∼60% higher than the corresponding non-cleaned-up reagent water sample.

Further investigation by CE (using different analysis conditions) revealed that the nontreated reagent water did contain traces of chloride, accounting for the cleanup enhancement. Quantitation Approaches. Because of the strong dependence of electrokinetic analyte injection efficiency on the quantity and specific nature of the ionic content of the sample, quantitation using an external standard approach is clearly not possible. One quantitation option is the use of an internal standard (IS). The effect of an IS on CE analysis variability with electrokinetic injection has been reported.34 This work used two internal standards and reported the analysis precision (∼1% RSD) obtained with samples in reagent water only, so the issue of substantial variation in sample ionic content was not addressed.34 An ideal IS for quantitation of these analytes would closely resemble them in the chemical characteristics that affect the CE separation but would not be a naturally occurring species or be related to any known CW agent, for example, alkylphosphonate analogues where the alkyl group attached to phosphorus is ethyl or propyl rather than methyl. Although we have not employed such an IS analogue in this work, the data in Table 2 offer some indication of the potential usefulness of this option. To check the applicability of an IS for quantitation, any one of the four Table 2 analytes can be designated as the IS for quantitation of the other three analytes. A straightforward measure of the success of an internal standard approach to quantification across this wide range of matrixes is the variance in the peak areas ratios (analyte peak area divided by the peak area of the analyte taken to be the IS). For the most disparate case (MPA as an IS for PMPA, or vice versa), the peak area ratio variance across the seven matrixes is 14% RSD. For the more closely eluting pairs, the variances are somewhat less, as would be expected for a closer match in electrophoretic mobility: 11.0, 10.5, and 10.1% RSD for the pairs, MPA/EMPA, EMPA/IMPA, and IMPA/PMPA, respectively. This degree of accuracy is sufficient for most environmental analysis applications. CE Buffer Stability and Freedom from Interferences. The CE buffer used in this work is remarkably stable. No degradation (as visualized by color or pH change, appearance of precipitate, or changes in the electropherograms obtained) has been observed over months of storage/use of the stock buffer solution. An additional favorable property of this buffer system is that it eliminates several major interferences. One of these interferences is fluoride, which is produced in equimolar quantities (versus the alkylphosphonate) in hydrolytic degradation of phosphonofluoridate-type CW agents (for example, GB, GD, and GF). Because of the acidic pH of the buffer, fluoride is not observed in the separation due to its reaction with the silica inner wall of the capillary.31,35 In earlier work with environmental samples, no interferences were observed when spiking with 100 µg/mL of fluoride ion.31,33 A second interference is the weakly acidic carbonate anion, which is also not observed in the separation, as it has a pK value (6.3) that is well above the buffer pH of 4.0. This acidic pH also eliminates any possible interference from humic acids that might be recovered in the soil leachates. A fourth potential source of interference, the chloride ion, is essentially completely removed by the On-Guard-Ag cartridge, but it does not eluate near the phosphonic acids anyway.33 (34) Dose, E. V.; Guiochon, G. A. Anal. Chem. 1991, 63, 1154-1158. (35) Thermo BioAnalysis Corp., Franklin, MA 02038, unpublished results.

Analytical Chemistry, Vol. 71, No. 7, April 1, 1999

1291

CONCLUSIONS We have demonstrated a CE-electrokinetic injection analysis method which offers increased analysis sensitivity in a highly direct analysis method with simple and minimal sample preparation steps. The method is well-suited for the determination of CW agent degradation products in environmental and possibly other types of complex matrixes. The method is simple, easy, fast, sensitive, and essentially interference-free and does not require expensive instrumentation. The coelectroosmotic conditions reduced the analysis time and eliminated the observed baseline anomalies observed when EOF modifier was absent, and this reversed EOF approach achieved complete separations in a 3-min electrophergram. Significant detection limit enhancement with electrokinetic injection (up to 100-fold versus pressure injection) has been shown, and the detection limit extends to the single to low two-digit ppb for aqueous environmental samples and to the low two-digit ppb for aqueous leachates of soils. For all seven environmental matrixes, excellent separation efficiency and no interfering background substances were observed. Complete separations were achieved in a 3-min electropherogram. The data

1292 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999

obtained support the proposed application of an internal standardbased quantitation strategy over a wide range of environmental matrixes giving practical detection limits in the mid- and low-ppb range. In addition, the acidic CE electrolyte buffer increases the shelf life because it does not absorb atmospheric CO2 (which drops the pH) as readily as alkaline pH electrolytes, and we have realized useable stability for more than one year, to date. Another important benefit of the use of this acidic buffer system is that it eliminates any possibility of interference from carbonate/ bicarbonate and, importantly for the CW agent detection scenario, fluoride. ACKNOWLEDGMENT We thank the U. S. Army Material Command Treaty Laboratory, Aberdeen Proving Ground, MD, for their support.

Received for review August 10, 1998. Accepted January 21, 1999. AC980886D