Anal. Chem. 2005, 77, 711-717
Application of Single Drop Microextraction for Analysis of Chemical Warfare Agents and Related Compounds in Water by Gas Chromatography/Mass Spectrometry Meehir Palit, Deepak Pardasani, A. K. Gupta, and D. K. Dubey*
Vertox Laboratory, Defence Research and Development Establishment, Gwalior, India
Retrospective detection and identification of chemical warfare agents (CWAs) is important from the verification point of view of the Chemical Weapons Convention. In the present work, a novel method for determination of CWAs and their markers in water has been described. It is based on a single drop micro extraction (SDME) of analytes and gas chromatography/mass spectrometric identification. Extraction conditions, such as solvent selection, agitation, extraction time, and salt content, were found to have significant influence on SDME. The conditions optimized for extraction of CWAs were 1 µL CH2Cl2/CCl4 (3:1 v/v), 30-min extraction time, 300-rpm stirring rate, and with or without NaCl addition. Under optimized conditions, comparison of SDME, solid-phase microextraction, and liquid-liquid extraction was also made. The limit of detection by SDME ranged from 75 to 10 µg L-1 at a signal-to-noise ratio of 10:1. Trace analysis of chemical warfare agents (CWAs) and their related compounds (the environmental markers) is an important problem area of current research.1,2 It has gained impetus in light of a verification program of the Chemical Weapons Convention (CWC).3,4 The CWC came into force in April 1997 with the objective of prohibition of proliferation of chemical weapons.5,6 The goal of CWC is being met by execution of its unique verification program. Signatory states (so far, 164) of the CWC have formed an organization, known as the Organization for Prohibition of Chemical Weapons (OPCW), whose headquarters is in The Netherlands and is responsible for implementation of the CWC * Corresponding author. Address: Vertox Lab, Defence R&D Establishment, Jhansi Road, Gwalior-474002 India. Phone: 91-751-2233488. Fax: 91-751-2341148. E-mail:
[email protected]. (1) Hooijschuur, E. W. J.; Kientz, C. E.; Brinkman, U. A. J. Chromatogr., A 2002, 982, 177. (2) Mesilaakso, M.; Ratio, M. Encyclopedia of Analytical Chemistry; Wiley: New York, 2000; p 899. (3) Kientz, Ch. E. J. Chromatogr., A 1998, 814, 1. (4) Kostiainen, O. In Forensic Science, Handbook of Analytical Separations; Bogusz, M. J., Ed.; Elsevier Science: Amsterdam, 2000; Vol. 2, p 405. (5) Convention on the Prohibition of the Development, Production, Stockpiling and Use of Chemical Weapons and their Destruction, Technical Secretariat of the Organization for Prohibition of Chemical Weapons. The Hague, 1997; accessible through Internet http://www.opcw.nl. (6) Krutzsch, W.; Trapp, R. A commentary of CWC; Martinus Nijhoff: The Netherlands, 1994. 10.1021/ac0486948 CCC: $30.25 Published on Web 12/02/2004
© 2005 American Chemical Society
through its verification mechanism.1-7 Verification of the CWC involves unequivocal detection and identification of CWAs and their precursors, starting compounds, and degradation products (so-called markers) in samples collected by OPCW inspectors from production, storage, and suspected sites of CW activity. Inspectors generally collect samples of soil, water, air, organic liquids, etc., and perform on-site analysis for detection and identification of CWAs and their markers. In the case of any ambiguity, these samples are sent to designated laboratories appointed by the OPCW for unequivocal identification of CWAs and their markers.5-7 On the basis of the potential of being used as CWAs, their precursors, starting materials and degradation products, chemicals are enlisted in three scheduled categories of the CWC;5,6 these compounds are also termed convention-related chemicals (CRCs). The concentration of CRCs in samples collected for verification may range from neat liquid to parts per billions. Therefore, there is a dire need to develop efficient and simple extraction protocols for the detection and identification of CRCs in these samples. Among various environmental samples, water is an important matrix because it can be contaminated during deliberate or inadvertent spread of CWAs.2,4 Hence, detection and identification of CRCs in water is of paramount importance for verification analysis of the CWC.7 The complete analytical procedure of CRCs in water involves three major steps: (i) extraction, (ii) concentration, and (iii) detection and identification. Identification of extracted CRCs is preferably carried out by gas chromatography/ mass spectrometry (GC/MS) because this technique provides both characteristic spectra (in EI mode) and molecular weight information (in CI mode).8-10 Conventional methods of extraction of CRCs from water involve liquid-liquid extraction (LLE), solidphase extraction (SPE), and solid-phase microextraction (SPME).9-12 (7) Hooijshuur, E. W. J.; Hulst, A. G.; De Jong, Ad L.; De Reuver, L. P.; Van Krimpen, S. H.; Van Baar, B. L. M.; Wils, E. R. J.; Kientz, C. E.; Brinkman, U. A. Th. Trends Anal. Chem. 2002, 21 (2), 116. (8) Wils, E. R. J. In Encyclopedia of Analytical Chemistry; Meyers, R. A., Ed.; J. Wiley and Sons: Chichester, 2000; p 979. (9) Ratio, M. Recommended Operating Procedures for Sampling and Analysis in the Verification of Chemical Disarmament; Ministry for Foreign Affairs of Finland: Helsinki, 1994. (10) Kuitunen, L. M Encyclopedia of Analytical Chemistry; Wiley: Chichester, 2000; p 1055. (11) Lakso, H. A. Anal. Chem. 1997, 69, 1866. (12) Sng, M. T.; Ng, W. F. J. Chromatogr., A 1999, 832, 173.
Analytical Chemistry, Vol. 77, No. 2, January 15, 2005 711
LLE is the most widely used sample pretreatment method for liquid samples; however, the main drawback of LLE is that it is time-consuming and multistep, requires a large volume of toxic organic solvents, often leads to emulsion formation, results in analyte loss during concentration, and has poor potential of automation.13 SPE is less time-consuming than LLE but needs column pretreatment and elution with toxic organic solvents. In addition, the extraction of extremely toxic CWAs by LLE and SPE needs special protection with burdensome protective gear. SPME is a solvent-free and simple technique, free from several limitations of conventional techniques; however, it suffers from some drawbacks, such as high cost, memory effects, and a decline in performance with time.13,14 Recently, a fast, simple, inexpensive, and virtually solvent-free sample preparation method has been devised for extraction of analytes from water; this technique is known as liquid-phase microextraction (LPME) or single drop microextraction (SDME).13,15-18 SDME combines extraction, concentration, and sample introduction in a single step. The technique is based on the distribution of analytes between a microdrop of organic solvent at the tip of a microsyringe needle and the aqueous phase. The organic drop is exposed to an aqueous sample where target analyte is extracted into the drop. After attainment of equilibrium, the drop is retracted into the microsyringe and injected into the injection port of the GC or GC/MS. The SDME technique has been used for extraction of dialkylphthalates,19 nitroaromatics,14,20 polycyclic aromatic hydrocarbons,21 organochlorine compounds,22,23 triazine herbicides,24 cocaine,25 and endosulfans.26 There is no report on extraction of CWAs and their related compounds (markers) from water employing SDME as the sample preparation method. Prompted by the virtues of SDME, herein we report optimization of this technique for the extraction of some of important CWAs and their markers from water. The compounds selected for this study included five nontoxic markers of organophosphorus-based nerve agents and four toxic CWAs. The compounds taken were O,O-dibutyl n-propylphosphonate (DBPP or A), O-ethyl O-cyclohexyl n-propylphosphonate (ECPP or B), O,O-dimethyl methylphosphonate (DMMP or C), O,O-dimethyl ethylphosphonate (DMEP or D), O,O-diethyl N,N-diethylphosphoramidate (DEDEP or E), O,O-dicyclohexyl methylphosphonate (DCMP or F), O-isopropyl methylphosphonofluoridate (Sarin or G), O-cyclohexyl ethylphosphonofluoridate (CEPF or H or cyclosarin) and bis-(2-chloroethyl)sulfide (SM or I). Compounds A-F are nontoxic “markers” (byproducts, precursors, starting materials, or degradation products) of phosphorus-based CWAs (the nerve agents), hence included in schedule 2B category of the CWC.5,6 Compounds G and H are highly toxic schedule 1A1 (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26)
712
Psillakis, E.; Kalogerakis, N. Trends Anal. Chem. 2002, 27, 53. Zhao, R.; Chu, S.; Xu, X. Anal. Sci. 2004, 20, 663. He, Y.; Lee, H. K. Anal. Chem. 1997, 69, 4634. Jeannot, M. A.; Cantwell, F. F. Anal. Chem. 1997, 69, 235. Jeannot, M. A.; Cantwell, F. F. Anal. Chem. 1997, 69, 2935. Jeannot, M. A.; Cantwell, F. F. Anal. Chem. 1996, 68, 2236. Batlle, R.; Nerin, C. J. Chromatogr., A 2004, 1045, 29. Psillakis, E.; Kalogerakis, N. J. Chromatogr., A 2001, 938, 113. Hou, L.; Lee, H. K. J. Chromatogr., A 2002, 976, 377. De Jager, L. S.; Andrews, A. R. Analyst 2000, 125, 1943. De Jager, L.S.; Andrews, A. R. Chromatographia 1999, 50, 733. Shen, G.; Lee, H. K. Anal. Chem. 2002, 74, 648. De Jager, L. S.; Andrews, A. R. J. Chromatogr., A 2001, 911, 97. Lopez-Blanco, M. C.; Blanco-Cid, S.; Cancho-Grande, B.; Simal-Gandara, J. J. Chromatogr., A 2003, 984, 245.
Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
nerve agents, and I is a cytotoxic, blistering agent belonging to the schedule 1A4 category.
The main objective of this study was to optimize the extraction parameters of SDME from water with nontoxic CRCs (compounds A-F) and apply them for the extraction of toxic agents (compounds G-I). The factors affecting the extraction of selected analytes, such as selection of solvent, sample stirring, salt addition, and sampling time, were assessed and optimized. The comparison of SDME with LLE and SPME is also made for compounds E-I. EXPERIMENTAL SECTION Chemicals, Solvents, and Materials. Compounds A-I were prepared in-house as per reported procedures.27,28 Caution: Compounds G-I are extremely toxic chemical warfare agents; they should be prepared and handled by trained professionals in an efficient fuming hood equipped with an alkali scrubber. Individuals handling them should wear a facemask, gloves, and a protective suit. Organic solvents (analytical grade) toluene, hexane, dichloromethane, and carbon tetrachloride were obtained from Aldrich Chemical Co. (U.S.A.). Toluene was redistilled in an all-glass apparatus prior to use; other solvents were used as received. For spiking the water samples, deionized, triply glass distilled water was used. SDME was performed with a 5-µL microsyringe otained from Hamilton (U.S.A.). Extractions were performed in 2-5 mL Teflon-capped glass vials from Supelco. For SDME, stirring with a Teflon-coated magnetic flea was performed on a stirrer purchased from Schott (Germany). Glaxo Qualipette pipets (200 µL) from Finland were used for transferring the aliquots in LLE and the other experiments. Standard and Spiking Solutions. Stock standard solutions (1000 mg/L) were prepared in acetonitrile separately by accurately weighing ∼0.01 g of analyte into 10-mL volumetric flasks and diluting to volume. Intermediate mix standard solutions were prepared by diluting the stock standard solutions in acetonitrile. Stock and intermediate standard solutions of internal standard, heptadecane, were prepared in the same way in extracting solvents. The final concentration of internal standard was kept at (27) Reid, E. E., Ed. Chemistry of Bivalent Sulfur; Chemical Publishing Co.: New York, 1960, Chapter 5. (28) Barnaby, F. The Problem of Chemical and Biological Warfare; Technical Aspects of Early Warning and Verification; Stockholm International Peace Research Institute (SIPRI); Humanities Press: New York, 1975; pp 192193.
10 mg/L. Water samples were spiked at a concentration of 1-10 mg/L with standard solutions of analytes and were used for the extraction experiments. To avoid the loss of analytes by hydrolysis, stock and intermediate solutions were prepared freshly every day. SDME Procedure. SDME experiments were performed by taking all the required precautions.13,16 Absence of air bubbles was ensured by washing the syringe several times (50) with organic solvent. Precision of method was improved by positioning the needle in an aqueous sample at a fixed length with stands and clamps. After each extraction, the syringe was washed several times with extractant containing internal standard. A 2-mL vial with stir bar was placed on a magnetic stirrer; 1.8 mL of water containing analytes was added to the vial. It was sealed with a PTFE-coated silicon septum. The 5-µL microsyringe filled with 1 µL of extracting solvent was inserted into the vial by piercing the septum. The needle tip was immersed into the solution and fixed ∼1 cm below the surface of the liquid, then 1 µL of extractant containing 10 mg/L internal standard was extracted out of the needle and kept suspended at the needle tip. The solution was stirred at 300 rpm for 30 min. When the extraction was complete, the drop was sucked back in the microsyringe, and the needle was removed from the sample vial. The plunger of the syringe was then depressed to the 0.5-µL position, and the tip of the needle was wiped with a tissue to remove any possible water contamination. The extracted liquid was injected into the GC/MS for analysis. SPME Procedure. For SPME, salted water (30% NaCl w/v) was spiked at a concentration of 1 mg/L with working standard of compounds E-I. Fresh spiking was done daily to avoid loss of analytes (particularly G, H, and I) due to hydrolysis. SPME was performed using a 65-µm poly(dimethylsiloxane)/divinylbenzene (PDMS/DVB) SPME fiber and a SPME holder assembly, all purchased from Supelco (Sigma Aldrich). Before starting the extractions, the fiber was conditioned as per the instructions of the supplier. The SPME fiber holder assembly was then clamped at a fixed location above the 2-mL glass vial containing the stirred spiked solution. The fiber was exposed to the salted solution for 30 min, and the fiber was retracted and transferred to the heated injection port of the GC/MS for 5 min. LLE Procedure. Similar to SPME, 1.8 mL of freshly prepared salted (30% w/v NaCl) water solution was taken in a 4-mL poly(tetrafluoroethylene) (PTFE) capped vial. It was extracted with 4 × 400 µL of dichloromethane. Organic layers were transferred into a second sampling vial after each extraction. The combined extracts were concentrated to 200 µL by a gentle stream of nitrogen. The concentrated aliquot was analyzed by GC/MS by injecting 0.5 µL of sample. GC/MS Analysis. The GC/MS analyses were performed in EI mode (70 eV) with an Agilent 6890 GC equipped with a model 5973 mass selective detector (Agilent Technologies, U.S.A.). An SGE BPX5 capillary column with 30-m length × 0.32-mm i.d. × 0.25-µm film thickness was used at a temperature program of 50 °C (2 min)-15 °C/min-240 °C-40 °C/min-280 °C (2 min). Helium was used as the carrier gas at a constant flow of 1.2 mL/ min. The samples were analyzed in splitless mode at injection temperature of 250 °C, EI source temperature of 230 °C, and quadrupole analyzer at 150 °C.
Figure 1. Extraction efficiency of organic solvents for compounds A and B.
RESULTS AND DISCUSSION SDME Optimization. The SDME process is driven by concentration differences of analytes between the aqueous and organic phases. Mass transfer of the analytes from the aqueous to organic microdrop continues until thermodynamic equilibrium is attained or extraction is stopped. To understand the dynamic characteristics of the micro extraction process, a mathematical model has been developed,13,16,18 which helps in optimizing the SDME parameters. According to this theoretical treatment, parameters that control the mass transfer of the analyte from the aqueous to organic microdrop should be assessed and optimized. The parameters that control the mass transfer include drop size, extraction time, stirring rate, addition of salt, and organic solvent. Hence, to develop SDME for determining CRCs in a water sample, extraction profiles with respect to these parameters were assessed. To obtain optimized extraction conditions, the ratio of peak area of analyte and that of internal standard (heptadecane) were used in GC/MS analysis of the extracts. Selection of the Extracting Solvent. To achieve good extraction, several solvents differing in polarity and water solubility were screened on the basis of the principle of “like dissolves like”. The final choice of solvent was based on extraction efficiency, rate of drop dissolution, and excellent gas chromatographic behavior. Solvent selection was performed by extraction of the spiked water sample (1.8 mL at 10 mg/L) with the organic solvent drop (1 µL). Two sets of experiments were performed. In the first experiment, compounds A and B were spiked and extracted with single solvents, such as dichloromethane, carbon tetrachloride, toluene, n-hexane, and isooctane. In the second set, compounds C-F were spiked and extracted with single and mixed solvents, such as dichloromethane, carbon tetrachloride, toluene, and dichloromethane/carbon tetrachloride in ratios of 3:1, 1:1, and 1:3 v/v and dichloromethane/toluene in ratios of 3:1 and 1:1. Spiked water samples were extracted at room temperature (25 °C) for 30 min with stirring at 300 rpm. After extraction, the plunger was withdrawn and the microdrop was analyzed by GC/MS as per the procedure given in the Experimental Section. Three replicate analyses were performed for each solvent and the mixed solvents. Averages of the ratios of the peak areas of the analytes versus the internal standard for each solvent system are shown in Figures 1 and 2. As is evident from Figure 1, higher extraction efficiencies were achieved with CH2Cl2, CCl4, and toluene than with hexane and isooctane. It seems that the slight polarity of phosphorus esters (owing to a polarizable PdO bond) had a better affinity with slightly polar solvents. Keeping this logic in mind, extraction Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
713
Figure 2. Extraction efficiency of organic solvents for compounds C-F.
Figure 3. Plots of observed concentration of DBPP in CH2Cl2/CCl4 (3:1) microdrop versus stirring time at various stirring rates.
of slightly more polar phosphorus esters (C-F) bearing smaller O-alkyl groups (such as DMMP and DMEP) was performed with mixed solvents also, where the polarity of the extracting solvent could be fine-tuned by changing the ratios of the two solvents. Figure 2 clearly demonstrates that the best extraction of analytes was accomplished with the mixed solvent CH2Cl2/CCl4 at a ratio of 3:1 v/v. It appears that the modified polarity of the mixed solvent CH2Cl2/CCl4 at a ratio of 3:1 is best suited to extract the esters of varying polarity, that is, more polar DMMP and less polar DCMP. On the basis of these observations, CH2Cl2/CCl4 (3:1) was chosen as the extracting solvent in further experiments. Rise of analytes with increasing drop size has been observed by several workers;14,19,20,27 however, since a larger drop tends to dislodge from the needle of the microsyringe, the drop size was kept to 1 µL in all the extractions. Optimization of Agitation. Agitation of the sample reduces time to reach thermodynamic equilibrium and increases extraction efficiencies. To evaluate the effect of sample stirring, water samples (spiked at 10 mg/L with analytes) were extracted in triplicate with CH2Cl2/CCl4 at ratios of 3:1 v/v at different time intervals with varying stirring rates (100 to 400 rpm). Stirring rates above 400 rpm were not evaluated because they destabilized the drop. The results are typically shown in Figure 3 with compound A (DBPP). It is evident from this graph that at a stirring rate of 100 rpm, it took a longer time to attain the equilibrium, and the extraction was also less. At 400 rpm, a maximum amount of analyte was extracted with the fastest attainment of equilibrium, but in some cases, it destabilized the drop, also; therefore, a stirring rate of 300 rpm was fixed for further microextractions. 714 Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
Figure 4. Plots of ratios of peak areas of compound A and internal standard versus time at different NaCl concentrations.
Optimization of Time and Salt addition on Extraction. It is obvious that sufficient time must be allowed to attain an equilibrium of analytes between the aqueous and organic drop. The sorption time profiles for the analytes were studied by plotting the ratios of the peak areas of the analytes versus the internal standard with respect to time. The optimum time was found to be 30 min to reach equilibrium. Addition of salt to the analyte may have several effects on the extraction as it increases the ionic strength of the solution. Salting out effect has been used in liquid-liquid extraction. Generally, addition of a certain amount of salt can decrease the solubility of analytes in the aqueous phase and enhance their partitioning into the organic phase. For SDME, however, the presence of salt was found to restrict the extraction of analytes.13,14,16,26 This negative effect of salt on the extraction was attributed to a change of physical properties of the Nernst diffusion film, which reduced the rate of diffusion of the analytes into the microdrop. We have also assessed the effect of salt on extraction efficiency of the organic drop for analytes at different time intervals. The effects of NaCl addition (0-30%) on extraction efficiency were evaluated in triplicate by spiking the compounds at 10 mg/L and extracting with 1 µL of CH2Cl2/CCl4 (3:1) at different time intervals. The effects of salt addition on SDME are typically depicted in Figure 4. It is evident from Figure 4 that at zero salt concentration, the equilibrium is attained more quickly, but the amount of extracted analyte is much less. Addition of salt requires a longer time to reach to equilibrium, which is evident from the shifting of the equilibrium time to the higher side with an
Figure 5. TICs of compounds E-I in GC/MS analysis of extracts obtained after (a) SDME, (b) SPME, and (c) LLE.
increasing concentration of salt. At 30% NaCl, the required time to attain equilibrium was 30 min. It is important to note that the presence of salt required a longer time to establish the equilibrium, but the overall amount of extracted analyte was found to increase with increasing salt concentration. These observations evidently demonstrate both phenomena, that is, the reduced diffusion and salting out of analyte from the aqueous to organic microdrop in the presence of salt, provided sufficient time is
allowed to attain equilibrium. On the basis of these findings, 30 min was fixed for SDME in further experiments. Comparison of SDME versus SPME and LLE for Extraction of CWAs. The optimized SDME procedure was compared with SPME and LLE for determination of CWAs in spiked water samples. Water containing 30% NaCl was fortified with compounds E-I at a concentration of 1 mg/L. Extractions were performed as per the procedures described in the Experimental Section. Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
715
Figure 6. TIC of Sarin by SDME at concentrations of (a) 1 mg/L, (b) 100 µg/L, and (c) 75 µg/L.
Extraction time for SDME and SPME was 30 min, with a stirring rate of 300 rpm. In SDME, the optimized solvent (1 µL of CH2Cl2/CCl4 at a ratio of 3:1) was used as the extracting liquid. A typical GC/MS analysis profile (the total ion chromatogram, TIC) of three extracts, that is, SDME, SPME and LLE, is shown in Figure 5 for compounds E-I. The important observations with 716
Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
the three techniques are noteworthy. The response (in terms of peak area count) of DEDEP (E), DCMP (F), and SM (I) was higher with SPME than SDME and LLE. Compounds E, F, and I showed a higher response with SPME than with SDME. Probably, higher hydrophobicity of these analytes facilitated their partitioning into the more hydrophobic PDMS fiber. On the other hand,
Table 1. Linearity, Limit of Detection (LOD) and Precision (RSD) for Determination of CWAs and Their Markers by SDME Followed by GC/MS Analysis compd
linearity range mg/L
RSD % (n ) 6)
r2
LOD (µg/L)
DBPP ECPP DMMP DMEP DEDEP DCMP SARIN CEPF SM
0.1-10 0.1-10 0.1-10 0.1-10 0.1-10 0.1-10 0.5-10 0.1-10 0.1-10
10.0 9.4 10.3 11.1 9.7 12.0 14.0 8.9 9.5
0.9995 0.9989 0.9991 0.9990 0.9988 0.9996 0.9984 0.9997 0.9997
10.0 10.0 20.0 15.0 10.0 10.0 75.0 25.0 30.0
SPME was found to be similar in extracting CEPF (H) and did not extract Sarin (I) at al. (Figure 5b), whereas SDME extracted all the compounds, with a good peak area count in the GC/MS analysis (Figure 5a). Better extraction of polar phosphonofluoridates (due to a P-F bond) with SDME than with SPME can be explained by the polarity difference of the microdrop (CH2Cl2/ CCl4) and the PDMS fiber. The microdrop containing CH2Cl2/ CCl4 (3:1) was more polar than the PDMS fiber; hence, the extracted polar phosphonofluoridates by the principle of like dissolves like. Another reason for not detecting the Sarin by SPME could be its loss during transfer of the fiber to the injection port. Another important difference between SDME and SPME is the presence of background peaks in TIC obtained with SPME (Figure 5 a,b). The background peaks of siloxanes often appear in chromatograms of SPME, which could interfere with the identification of analytes; however, no such interference is observed with SDME. LLE also extracted all the analytes, but the peak area count of all the analytes was lowest under the employed analytical conditions. However, the major problem of using LLE for determination of extremely toxic CWAs in water is the special caution that is necessary to avoid the exposure that can take place during transfer and concentration of organic solvent. With SDME, no such problem occurs because in this technique, extraction is done in an isolated system. These results clearly demonstrate the superiority of SDME over LLE in terms of performance, safety, and time for detection and identification of CWAs in water. As compared to SPME, SDME has the edge in terms of versatility by choice of selecting the appropriate solvent(s).
Linearity, Limit of Detection, and Reproducibility. Under the optimized conditions, linearity, limit of detection (LOD), and reproducibility of the analytes were determined; the results are shown in Table 1. The linearity was obtained in the range of 0.110 mg/L for CWAs and their markers. Correlation coefficients (r2) varied from 0.9997 to 0.9984. LOD in GC/MS analysis (full scan mode) after SDME ranged from 10 to 75 µg/L at a signalto-noise ratio of 10:1. These LODs are much below those required (1-10 mg/L) in Official Proficiency Tests conducted by OPCW for designating the laboratories for off-site analysis of CRCs.7 A typical TIC for Sarin is given in Figure 6 at 0.075, 0.1, and 1.0 mg/L concentrations after performing SDME. To asses the precision of the SDME method, the reproducibility was determined by performing six experiments. The RSD was found to be ∼10% for most of the analytes, except that the RSD for Sarin was 14%. 4. CONCLUSION This study has dealt with development of a novel technique known as single drop microextraction (SDME) for analysis of CWAs and their markers in water. SDME is a simple, inexpensive, fast, effective, and virtually solvent-free sample preparation technique. The parameters such as selection of solvent, extraction time, stirring rate, and salt content that control the extraction of CWAs from water to the microdrop were optimized. The optimized conditions were 1 µL of CH2Cl2/CCl4 (3:1 v/v), 30 min extraction time, 300 rpm stirring rate, with or without salt addition. The LODs under optimized conditions ranged from 75 µg/L (for Sarin) to 10 µg/L. The comparison of SDME, SPME, and LLE was also made, which revealed that SDME extracted analytes of diverse structure by virtue of solvent selection, whereas SPME, although it showed a better response in TIC for more hydrophobic compounds, was not effective against polar analyte. LLE is an exhaustive extraction technique with several limitations (longer time, higher volume of organic solvents, and often impeded by emulsification), whereas SDME and SPME are nonexhaustive equilibrium-based methods. Thus, SDME is an alternative and better sample pretreatment technique in terms of cost, speed, safety, performance, and versatility.
Received for review September 2, 2004. Accepted October 21, 2004. AC0486948
Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
717
