Anal. Chem. 1996, 68, 2533-2540
Determinations of N-Nitrosodimethylamine at Part-per-Trillion Concentrations in Contaminated Groundwaters and Drinking Waters Featuring Carbon-Based Membrane Extraction Disks Bruce A. Tomkins* and Wayne H. Griest
Organic Chemistry Section, Chemical and Analytical Sciences Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, Tennessee 37831-6120
A new solid-phase extraction procedure extracts N-nitrosodimethylamine (NDMA) at part-per-trillion (ng/L) concentrations from aqueous samples using a C18 (reversed-phase) membrane extraction disk layered over a recently introduced carbon-based extraction disk. The reversed-phase disk removes nonpolar water-insoluble neutrals and is set aside; the carbon-based disk is extracted with a small volume of dichloromethane. NDMA is quantified in the organic extract using a gas chromatograph equipped with both a short-path thermal desorber and a chemiluminescent nitrogen detector. The detection limit for the procedure, calculated using two statistically unbiased protocols, is 3 ng of NDMA/L; the analyte recovery is ∼57%. A related procedure substitutes a standard automatic sampler for the short-path thermal desorber and is suitable for determining NDMA in heavily contaminated (>300 ng of NDMA/L) aqueous samples. The detection limit for the procedure, calculated in the same manner as above, is 300 ng of NDMA/L, with an analyte recovery of ∼64%. The detection limits and measured recovery values are comparable to those observed in earlier work, in which a conventional continuous overnight extraction with dichloromethane was used to remove NDMA from the aqueous samples. The newer procedures described herein offer a 50-fold savings in extraction time and a 100-fold reduction in dichloromethane consumed per sample while maintaining the wide range (3-4 orders of magnitude concentrations of NDMA) observed for the original procedures used in tandem. Authentic contaminated groundwaters are extracted using both the conventional and disk-based extraction procedures and analyzed; the observed NDMA concentrations are virtually identical over a target range spanning 100-10000 ng of NDMA/L. N-Nitrosodimethylamine (NDMA, CAS Registry No. 62-75-9) is a contaminant and an oxidative degradation product of unsymmetrical dimethylhydrazine (UDMH),1 a component of rocket fuel. Prior waste management and general processing procedures, now considered both outdated and ineffective, introduced both UDMH (1) Fine, D. H.; Rounbehler, D. P.; Rounbehler, A.; Silvergleid, A.; Sawicki, E.; Krost, K.; DeMarrais, G. A. Environ. Sci. Technol. 1977, 11, 581-584. S0003-2700(96)00157-6 CCC: $12.00
© 1996 American Chemical Society
and NDMA into the groundwater supply.2 NDMA is also a wellknown volatile carcinogen3 that has been detected in the atmosphere surrounding UDMH-producing facilities.1,4 For these reasons, there is an ongoing need to reduce the concentration of NDMA to acceptable, often ultratrace, concentrations in groundwaters near older, and frequently outdated, UDMH-producing facilities. As an example, the preliminary remediation goal for NDMA in groundwater near one such facility is 7.0 ng of NDMA/ L.5 In previous work,6 we demonstrated that NDMA can be quantified readily at the part-per-trillion (ng/L) level in groundwaters after they are extracted continuously overnight with dichloromethane.7 Briefly, a substantial portion of the resulting dichloromethane concentrate is then spiked onto a dual-sorbent carbon trap and dried before the residues (including NDMA) are desorbed onto a fused-silica gas chromatographic column. These residues are then separated in the usual manner; NDMA is detected selectively using a chemiluminescent nitrogen detector (CLND) optimized in its nitrosamine-selective mode. This procedure, while effective, exhibits a lengthy sample turnaround time (∼3 days per batch of eight samples, including a blank and appropriate QA/QC sample) and generates a substantial volume (∼300 mL per 1-L aqueous sample) of dichloromethane waste. For these reasons, it was most desirable to develop alternate procedures that would be effective and rapid and generate smaller volumes of dichloromethane waste.8 (2) Mohrman, G. B., Program Manager for Rocky Mountain Arsenal, Commerce, CO, private communication, 1994. (3) Magee, P. N. In The Experimental Basis for the Role of Nitroso Compounds in Human Cancer; Forman, D., Shuker, D. E. G., Eds.; Cancer Surveys 8; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1989; pp 208-239. (4) Fine, D. H.; Rounbehler, D. P.; Pellizzari, E. D.; Bunch, J. E.; Berkley, R. W.; McCrae, J.; Bursey, J. T.; Sawicki, E.; Krost, K.; DeMarrais, G. A. Bull. Environ. Contamin. Toxicol. 1976, 15, 739-746. (5) Record of Decision for the On-Post Operable Unit: Volume 1, Sections 1-11, Version 2.0; Foster Wheeler Environmental Corp., prepared for U.S. Army Program Manager’s Office for the Rocky Mountain Arsenal, March 1996; Table 4.0-1. (6) Tomkins, B. A.; Griest, W. H.; Higgins, C. E. Anal. Chem. 1995, 67, 43874395. (7) Method 3520A, Continuous Liquid-Liquid Extraction. Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, SW-846, 3rd ed.; Final Update 1; U.S. Environmental Protection Agency: Washington, DC, July 1992; Revision 1. (8) Ho, J. S.; Tang, P. S.; Eichelberger, J. W.; Budde, W. L. J. Chromatogr. Sci. 1995, 33, 1-8.
Analytical Chemistry, Vol. 68, No. 15, August 1, 1996 2533
Both Taguchi et al.9,10 and Jenkins et al.11 have described extraction procedures for NDMA in which a small mass (200 mg) of Ambersorb 572, a carbonaceous sorbent, was rolled for 1 h with an aqueous sample (500 mL) fortified with NDMA-d6 (internal standard). The sorbent is then filtered, thoroughly dried, and transferred to a 2-mL autosampler vial, where NDMA is extracted with a small volume (400 µL) of dichloromethane. The analyte is quantified by gas chromatography/high-resolution mass spectrometry using the method of internal standards. The authors calculated a detection limit of 1.0 ng of NDMA/L; recoveries, based on water samples fortified to 10.0 ng of NDMA/L, ranged between 26 and 41%. In our experience, similar recoveries were obtained using 870-mg portions of Ambersorb 572 and a 1-L water sample when the latter was stirred with a magnetic stirrer, rather than rolled, for 1 h and quantified using the method of external standards. We also observed some fragmentation of the sorbent, which is apparently quite fragile, in as little as 30 min. In addition, we noted substantial difficulty in obtaining facile quantitative transfer of the carbonaceous sorbent to the filtration apparatus, and thence to the 2-mL autosampler vial. Longer stirring times did not improve extraction recoveries and usually resulted in an extensive breakdown of the sorbent. Approximately half the Ambersorb 572 mass was extracted with 1 mL of dichloromethane per trial; the remaining half was set aside as an archival sample. These concerns were addressed initially by extracting 1-L water samples fortified to either 10 or 40 ng of NDMA/L through small glass solid-phase extraction columns packed with either 500 or 1000 mg of Ambersorb 572. In general, the sorbent was obviously much easier to handle in this packed format, and the extraction recoveries observed consistently exceeded 50%. Typically, the columns containing 500 mg of sorbent outperformed those containing 1000 mg of sorbent because the former was easier to extract with a small volume (not greater than 2 mL) of dichloromethane. However, the aqueous sample must pass through the sorbent bed at flow rates slower than 5-10 mL/min. For this reason, a single extraction would still be time consuming, requiring 3-4 h. Further, some provision must be made for removing high concentrations of potentially interfering neutral species present at some sites, particularly diisopropylmethylphosphonate (DIMP), prior to the extraction with the carbon sorbent column. All of these questions were addressed elegantly by using a carbon-based membrane extraction disk for removing NDMA from aqueous samples.12 The procedure described herein employs a preconditioned C18 membrane extraction disk, laid over the carbon-based disk, to remove the unwanted neutral waterinsoluble species. The former is set aside; the latter is dried and extracted with a small volume of dichloromethane. NDMA may then be quantified using the various procedures described in ref 6. Both the extraction time for a 1-L aqueous sample and the volume of waste dichloromethane are reduced ∼100-fold when (9) Taguchi, V. Y.; Jenkins, S. W. D.; Wang, D. T.; Palmentier, J.-P. F. P.; Reiner, E. J. Can. J. Appl. Spectrosc. 1994, 39, 87-93. (10) Taguchi, V. The Determination of N-Nitrosodimethylamine (NDMA) in Water by Gas Chromatography-High Resolution Mass Spectrometry (GC-HRMS); Environment Ontario (Canada), Laboratory Services Branch, Quality Management Office: Ontario, 1994; Method catalogue code NDMA-E3291A (originally approved February 19, 1993; revised and approved April 7, 1994). (11) Jenkins, S. W. D.; Koester, C. J.; Taguchi, V. Y.; Wang, D. T.; Palmentier, J.-P. F. P.; Hong, K. P. Environ. Sci. Pollut. Res. Int. 1995, 2, 207-210. (12) Markell, C. G., 3M Industrial and Consumer Sector, St. Paul, MN, private communication, 1995.
2534
Analytical Chemistry, Vol. 68, No. 15, August 1, 1996
the membrane disks are used instead of the continuous overnight extraction procedure. The reporting limit for the ultratrace procedure, 3 ng of NDMA/L of groundwater, was calculated using two statistically unbiased procedures.13,14 The certified range for this procedure is 2-40 ng of NDMA/L of groundwater. A related procedure, described herein, permitted unattended analysis of samples for NDMA concentrations exceeding 300 ng of NDMA/L of groundwater. The sample preparation is the same as that described above; however, 2-µL aliquots of dichloromethane are injected onto the gas chromatographic column using an automatic sampler. The use of the two procedures in tandem permits the quantification of NDMA concentrations ranging over ∼3 orders of magnitude, from nanogram-per-liter to microgram-per-liter levels. EXPERIMENTAL SECTION Reagents, Standards, and Samples. Sample Collection and Storage. Groundwater samples should be collected in commercially available, precleaned 1-L narrow-neck amber glass bottles, stored at 4 °C, and extracted within 7 days of sampling.13 NDMA Standard Solutions. Stock solutions with certified concentrations of NDMA in methanol may be purchased in a variety of concentrations from several vendors. These include NSI Environmental Solutions, Inc. (Research Triangle Park, NC; 5000 µg of NDMA/mL), AccuStandard (New Haven, CT; 1000 of µg of NDMA/mL), Polysciences (Niles, IL; 1000 µg of NDMA/ mL), and Ultra Scientific (North Kingstown, RI; 100 µg of NDMA/ mL). Aliquots are diluted sequentially to final concentrations ranging between 2 and 240 ng of NDMA/mL in highest purity dichloromethane (GC2 grade, Burdick & Jackson, Muskegon, MI, or equivalent) to form calibration standards for the “low-level” method, which employs the short-path thermal desorber. Calibration standards ranging between 50 and 4000 ng of NDMA/mL dichloromethane, prepared similarly, are required for the “highlevel” method, which employs the automatic sampler. Fresh standards should be prepared every 60 days. Caution: NDMA is a listed regulated OSHA carcinogen. Current regulations stipulate that neat NDMA must be dispensed only by individuals with specific training, and then only in contained areas. Dilute solutions of NDMA may be handled with the normal care and caution due any hazardous chemical. NDMA is a listed waste (No. P082) on the U.S. Environmental Protection Agency’s “P” list.15 Consult the local Environmental Protection Officer for guidance concerning the disposal of outdated standards. Preparation of Blanks and Spiked Samples for Method Certification and Quality Assurance. Aliquots (1 L) of synthetic groundwater are prepared from HPLC-grade water (Burdick & Jackson, J. T. Baker (Phillipsburg, NJ), or equivalent) which is fortified to 100 mg/L each in sulfate and chloride, as described elsewhere.13 These are spiked to concentrations ranging between 2 and 40 ng (13) Program Manager for Rocky Mountain Arsenal. Chemical Quality Assurance Plan, Version 1; U.S. Government Printing Office: Washington, DC, September 1993. (14) Appendix B to Part 136: Definition and Procedure for the Determination of the Method Detection Limit, Revision 1.11. Code of Federal Regulations: Protection of the Environment, Parts 100-149; Title 40; U.S. GPO: Washington, DC, Revised July 1, 1990. (15) Part 261: Identification and Listing of Hazardous Waste, and Paragraph 261.33, Discarded Commercial Chemical Products, Off-Specification Species, Container Residues, and Spill Residues Thereof. Code of Federal Regulations: Protection of the Environment, Parts 260-299; Title 40; U.S. GPO: Washington, DC, Revised July 1, 1994.
of NDMA/L using a master spiking solution containing 0.1 µg of NDMA/mL of water for the low-level procedure. Spiked samples for the high-level procedure are prepared from a master spiking solution containing 10 µg of NDMA/mL water; the final concentrations range between 100 and 4000 ng of NDMA/L. Water produced from typical “on-demand” systems may not be used, as described in ref 6. “Check Standards”. Check standards should be prepared for both the low-level and high-level procedures to evaluate the stability of the instrument and calibration curve. The concentration of such a standard, prepared in highest purity dichloromethane, should be near the midpoint of the calibration curve. A check standard should be prepared from a stock solution different from that used to prepare the calibration standards. It is strongly preferred that check standards be prepared on a staggered schedule compared to that of the calibration standards, to enable evaluation of possible degradation and decomposition of the latter. Apparatus. Sample Preparation and Extraction. Aqueous samples are filtered through Empore C18 filter disks (47-mm diameter, J. T. Baker or Varian Sample Products Division (Harbor City, CA)) layered over a carbon-based Empore disk (3M, New Products Department, 3M Industrial and Consumer Sector, St. Paul, MN; Part No. 98-0405-0047-6).16 An all-glass funnel/support assembly (DELTAWARE, with a 1-L filtering flask and a 300-mL reservoir, handles 47-mm-diameter disks, VWR, Part No. KT 93825-47 or equivalent) is used to support the disk and hold the filtrate. The two disks are preconditioned simultaneously with two 10-mL portions each of methanol and water. Each aliquot remains on the disk for 1 min prior to removal. The disk must not be allowed to become dry during the conditioning sequence. A separate all-glass funnel/support assembly should be arranged for drying the carbon-based disks. Here, the 1-L flask is replaced with a custom-prepared receiver cup fashioned from an “inner” standard taper 40/35 glass joint; dimensions are 37 mm o.d., 60 mm high. These receiver cups are large enough to accommodate precleaned, tared 20-mL screw cap vials. Desorption Equipment. The sorbent trap contains 125 mg each of Carbotrap and Carbotrap C (Supelco, Inc., Bellefonte, PA), which are packed and conditioned as described in ref 6. The short-path thermal desorber (SPTD), Model TD-1 (Scientific Instrument Services, Ringoes, NJ), is installed and arranged as described in ref 6. Sample residues are first dried at room temperature using ultrapure (99.9999%) helium for 2 min at 2530 mL/min and then desorbed for 5 min at 150 °C using ultrapure helium (3 mL/min). Gas Chromatograph, Automatic Sampler, Chemiluminescent Nitrogen Detector (CLND), and Data System. An Antek Model 705D pyrochemiluminescent nitrogen detector was retrofitted to a gas chromatograph (described below), with details given in ref 6. The detector was normally operated in its nitrosamine-selective mode. The factory-optimized parameters were as follows: furnace temperature, 250 °C; photomultiplier voltage and gain, 750 V and “high”; photomultiplier attenuation and output gain, 8 and 50; pyro oxygen flow rate, 30 mL/min; ozone oxygen flow rate, 5 mL/ min. (16) Physical parameters of the disks: carbon type, acid-washed coconut charcoal; surface area, >1000 m2/g; nominal particle diameter, 15-20 µm; weight of carbon in the disk, ∼90% w/w of the disk (about 400-450 mg); pore size distribution is proprietary.
A Hewlett-Packard Model 5890 Series II gas chromatograph equipped with a split/splitless injector both supported and was physically interfaced to the SPTD (low-level procedure), CLND, data system, and automatic sampler (high-level procedure). One of the existing data interface boards was replaced with a HewlettPackard analog input board so that the output from the CLND could become part of the overall system network. The gas chromatographic parameters and conditions were optimized as follows: injector, equipped with “double gooseneck” liner, 150 °C; injector purge valve OFF at 0.00 min and ON at 5.00 min; guard column, 0.53 mm i.d. × 5 m, deactivated and uncoated fused-silica column (Restek Corp., Bellefonte, PA) connected to the analytical column using a Universal Press-Tight connector (Restek); analytical fused-silica capillary column, Rtx-200 (crossbond trifluoropropylmethyl), 3.00 µm film thickness, 0.53 mm i.d. × 30 m (Restek) (Note: The polyimide coating must be removed from ∼25 mm of the outlet end of the column, where the column enters the CLND furnace. This is achieved conveniently using either a match or a cigarette lighter.); carrier gas, ultrahigh-purity helium, 4 mL/min (13.0 cm/s); column oven temperature, increased linearly from 35 °C (hold for 5 min) to 175 °C (hold for 5 min) at 6 °C/min. When the high-level procedure was used, a Hewlett-Packard Model 7673 automatic sampler replaced the short-path thermal desorber and injected 2 µL on-column. The gas chromatographic operating parameters are the same as those described above with the exception of the purge valve settings, which become OFF at 0.00 min and ON at 0.25 min. The Hewlett-Packard Model 3365 ChemStation software package loaded on an HP 486/50 personal computer or equivalent permitted operation of both the gas chromatograph and the automatic sampler (“high level” procedure) as well as facile collection, storage, reporting, and manipulation of the analytical data. Procedure. Aqueous samples are filtered simultaneously through C18 and carbon-based Empore disks under vacuum at ∼40-50 mL/min (a 1-L sample such as drinking water or filtered groundwater, which contains little or no particulate matter, was typically finished in ∼20 min). The extraction apparatus was then dismantled; the C18 disk was set aside or may be used for related independent analyses. The carbon-based Empore disk was transferred to the second all-glass filter support, where the receiver cup replaced the normal 1-L filtration flask, and was dried for 10 min under vacuum. When the disk was thoroughly dry, a precleaned, tared 20-mL screwcap vial was placed in the receiver cup, which was then reattached to the all-glass filter support. A 3-mL aliquot of highest purity dichloromethane was added to the reservoir. After the solvent had remained over the carbon-based disk for 1 min, it was drawn through the disk under vacuum (slowly, if at all possible); ∼2 mL was collected in the screw cap vial. The vial was then reweighed to determine the mass (and later, the accurate volume) of the dichloromethane extract. The determination of NDMA at ultratrace concentrations (between 2 and 40 ng of NDMA/L of water) proceeded in the manner described in ref 6. A 100-µL aliquot of the dichloromethane concentrate was added slowly to the Carbotrap C end of the dual-sorbent trap, allowing sufficient time for the liquid to soak into the trap bed. The trap was then mounted into the shortpath thermal desorber (Carbotrap C end was connected to the “top tube” without the needle attached) and dried with ultrahighpurity helium for 2 min at 25-30 mL/min. This procedure was Analytical Chemistry, Vol. 68, No. 15, August 1, 1996
2535
repeated, leaving the residues from 200 µL of concentrate on the trap. The trap was then disconnected from the top tube and inverted. The needle was connected to the Carbotrap C end of the trap, and the remaining Carbotrap end was screwed into the top tube. The residues present on the trap were then desorbed directly from the sorbent trap onto the guard and analytical gas chromatographic columns, both maintained at 35 °C, using a 150 °C desorption temperature maintained for 5 min and a helium desorption flow rate of 3 mL/min. The column oven temperature program commenced immediately after completion of desorption. The extraction procedure for high levels of NDMA, typically exceeding 300 ng of NDMA/L water, was the same as that described above. Upon completion of the carbon disk extraction, the dichloromethane extract was transferred to a 2-mL amber autosampler screw cap vial equipped with a Teflon/silicone/Teflon septum (National Scientific Co., Lawrenceville, GA). A 2-µL aliquot was injected onto the column using the automatic sampler and analyzed using the gas chromatographic conditions described above. Preliminary Sample Stability Study. Synthetic groundwaters fortified with 1000 ng of NDMA/L were extracted as described above. The carbon-based membrane disks were then dried and stored in the dark, either at ambient temperature or at 4 °C, for 6 days. NDMA was then removed from the disks using dichloromethane and quantified as described, and the resulting concentrations were compared with those observed at time 0.
RESULTS AND DISCUSSION Method Optimization. Extraction kinetics and mass transfer determine the maximum flow rate of a sample through any given solid-phase extraction medium. For bulk Ambersorb 572, which was supplied as 20-50 mesh (300-850 µm) beads, the recoveries observed for flow rates not faster than 5-10 mL/min were clearly superior to those of samples which had been extracted at ∼3040 mL/min, and even slower flow rates would have been preferred. The carbon-based Empore disks, by contrast, contain acid-washed coconut charcoal particles with typical particle diameters ranging between 15 and 20 µm, or 20-40 times smaller than those of bulk Ambersorb 572. It can be shown12 that the analyte recovery is nearly independent of flow rate if the solid-phase extraction medium is chosen properly (i.e., appropriate partitioning between sample and extraction media) and consists of small particles. For this reason, it was possible to recover NDMA, which is highly soluble in water and is therefore difficult to extract in a reasonable yield with rapid flow rates. Overall, a 1-L groundwater sample could be extracted in ∼20 min, or ∼50-fold faster than the 18-h traditional overnight continuous extraction. It was critical to ensure that the carbon-based medium was completely dry prior to analyte elution with dichloromethane. The reasons are twofold. First, because NDMA is so soluble in water, the latter competes successfully, even with small quantities present, with dichloromethane for available analyte. Second, because dichloromethane and water are immiscible, NDMA extraction from a solid medium is inhibited if the latter is wet. When small columns packed with Ambersorb 572 were employed, they were dried to constant weightsa process which frequently required 1-2 hsbefore eluting NDMA with dichloromethane. The same formal procedure could also be used to determine whether a carbon-based Empore disk is ready for elution. In practice, 2536
Analytical Chemistry, Vol. 68, No. 15, August 1, 1996
however, a 10-min drying period renders the disk reproducibly dry and ready for elution. Because NDMA is exceptionally soluble in water, it is difficult to obtain recoveries exceeding ∼60% using either the traditional procedure, featuring overnight continuous extraction with dichloromethane or a solid-phase extraction procedure, featuring the carbon-based Empore disk. The recovery of NDMA was virtually constant if a single 2-, 3-, or 4-mL aliquot of dichloromethane was used to extract the disk. Thus, merely increasing the elution volume was ineffective in extracting additional NDMA. If a second 3-mL aliquot of dichloromethane was used to extract the disk, an additional 15% of NDMA was recovered. The reason for this observation is not clear. These data had to be balanced against the need for small elution volumes (to improve sensitivity) and minimal “blow-down” of the extract (to reduce losses of volatile NDMA). Overall, the optimal extraction procedure was to elute the disk once with 3 mL of dichloromethane (nominally, 2 mL was recovered). This volume is a factor of 100 smaller than that normally employed for continuous overnight extraction. Method Certification and the Statistical Determination of the Reporting Limit Using the Short-Path Thermal Desorber as the Injector. The reporting limit for this procedure was evaluated initially using the protocol established by the U.S. Army13 for calculating the method reporting limit (MRL) and discussed in detail in ref 6. This procedure requires a target reporting limit (TRL), which was set at 2 ng of NDMA/L (believed to be achievable), or ∼3 times the health-based criterion of 0.7 ng NDMA/L. The first part of the protocol evaluated the detector response and calibration curve expected. Briefly, a calibration curve was prepared on two consecutive days from standards ranging in concentration from 2 to 240 ng of NDMA/mL dichloromethane. In all cases, a 50-µL aliquot was spiked onto the sorbent trap and dried, and the residues were thermally desorbed as described above. The resulting data represented the range of 100-12000 pg of NDMA detected by the CLND. A spreadsheet statistical program, available from the U.S. Army, calculates lack of fit based on a 95% significance level for (a) a linear model with zero intercept, (b) a linear model with nonzero intercept, and (c) an intercept statistically different from zero. Taken together, the statistical tests demonstrated that a linear model with zero intercept was appropriate for the calibration data. The second part of the protocol evaluated the analytical methodology itself and calculated the MRL.17 Two sets of spiked synthetic groundwater samples, analyzed on consecutive days, reflecting concentrations of NDMA ranging between 1 × TRL and 20 × TRL (i.e., 2-40 ng of NDMA/L) were subjected to the extraction procedure. Aliquots (200 µL) of the final dichloromethane concentrate were processed using short-path thermal desorption followed by gas chromatography and detection of NDMA using the CLND. The resulting integrated peak areas were converted to mass of NDMA in picograms using the calibration data to determine the “found” mass of NDMA in each case. These “found” masses of NDMA were compared to the “target” masses of NDMA, which must be calculated knowing the following data: (a) the aqueous sample volume (fixed at 1 L); (b) C, the concentration of the aqueous sample, in ng/L; (c) w, the mass of the dichloromethane extract, in g; (d) the volume of the extract spiked onto the trap (fixed at 200 µL or 0.2 mL); (e) ρ, (17) Tomkins, B. A.; Merriweather, R.; Jenkins, R. A.; Bayne, C. K. J. Assoc. Off. Anal. Chem. Int. 1992, 75, 1091-1099.
Table 1. Summary of Target and Found NDMA Masses on the Two Days of Certification Using the Short-Path Thermal Desorber as the Injectora target NDMA concn (ng/L)
certification day
0
1
0
70
2
1
250
350
a
target mass of NDMA (pg)
found mass of NDMA (pg)
4
1
490
490
10
1
1100
790
20
1
2500
1700
40
1
4900
2900
0
2
0
100
2
2
240
300
4
2
470
430
10
2
1200
800
20
2
2400
1500
40
2
4800
2800
Table 2. Calculation of the Method Detection Limit Using Replicate Water Samples Containing 10 ng of NDMA/L (Short-Path Thermal Desorber Used as the Injector) replicate no.
calcd NDMA concn (ng/L)a
1 2 3 4 5 6 7 8
7.4 7.0 6.8 6.6 8.6 7.0 7.0 7.6
standard deviation, all samples MDL, all samples,b ng/L standard deviation, “best seven” samples MDL, “best seven” samples,c ng/L
0.6 2 0.3 1
a Values not corrected for recovery. b Student’s t value (one-tailed) at 99% confidence for eight samples (seven degrees of freedom) ) 2.988. c Student’s t value (one-tailed) at 99% confidence for seven samples (six degrees of freedom) ) 3.143.
Values not corrected for recovery.
the density of dichloromethane, 1.326 g/mL; and (f) a units conversion, 1000 pg/ng. The “target” mass is then calculated using
(C/w)ρ(0.2)(1000) ) target mass of NDMA, in pg
(1)
Table 1 summarizes the “found” versus the “target” masses for the two days of method certification. All of these data were then entered into the second part of the spreadsheet program discussed above. The spreadsheet calculates the MRL using a 90% confidence interval, in which the likelihoods of a false positive error (R) or a false negative error (β) are both set to 5%. The MRL so obtained, 408 pg of NDMA, must be converted back to an aqueous concentration by knowing or assuming the following data: (a) volume of extract analyzed, 0.2 mL; (b) total extract volume, nominally 1.6 mL; and (c) the volume of an aqueous sample, 1 L:
408 pg of NDMA 1.6 mL ng ) 3.3 ng of NDMA/L 0.2 mL 1 L 1000 pg The agreement between the MRL calculated in the present work, ∼3 ng of NDMA/L of groundwater, and that reported in ref 6, 2 ng of NDMA/L of groundwater, is excellent and demonstrates that the disk extraction procedure may be substituted for the traditional continuous extraction procedure employing dichloromethane. The overall NDMA recovery, 57%, is slightly better than that observed in the earlier method, which was 50%. The detection limit was also determined using the method detection limit (MDL) defined by the U.S. Environmental Protection Agency.14 Briefly, eight (minimum of seven required) 1-L synthetic groundwater samples were fortified to 5 × TRL (i.e., 10 ng NDMA/L) and analyzed as described above. The standard deviation, which was calculated for both seven and eight replicates, was then multiplied by the appropriate one-tailed Student’s t statistic for 99% confidence. The resulting value was the MDL, ∼2 ng of NDMA/L when all eight experimental data points were
employed, as shown in Table 2. When the “best seven” values were employed (that is, those with the smallest spread), the MDL was reduced to 1 ng of NDMA/L of groundwater. These calculations also satisfied an additional criterion given in ref 13, viz., the magnitude of the MDL should be equal to or less than that of the TRL. Method Certification and the Statistical Determination of the Reporting Limit Using the Automatic Sampler as the Injector. To evaluate the performance of the high-level procedure for determining NDMA in heavily contaminated groundwaters, the short-path thermal desorber was removed and replaced with the HP 7673 automatic sampler, which was programmed to inject 2-µL volumes of either standards or extract concentrates on column. The evaluation of the calibration curve, using the lack of fit and zero intercept tests described above, required testing standards containing 50-4000 ng of NDMA/mL over two consecutive days. Once again, a linear calibration curve was obtained. The statistical evaluation showed no significant lack of fit when using a linear model either with or without a nonzero intercept, and the intercept of the calibration curve was not significantly different from zero. The MRL was then evaluated as described previously, using synthetic groundwater samples that had been fortified to 1004000 ng of NDMA/L, or 1-40 times the target reporting limit (TRL) of 100 ng of NDMA/L. The “found” NDMA concentrations were obtained using the calibration data evaluated previously and the appropriate measured integrated peak areas. The “target” NDMA concentration expected in the extract must be calculated knowing (a) C, the NDMA concentration in the spiked synthetic groundwater sample, in ng/L; (b) the volume of the spiked synthetic groundwater sample (1 L); (c) w, the weight of the dichloromethane extract, in g; and (d) ρ, the density of dichloromethane, 1.326 g/mL:
(C/w)ρ ) target NDMA concentration, in ng/mL (2) Table 3 summarizes the “target” and “found” data obtained from the two days of method certification. When all of the data values Analytical Chemistry, Vol. 68, No. 15, August 1, 1996
2537
Table 3. Comparison between the Target and Found Concentrations of NDMA When the Automatic Sampler Is Used for Sample Injection target NDMA concn (ng/L)
certification day
target NDMA concn (ng/mL)
found NDMA concn (ng/mL)
0 100 200 500 1000 2000 4000 0 100 200 500 1000 2000 4000
1 1 1 1 1 1 1 2 2 2 2 2 2 2
0 50 120 280 530 1200 2100 0 60 120 280 550 1100 2200
10 40 90 170 350 750 1400 10 40 90 150 300 730 1500
a
Values not corrected for recovery.
were used, the spreadsheet program returned an extract concentration of 189 ng of NDMA/mL of dichloromethane. This value can be converted into an aqueous NDMA sample concentration by assuming an aqueous sample volume (1 L) and a typical extract volume (1.8 mL):
Table 4. Calculation of the Method Detection Limit Using Replicate Spiked Synthetic Groundwater Samples Containing 1000 ng of NDMA/L (Automatic Sampler Used as the Injector) replicate no.
calcd sample concn (ng/L)a
1 2 3 4 5 6 7 8
710 680 660 710 680 740 560 750
sample standard deviation, all values (ng/L) MDL, all eight samplesb (ng/L) sample standard deviation, “best seven” values (ng/L) MDL, “best seven” valuesc (ng/L)
60 180 30 100
a Values not corrected for recovery. b Student’s t value (one-tailed) at 99% confidence for eight samples (seven degrees of freedom) ) 2.998. c Student’s t value (one-tailed) at 99% confidence for seven samples (six degrees of freedom) ) 3.143.
Table 5. Determination of NDMA in Drinking Water and Contaminated Groundwater Samples: Comparison with Conventional Extractions and Historical Data NDMA concn (ng/L)a sampling period
disk extraction
groundwater A
May 1995
groundwater B groundwater C groundwater D
May 1995 May 1995 May 1995
600 200 6400 200 600 800 230 210 220
