6610 CARBAMATE PESTICIDES*

6610 A. Introduction

1. Sources and Significance waters: aldicarb sulfoxide, aldicarb sulfone, oxamyl, methomyl,
3-hydroxycarbofuran, aldicarb, propoxur, carbofuran, carbaryl,
Carbamates are used as insecticides, nematicides, and aracides 1-naphthol, and methiocarb.
to control pests on agricultural crops, as well as to control lawn
and garden insects. Their toxicity comes from their ability to act 3. References
as cholinesterase inhibitors. Residues of several carbamates have
been found in groundwater in a number of states.1,2 Two of the 1. HOWARD, P.H., ed. 1989. Handbook of Environmental Fate and
target compounds in this method, carbofuran and oxamyl, are Exposure Data for Organic Chemicals. Lewis Publishers, Chelsea,
regulated by the U.S. Environmental Protection Agency, with Mich.
aldicarb and its metabolites, aldicarb sulfoxide and aldicarb 2. JONES, R.L. & T.L. ESTES. 1995. Summary of aldicarb monitoring and
research programs in the U.S.A. J. Cont. Hydr. 18:107.
sulfone, under consideration for regulation.3
3. U.S. ENVIRONMENTAL PROTECTION AGENCY. 2000. Drinking Water Stan-
dards and Health Advisories. EPA 822-B-00-001, U.S. Environmental
2. Selection of Method Protection Agency, Off. Water, Washington, D.C.
4. MOYE, H.A., S.J. SCHERRER & P.A. ST. JOHN. 1977. Dynamic labeling
High-performance liquid chromatography (HPLC) is the of pesticides for high performance liquid chromatography: Detection
method of choice for analysis for carbamates, many of which are of N-methylcarbamates and o-phthalaldehyde. Anal. Lett. 10:1049.
thermally labile. This HPLC analysis method is an updated 5. FOERST, D.L. & H.A. MOYE. 1984. Aldicarb and related compounds
version of a previous method developed using direct injection, in drinking water via direct aqueous injection HPLC with post
post-column derivatization, and fluorescence detection to yield column derivatization. In Advances in Water Analysis and Treat-
sensitivity and selectivity while keeping sample preparation to a ment, Proc. 12th Annu. AWWA Water Quality Tech. Conf., Denver,
Colo., Dec. 2-5, 1984, p. 189. American Water Works Association,
minimum.4 – 6 This method is suitable for the analysis of the
Denver, Colo.
following carbamate compounds and metabolites in drinking 6. U.S. ENVIRONMENTAL PROTECTION AGENCY. 2001. Method 531.2—Mea-
surement of N-methylcarbamoyloximes and N-methylcarbamates in wa-
* Approved by Standard Methods Committee, 2004.
ter by direct aqueous injection HPLC with postcolumn derivatization.
Joint Task Group: Steven C. Wendelken (chair), Margarita V. Bassett, Linda EPA 815-B-01-002, U.S. Environmental Protection Agency, Off.
Henry, David J. Munch, Barry V. Pepich. Ground Water and Drinking Water, Cincinnati, Ohio.

6610 B. High-Performance Liquid Chromatographic Method

1. General Discussion TABLE 6610:I. DETECTION LEVELS IN REAGENT WATER

Fortification Detection
a. Principle: After the addition of a surrogate compound and Level Level* Signal-to-Noise
filtration, water samples are injected directly onto a HPLC and Analyte ␮g/L ␮g/L Ratio
separated by use of a gradient and a C18 column. The carbamates
analyzable by this method (Table 6610:I), which are generally Aldicarb sulfoxide 0.20 0.059 8:1
classified as phenyl and oxime carbamates, all have an N-methyl Aldicarb sulfone 0.10 0.057 3:1
Oxamyl 0.20 0.065 10:1
group in common. After chromatographic separation, these com- Methomyl 0.20 0.050 10:1
pounds are hydrolyzed with 0.05N NaOH at 80 to 95°C, yielding 3-Hydroxycarbofuran 0.20 0.029 18:1
a methyl amine which is then reacted with o-phthalaldehyde Aldicarb 0.20 0.026 9:1
(OPA) and 2-mercaptoethanol (or N,N-dimethyl-2-mercaptoeth- Propoxur 0.20 0.037 6:1
ylamine) to form a highly fluorescent isoindole that is detected Carbofuran 0.20 0.043 9:1
instrumentally. The external standard technique is used for quan- Carbaryl 0.20 0.045 13:1
titation. 1-Naphthol 0.20 0.063 10:1
b. Interferences: Method interferences may be caused by Methiocarb 0.20 0.061 11:1
BDMC (SUR) 2.00 N/A N/A
contaminants, especially primary amines and ammonia, in sol-
vents, reagents (including reagent water), sample bottles and * Detection levels were determined by analyzing 7 replicates over 3 d under the
conditions outlined in Table 6610:IV with a 1000-␮L injection.
caps, and other sample-processing hardware, that lead to discrete
N/A ⫽ not applicable
artifacts and/or elevated baselines in the chromatograms. The
samples or analytical system may be contaminated from being

S-1
S-2 INDIVIDUAL ORGANIC COMPOUNDS (6000)

TABLE 6610:II. SINGLE-ANALYST PRECISION AND ACCURACY OF COMPOUND DETECTION IN VARIOUS WATERS AT LOW (0.20 ␮G/L) AND HIGH (10 ␮G/L)
FORTIFICATION LEVELS*
Reagent Water Drinking Water, Surface Water Drinking Water, Groundwater
0.20 ␮g/L 10.0 ␮g/L 0.20 ␮g/L 10.0 ␮g/L 0.20 ␮g/L 10.0 ␮g/L
Compound MR% %RSD MR% %RSD MR% %RSD MR% %RSD MR% %RSD MR% %RSD

Aldicarb sulfoxide 112 6.2 106 1.8 113 7.0 104 2.8 111 7.3 106 1.1
Aldicarb sulfone 92 9.5 106 2.6 104 5.5 106 1.4 98 9.2 106 0.9
Oxamyl 101 8.6 106 2.2 107 6.4 104 2.2 99 8.4 105 1.2
Methomyl 101 6.5 106 2.9 110 9.8 104 1.6 99 10.2 105 1.4
3-Hydroxycarbofuran 105 6.8 108 1.2 128 3.9 107 1.1 107 3.0 108 0.4
Aldicarb 95 7.4 106 1.3 123 2.7 105 1.5 100 6.3 105 0.6
Propoxur 109 5.9 109 2.0 128 6.0 106 2.1 112 6.1 107 0.8
Carbofuran 112 6.7 110 2.2 140 5.6 105 2.5 112 4.1 107 1.6
Carbaryl 112 7.0 107 2.1 112 9.7 106 0.9 119 5.1 108 1.3
1-Naphthol 113 12.6 108 3.1 113 12.1 101 1.3 109 8.2 109 1.2
Methiocarb 105 5.9 107 1.5 104 13.3 107 1.1 105 3.9 107 1.0
BDMC (SUR)† 108 4.3 101 2.3 108 2.1 96 3.9 109 2.0 97 4.3
* MR% ⫽ mean recovery expressed as % recovery; %RSD ⫽ percent relative standard deviation
† Surrogate concentration in all samples was 2.0 ␮g/L; all data from n⫽7 replicates using a 1000-␮L injection volume.

handled with bare fingers. Routinely demonstrate that all items shipping sample bottles to the field, add preservatives, as dry
used in analysis are free from interferences under the conditions solids, to each bottle. To adjust sample pH to ⬃3.8 to prevent
of the analysis by analyzing laboratory reagent blanks. Do not hydrolysis of oxamyl, 3-hydroxycarbofuran, carbaryl, and me-
subtract blank values from sample results. Clean all glassware thiocarb and to guard against biodegradation, add a sufficient
meticulously: wash glassware with detergent and tap water, rinse amount of potassium dihydrogen citrate (C6H7KO7) to yield a
with tap water, and rinse again with reagent water. A final rinse concentration in the sample of 9.2 to 9.5 g/L. To eliminate the
with solvents may be needed. In place of a solvent rinse, non- residual free chlorine in the samples, which rapidly degrades
volumetric glassware can be heated in a muffle furnace at 400°C aldicarb and methiocarb, add sodium thiosulfate (Na2S2O3) to
for 2 h. Do not heat volumetric glassware above 120°C. yield a sample concentration in the range of 80 to 320 mg/L.
Samples that are not properly preserved (¶ 2 below) may yield b. Sample collection: Collect grab samples in accordance with
poor target analyte recovery due to degradation caused by chlorine conventional sampling practices. Do not prerinse sample bottles
residual and base-catalyzed hydrolysis at neutral and high pH. with sample before collection, because prerinsing will wash out
c. Detection levels: Detection levels are compound, instru- the preservatives added to the bottles before shipment.
ment, and matrix dependent. The detection level is defined as the When sampling from a cold water tap, remove aerator so that
statistically calculated minimum concentration that can be mea- no air bubbles will be trapped in the sample. Open tap and let
sured with 99% confidence that the reported value is greater than system flush until water temperature has stabilized (usually
zero.1 Experimentally determined detection levels for the target about 3 to 5 min). Collect samples from the flowing system.
analytes are provided in Table 6610:I. The detection level differs Fill sample bottles, taking care not to flush out sample pres-
from, and is lower than, the minimum reporting level (MRL). ervation reagents. Samples do not need to be collected head-
The concentration range for target analytes in this method was space-free. After collecting sample, cap carefully to avoid spill-
evaluated between 0.2 ␮g/L and 10 ␮g/L. Precision and bias data age and agitate by hand for 1 min. Keep samples sealed from
are presented in Table 6610:II. collection time until analysis.
d. Safety: The toxicity or carcinogenicity of each reagent used c. Sample storage and holding time: Keep all samples iced
in this method has not been precisely defined; treat each chem- during shipment and do not let temperature exceed 10°C during
ical compound as a potential health hazard, and minimize expo- the first 48 h after collection. Assuming that samples are in
sure. Handle pure standard materials and stock standards of these transit for 48 h or less, confirm that they are at or below 10°C
compounds with suitable protection to skin and eyes. Take care when they are received at the laboratory. In the laboratory, store
not to breathe the vapors or ingest the materials. Maintain a samples at or below 6°C and protect from light until analysis. Do
current-awareness file of OSHA regulations regarding the safe not freeze samples. Results of the sample storage stability study
handling of the chemicals specified in this method. Make a of all method analytes indicated that all compounds are stable for
reference file of MSDSs available to all personnel involved in 28 d in water samples that are collected, dechlorinated, pre-
the chemical analysis. Additional references to laboratory safety served, shipped, and stored as described above.5 Analyze sam-
are available.2– 4 ples within 28 d.

2. Sampling and Storage 3. Apparatus

a. Sample bottle preparation: Use amber glass bottles fitted a. High-performance liquid chromatograph (HPLC): A sys-
with polytetrafluoroethylene (PTFE)-lined screw caps. Before tem capable of reproducibly injecting up to 1000-␮L portions,
CARBAMATE PESTICIDES (6610)/High-Performance Liquid Chromatographic Method S-3

and performing ternary linear gradients at a constant flow rate of c. Acetonitrile, CH3CN, high-purity, demonstrated to be free
approximately 1.5 mL/min. A column heater is desirable. Data of analytes and interferences (HPLC grade or better).
included in this method were obtained with a heater set at 30°C. d. Methanol, CH3OH, high-purity, demonstrated to be free of
1) Analytical column: Any column that provides adequate analytes and interferences (HPLC grade or better).
resolution, peak shape, capacity, accuracy, and precision (¶ 6 e. Hydrolysis solution (postcolumn reagent 1): Make sodium
below) may be used. For development of this method, an HPLC hydroxide, NaOH, 0.05N, by diluting 4 mL 50% (w/w) sodium
“carbamate” column, 3.9 ⫻ 150 mm, packed with 4-␮m dp C18 hydroxide solution to 1 L with reagent water. Because hydrolysis
solid-phase particles* was used. solution concentration can dramatically affect analyte response,
2) Postcolumn reaction system capable of mixing reagents use extra care in preparing. Filter and degas (with helium or
into the mobile phase. Use a reactor constructed of polyether- other appropriate gas) just before use.
etherketone (PEEK) or PTFE tubing and equipped with two f. Sodium borate solution, 0.05N: Dissolve 19.1 g sodium
pumps capable of delivering up to 0.5 mL/min of each reagent; tetraborate decahydrate (Na2B4O7 䡠 10H2O) in a 1-L volumetric
mixing tees; and two reaction coils. Various postcolumn system flask. Bring volume up to 1.0 L with reagent water. The sodium
manufacturers recommend different reaction-coil temperatures borate will dissolve in less than 2 h if a stir bar is used. Filter and
for the carbamate hydrolysis reaction; therefore, the first reaction degas.
coil temperature may range from 80 to 95°C. Analyte signal can g. o-Phthalaldehyde (OPA): Dissolve 100 ⫾ 10 mg of OPA in
increase with temperature over this temperature range; however, 5 to 10 mL methanol.
baseline noise can also increase with increasing temperatures. h. OPA derivatization solution (postcolumn reagent 2): Pre-
The second reaction takes place at ambient temperature. pare by either 1) or 2) below. Both 2-mercaptoethanol and N,
3) Detector: Use a fluorescence detector capable of excitation N-dimethyl-2-mercaptoethylamine hydrochloride react with
at approximately 340 nm and detection of emission energy at OPA and the target methylamine to form the isoindole detected
approximately 465 nm. Specific optimum excitation and emis- by the fluorescence detector. Both reagents have characteristic
sion wavelengths may vary slightly for each system. strong odors. Prepare solutions in a fume hood. This reagent, as
4) Data system: Preferably use a computerized data system prepared by either method, is stable for at least 36 h. However,
for data acquisition and processing.† individual laboratory conditions vary and daily preparation of
b. Vials: Screw-cap or crimp-top glass autosampler vials with
this solution may be necessary.
PTFE-faced septa, amber or clear.
1) Preparation with 2-mercaptoethanol—Add dissolved OPA
c. Volumetric flasks, Class A, various sizes, used for prepara-
(¶ g above) to 1 L filtered and degassed sodium borate solution
tion of standards and samples.
(¶ f). Add 1.0 mL 2-mercaptoethanol and mix.
d. Microsyringes, various sizes.
2) Preparation with N,N-dimethyl-2-mercaptoethylamine hy-
e. Disposable syringes, 5- to 30-mL size, used to filter sample
drochloride—Dissolve 2.0 ⫾ 0.2 g N,N-dimethyl-2-mercapto-
extracts before analysis.‡
ethylamine hydrochloride in approximately 10 mL sodium bo-
f. Filters, disposable, used to filter samples before analysis.§
rate solution (¶ f above). Add dissolved OPA (¶ g) to 1 L filtered
g. Analytical balance, capable of weighing accurately to
0.0001 g. and degassed sodium borate solution. Add dissolved N,N-di-
methyl-2-mercaptoethylamine-hydrochloride to the sodium bo-
rate solution and mix.
4. Reagents i. Calibration standards: Prepare standard solutions either
from certified, commercially available solutions or from neat
Use only reagent-grade or better chemicals and HPLC-grade compounds. Compounds used to prepare solutions must be
solvents. Unless otherwise indicated, ensure that all reagents ⱖ 96% pure; their weights may be used without correction for
conform to the specifications of the Committee on Analytical purity to calculate the concentration of the stock standard. So-
Reagents of the American Chemical Society, where such spec- lution concentrations mentioned in this section were used to
ifications are available. develop this method and are included as an example, not a
a. Reagent water: Use purified water that does not contain requirement. Generally, prepare standards for sample fortifica-
measurable quantities of any target analytes or interfering com- tion in the smallest volume that can be measured accurately, to
pounds greater than 1⁄3 the MRL for each compound of interest. minimize the addition of organic solvent to aqueous samples.
b. Buffered reagent water: To reagent water, add sample NOTE: Although stability times for standard solutions are sug-
preservation reagents in the same concentrations present in the gested below, use standard QC practices to determine when
samples. To prepare 1 L buffered reagent water, add a sufficient standard solutions need to be replaced.
amount of potassium dihydrogen citrate to yield a concentration 1) Surrogate analyte (SUR) standard solution, 4-bromo-3,5-
of 9.2 to 9.5 g/L and sodium thiosulfate to yield a concentration dimethylphenyl n-methylcarbamate (BDMC):
in the range of 80 to 320 mg/L to a graduated bottle or volu- a) SUR stock standard: If preparing from neat material, ac-
metric flask. Fill to 1-L mark with reagent water. curately weigh approximately 25 to 35 mg neat SUR to nearest
0.1 mg into a tared, 5-mL volumetric flask. Dilute to the mark
with methanol. Stock solutions have been shown to be stable for
* Waters catalog No. WAT035577. 12 months when stored at ⫺10°C or less.
† A Waters Millennium software system was used to generate all data contained b) SUR primary dilution standard (PDS): Prepare by adding
in the tables provided in this method.
‡ B-D catalog No. 309603, 309650 or equivalent. enough of the SUR stock standard to a volumetric flask partially
§ Millipore 0.22-␮m PVDF membrane, catalog No. SLGV 013 NL, or equivalent. filled with methanol to make a final concentration of 10 ␮g/mL
S-4 INDIVIDUAL ORGANIC COMPOUNDS (6000)

TABLE 6610:III. PREPARATION OF CALIBRATION (CAL) CURVE SOLUTIONS

Analyte PDS Volume of Analyte Volume of 10-␮g/mL Final Volume of Final Conc. of CAL
Conc. PDS SUR PDS CAL Solution Solution
CAL Level ␮g/mL ␮L ␮L mL ␮g/L
1 1.0 5.0 5.0 25 0.20
2 1.0 12.5 5.0 25 0.50
3 1.0 25.0 5.0 25 1.00
4 10.0 5.0 5.0 25 2.00
5 10.0 12.5 5.0 25 5.00
6 10.0 25.0 5.0 25 10.0

when filled to the mark with methanol. The PDS has been shown resolution for closely eluting compounds that are not baseline
to be stable for 6 months when stored at ⫺10°C or less. resolved according to ¶ 6h.
2) Target analyte standard solution: b. Calibration:
a) Target analyte stock standard: Obtain analytes listed in 1) Initial calibration — Calibrate system by using the external
Table 6610:I as neat or solid standards or as commercially standard technique. Prepare a set of at least five calibration
prepared ampulized solutions. If preparing from neat material, standards as described in ¶ 4i3).
accurately weigh approximately 25 to 35 mg of pure material to
the nearest 0.1 mg into a tared, 5-mL volumetric flask. Dilute to
the mark with methanol. Repeat for each target analyte. TABLE 6610:IV. INSTRUMENT GRADIENT AND CONDITIONS
b) Target analyte primary dilution standard (PDS): Prepare Time
by dilution of the target analyte stock standards. Add enough of min % Water % Methanol % Acetonitrile
each target stock standard to a volumetric flask partially filled
with methanol to make a final concentration near 10 ␮g/mL Initial 88.0 12.0 0.0
5.30 88.0 12.0 0.0
when filled to the mark with methanol. A serial dilution of this
5.40 68.0 16.0 16.0
PDS, to make a 1.0-␮g/mL solution, is useful for low-level 14.00 68.0 16.0 16.0
fortification. The PDSs have been shown to be stable for 6 16.10 50.0 25.0 25.0
months when stored at ⫺10°C or less. 20.00 50.0 25.0 25.0
3) Calibration solutions (CAL): Prepare the initial calibration 22.00 88.0 12.0 0.0
curve with at least 5 calibration concentrations. Prepare working 30.00 88.0 12.0 0.0
calibration solutions over the concentration range of interest Instrument conditions:
from dilutions of the analyte PDSs in buffered reagent water. HPLC: A ternary gradient of water, methanol, and acetonitrile with a flow of 1.5
Add SUR PDS to each CAL. Filter CAL solutions in same mL/min as shown above.
manner as samples (¶ 5c). The lowest concentration of calibra- Injection volume: System-dependent; for the development of this method: 250 to
tion solution must be at or below the MRL, which may depend 1000 ␮L.
Column: See ¶ 3a1).
on system sensitivity. Prepare calibration standards using buff- Postcolumn reactor: Reactor coil: 80°C; reagent 1 and 2 flow rates are instrument-
ered reagent water (¶ 4b). An example of the dilutions used to dependent; 0.3 to 0.5 mL/min.
prepare the CALs used to collect the data in this method is Fluorescence detector: 340-nm excitation, 465-nm emission with a 18-nm band
shown in Table 6610:III. These standards also may be used as width; gain⫽100, attenuation⫽16; response⫽standard; 16-␮L flow cell.
continuing calibration checks (¶ 6d). TABLE 6610:V. RETENTION TIMES FOR ANALYTES*

5. Procedure Relative Standard

Retention Time Standard Deviation
Analyte min Deviation %
a. Chromatographic conditions: Establish operating condi-
tions as recommended in Table 6610:IV for the HPLC system. Aldicarb sulfoxide 4.369 0.0092 0.21
Other HPLC conditions may be used as long as all QC require- Aldicarb sulfone 5.072 0.0089 0.17
ments (¶ 6 below) are met. Establish an appropriate retention Oxamyl 5.744 0.0095 0.17
time window for each target and surrogate to identify them in the Methomyl 6.526 0.0077 0.12
QC and field samples. Base this retention time window on 3-Hydroxycarbofuran 9.824 0.0128 0.13
measurements of actual retention time variation for each com- Aldicarb 11.461 0.0129 0.11
Propoxur 14.321 0.0195 0.14
pound in standard solutions analyzed on the HPLC over the
Carbofuran 14.834 0.0244 0.16
course of time. Plus or minus three times the standard deviation Carbaryl 16.993 0.0264 0.16
of the retention time for each compound during initial calibration 1-Naphthol 18.579 0.0187 0.10
and completion of IDC can be used to calculate a suggested Methiocarb 21.826 0.015 0.07
window size; however, consider the experience of the analyst in BDMC (SUR) 22.341 0.015 0.07
determination of the appropriate retention window size. Confirm
* Retention time data calculated from precision and accuracy data results pre-
that retention times, compound separation, and resolution are sented in Table 6610:II reagent water injections and calibration curve used to
similar to those listed in Table 6610:V and Figure 6610:1. Check quantitate these data.
CARBAMATE PESTICIDES (6610)/High-Performance Liquid Chromatographic Method S-5

Figure 6610:1. Sample chromatogram of target analytes. Concentration 2.0 ␮g/L, 1000-␮L injection volume, analyzed under conditions stated in Table
6610:IV.

Generate a calibration curve for each analyte by plotting peak and the calculated amount for the lowest calibration point for
response (preferably area) against analyte concentration. The each analyte must be within ⫾ 50% of the true value. If the
instrument used during method development yielded linear values are outside these ranges, consider all data for the problem
curves for the target analytes over the concentration range of analyte invalid and take remedial action (possibly including
interest. However, data may be fitted with either a linear regres- recalibration). After adequate calibration has been restored, re-
sion (response vs. concentration) or quadratic fit (response vs. analyze any field or QC samples that have been analyzed since
concentration). Alternatively, if the ratio of the analyte peak area the last acceptable calibration verification.
to concentration (or response factor) is relatively constant (RSD c. Sample preparation: Preserve, collect, and store samples as
⬍ 30%), an average response factor may be used to calculate directed in 6610B.2, above. Ensure that all field and QC samples
analyte concentration. contain the required preservatives. Measure a portion of sample
When each calibration standard is calculated as an unknown into a volumetric flask (25-mL recommended). When the vol-
using the calibration curve, the results must be 70 to 130% of the umes of SUR and analyte PDS added to the field and QC
true value for all but the lowest standard, and the lowest standard samples are kept to a minimum as described below, no volume
must be 50 to 150% of the true value. adjustment is necessary.
2) Daily calibration verification — Verify accuracy of initial Add a portion of the SUR PDS [¶ 4i1)b)] to all samples and
calibration during each analysis batch (set of samples prepared mix by capping and inverting sample. For the development of
and analyzed on the same instrument during a 24-h period). this method, the addition of 5.0 ␮L of a 10-␮g/mL SUR PDS to
Begin an analysis batch with a continuing calibration check a 25-mL sample resulted in a SUR concentration of 2.0 ␮g/L.
(CCC) at or below the MRL, analyze a CCC after every 10 field If sample is a CCC or other target-analyte-fortified QC sample
samples (alternate between medium and high concentrations), (¶s 6f and g), add the necessary amount of analyte PDS [¶
and end batch with a CCC. Limit analysis batch to 20 field 4i2)b)]. Cap and invert each sample to ensure that all compo-
samples. See ¶ 6 for the required QC samples for each analysis nents are properly mixed.
batch. Filter samples before filling appropriate autosampler vials.
Inject a portion of the appropriate concentration calibration Preferably use filters specified in ¶ 3f.
solution and analyze with the same conditions used during initial d. Analysis of samples: Inject and analyze samples, including
calibration. CCCs and QC samples, using conditions identical with those
Calculate concentration of each analyte and surrogate in the used for initial calibration. Do not extrapolate beyond the estab-
CCC standard. The calculated amount for each analyte for me- lished calibration range. If an analyte peak area exceeds the
dium- and high-level CCCs must be ⫾ 30% of the true value, range of the initial calibration curve, the extract may be diluted
S-6 INDIVIDUAL ORGANIC COMPOUNDS (6000)

TABLE 6610:VI. SUMMARY OF REQUIREMENTS FOR INITIAL DEMONSTRATION OF CAPABILITY (IDC)

Method Reference Requirement Specification and Frequency Acceptance Criteria

¶ 6a1) Initial demonstration of Analyze laboratory reagent blank Target analytes below 1/3 intended
low system before any other IDC steps. MRL; possible interferences
background from reagents and glassware do
not prevent analyte identification
and quantitation.
¶ 6a2) Quality control sample Use second source standard to Verifies initial calibration
(QCS) fortify buffered reagent water. accuracy; recovery within
Analyze as a CCC after initial ⫾ 30% of true value.
calibration but before IDA
sample analysis.
¶ 6a2) Initial demonstration of Analyze 4–7 replicate LFBs/CCCs Mean recovery within ⫾ 20% of
accuracy (IDA) fortified at mid-range true value.
concentration. Calculate average
recovery for replicates used in
IDP.
¶ 6a3) Initial demonstration of Calculate average recovery for RSD ⱕ 20%
precision (IDP) replicates used in IDA.
¶ 6a4) Detection level Over 3-d period prepare at least 7 Data not required to meet method
determination replicate LFBs fortified at precision and accuracy
concentration estimated to be criteria.[See note, ¶ 6a4)].
near the detection level.
Analyze replicates through all
steps of analysis. Calculate
detection level according to
¶ 6a4).

with buffered reagent water. Determine acceptable surrogate the method is unacceptable; identify and correct the source of the
performance (¶ 6i) from the undiluted sample extract. Any problem. After calibration accuracy has been verified, prepare
dilutions will also affect analyte MRL. For final calculations, see and analyze four to seven replicate LFBs (or CCCs in this
¶ 7 below. method) fortified at 2 ␮g/L, or near the mid-range of the initial
calibration curve, according to the procedure described in ¶ 4 i3).
6. Quality Control Also add sample preservatives (¶ 2a) to these samples. Average
recovery of the replicate values must be within ⫾ 20% of the
Quality control (QC) requirements include the initial demon- true value.
stration of capability (IDC), determination of the detection level, 3) Initial demonstration of precision — Using the same set
and subsequent analysis in each analysis batch of a laboratory of replicate data generated for ¶ 2) above, calculate the
reagent blank (LRB), continuing calibration check (CCC) stan- standard deviation and percent relative standard deviation of
dards, a laboratory-fortified blank (LFB), a laboratory-fortified the replicate recoveries. The percent relative standard devia-
sample matrix (LFSM), either a laboratory-fortified sample ma- tion (% RSD) of the results of the replicate analyses must be
trix duplicate (LFSMD) or a field duplicate sample, and the less than 20%.
periodic analysis of a quality control sample (QCS). This section 4) Detection level determination — Prepare and analyze at
details the specific requirements for each QC parameter. The QC least seven replicate LFBs at a concentration estimated to be
criteria discussed in the following sections are summarized in near the detection level over at least 3 d by the procedure
Tables 6610:VI and VII. These criteria are considered the min- described in ¶ 4i3). This fortification level may be estimated
imum acceptable QC criteria; institute additional QC practices to by selecting a concentration with a signal of 2 to 5 times the
meet specific needs. noise level. The appropriate concentration will depend on the
a. Initial demonstration of capability: Requirements for the sensitivity of the HPLC system being used. Also add sample
IDC are described in ¶s 1) through 4) below and are summarized preservatives (¶ 2a) to these samples. Calculate the detection
in Table 6610:VI. level using the equation
1) Initial demonstration of low system background — Before
any field samples are analyzed, and any time a new set of reagents Detection level ⫽ St共n ⫺ 1, 1 ⫺ ␣ ⫽ 0.99兲
is used, demonstrate that a laboratory reagent blank is reasonably
free of contamination and that the criteria in ¶ 6c are met. where:
2) Initial demonstration of accuracy — Before analysis of the S ⫽ standard deviation of replicate analyses,
IDC samples, verify calibration accuracy with the preparation t(n⫺1,1⫺␣ ⫽ 0.99)⫽ Student’s t value for the 99% confidence level
and analysis of a mid-level QCS as defined in ¶ 6j. If the analyte with n⫺1 degrees of freedom, and
recovery is not within ⫾ 30% of the true value, the accuracy of n ⫽ number of replicates.
CARBAMATE PESTICIDES (6610)/High-Performance Liquid Chromatographic Method S-7

TABLE 6610:VII. SUMMARY OF QUALITY CONTROL REQUIREMENTS

Method
Reference Requirement Specification and Frequency Acceptance Criteria

¶ 2c Sample holding times Ship preserved samples at or below Do not report data for samples that have
10°C; hold in lab at ⱕ 6°C for up to not been properly preserved or stored,
28 d. Do not freeze. or that have exceeded specified
holding time.
¶ 6c Laboratory reagent blank (LRB) Include with each analysis batch (up to All target analytes below 1/3 intended
20 samples). Analyze before samples MRL; possible interferences from
and determine to be free from reagents and glassware do not prevent
interferences. analyte identification and quantitation.
If targets exceed 1/3 MRL, analytical
batch results for that analyte are
invalid.
¶ 5b2) Continuing calibration check (CCC) Verify initial calibration by analyzing Recovery for each analyte within 70–
CCC (at or below MRL) before 130% of true value for all but lowest
analyzing samples. Inject CCCs after calibration level. Lowest calibration
every 10 samples and after last level CCC within 50–150% of true
sample, rotating concentrations to value. Results not bracketed by
cover calibration range. acceptable CCCs are invalid.
¶ 6i Surrogate standards Add surrogate, ¶ 4i1), to all field and QC Surrogate recovery within 70–130% of
samples. true value. Report samples that fail
criteria as suspect.
¶s 6f & g Laboratory-fortified sample matrix (LFSM) With each analysis batch, extract and Recoveries not within 70–130% (50–
and laboratory-fortified sample matrix analyze at least one LFSM. Extract 150% at MRL) of fortified amount
duplicate (LFSMD) LFSMD when occurrence of target may indicate matrix effect. If LFSMD
analytes is low. Field duplicate is analyzed instead of a laboratory
analysis is not required for extraction duplicate, accept if target RPDs within
batches containing a LFSMD. ⫾ 30%. If all CCCs meet acceptance
criteria and LFSM or LFSMD do not,
designate sample “suspect/matrix.”
¶ 6g Field duplicates (FD) Analyze at least one duplicate with each RPDs within ⫾ 30%. If all CCCs meet
extraction batch (20 samples or less). acceptance criteria and FDs do not,
LFSMD may be substituted when designate sample “suspect/matrix.”
analyte occurrence is low.
¶ 6h Resolution check Monitor once/ 24-h analysis period. Resolution of ⱖ 1.0 for closely eluting
peaks that are not baseline resolved
(see ¶ 6h).
¶ 6j Quality control sample Analyze at least quarterly or when Same acceptance criteria as CCC.
preparing new standards, as well as
during IDC.
¶ 5b1) Initial calibration Use external standard calibration For each calibration standard, calculated
technique to generate calibration curve as an unknown using calibration curve,
with at least 5 standards. results within 70–130% of true value
for all but lowest standard; lowest
standard (concentration ⱕ MRL)
results within 50–150% of true value.

NOTE: Calculated detection levels need only be less than 1⁄3 of either three times the detection level or a concentration yielding
the laboratory’s MRL to be considered acceptable. Do not sub- a response less than a signal-to-noise ratio of five. Depending
tract blank values when performing detection-level calculations. upon the study’s data quality objectives, it may be set at a higher
The detection level is a statistical determination of precision concentration.
only.1 If the detection level replicates are fortified at a low c. Laboratory reagent blank (LRB): This is a direct injection
enough concentration, it is likely that they will not meet preci- method without a conventional extraction. A LRB, prepared
sion and accuracy criteria, and may result in a calculated detec- using buffered reagent water and filtering in the same manner as
tion level higher than the fortified concentration. the samples, is required with each analysis batch of samples to
b. Minimum reporting level (MRL): The MRL is the threshold determine any background system contamination. If, within the
concentration of an analyte that a laboratory can expect to retention time window of any analyte, the LRB produces a peak
quantitate accurately in an unknown sample. Do not establish the that would prevent determination of that analyte, determine
MRL at an analyte concentration lower than that of the lowest source of contamination and eliminate the interference before
calibration standards. The MRL also should not be less than processing samples. Reduce background contamination to an
S-8 INDIVIDUAL ORGANIC COMPOUNDS (6000)

acceptable level before proceeding. Keep background from associated with sample collection, preservation, storage, and
method analytes or contaminants that interfere with the measure- laboratory procedures. If target analytes are not routinely ob-
ment of method analytes below 1⁄3 the MRL. If target analytes served in field samples, analyze a LFSMD rather than a FD.
are detected in the LRB at concentrations equal to or greater than Calculate relative percent difference (RPD) for duplicate mea-
this level, consider all data for the problem analyte(s) invalid for surements (FD1 and FD2) using the equation
all samples in the analysis batch.
d. Continuing calibration check (CCC): Prepare a CCC in the 共FD1 ⫺ FD2兲
same manner as the initial calibration solutions, using buffered RPD ⫽ ⫻ 100
共FD1 ⫹ FD2兲/2
reagent water and filtering in the same manner as the samples.
Analyze during an analysis batch at a required frequency to If a LFSMD is analyzed instead of a FD, calculate relative
confirm that the instrument meets initial calibration criteria. See percent RPD for duplicate LFSMs (LFSM and LFSMD) using
¶ 5b2) for concentration requirements, frequency requirements, the equation
and acceptance criteria.
e. Laboratory-fortified blank (LFB): For this direct injection
method, a CCC is the same as an LFB. Consequently, the 共LFSM ⫺ LFSMD兲
RPD ⫽ ⫻ 100
analysis of an LFB is not required. 共LFSM ⫹ LFSMD兲/2
f. Laboratory-fortified sample matrix (LFSM): Analyze an
LFSM in each analysis batch to determine that the sample matrix RPDs for FDs and duplicate LFSMs should fall in the range of
does not adversely affect method accuracy. If the occurrence of ⫾ 30% for samples fortified at or above their original concen-
target analytes in the samples is infrequent, or if historical trends tration. Greater variability may be observed when LFSMs are
are unavailable, prepare a second LFSM, or LFMSD, from a fortified near the MRL. At the MRL, RPDs should fall in the
duplicate of the field sample used to prepare the LFSM, and range of ⫾ 50% for samples fortified at or above their original
analyze to assess method precision. Extraction batches that con- concentration.
tain LFSMDs do not require the analysis of a field duplicate. If h. Resolution check: In each analytical batch, monitor resolu-
a variety of different sample matrices are analyzed regularly, for tion of peaks in a calibration standard or CCC near the mid-level
example, drinking water from groundwater and surface water of calibration. During the development of this method, the
sources, establish method performance for each. Over time, 2-␮g/L level was monitored. Check that closely eluting peaks
document LFSM data for all routine sample sources for the that are not baseline-resolved have a resolution (Rs) of 1.0 or
laboratory. greater, calculated by the equation6
Within each analysis batch, fortify a minimum of one field
sample as an LFSM for every 20 samples processed. Prepare 1.18共t 2 ⫺ t 1兲
LFSM by adding an appropriate amount of the analyte PDS Rs ⫽
W 0.5,1 ⫹ W 0.5,2
[¶ 4i2)b)] to a sample. Select a fortifying concentration at least
twice the matrix background concentration, if known. Use his-
where:
torical data or rotate through the designated concentrations when
t 1 and t 2 ⫽ retention times of the first and second adjacent peaks,
selecting a fortifying concentration. Selecting a duplicate bottle
and
of a sample that has already been analyzed aids in the selection
W0.5,1 and
of appropriate fortifying levels.
W0.5,2 ⫽ widths of the adjacent peaks at half height.
Calculate the percent recovery (R) for each analyte with the
equation Monitor resolution once for every 24-h analytical batch at any
time during the 24-h period. Preferably check resolution before
共A ⫺ B兲 sample analysis, especially if the system in use has a history of
R⫽ ⫻ 100 resolution problems. If a resolution check fails, reanalyze all
C
samples in the analytical batch, including the QC samples, after
the problem is corrected.
where:
i. Surrogate recovery: Fortify all samples, blanks, LRBs, and
A ⫽ measured concentration in the fortified sample,
LFSMs and LFSMDs with surrogate standard before filtration
B ⫽ measured concentration in the unfortified sample, and
and analysis. Also add it to the calibration curve and calibration
C ⫽ fortification concentration.
check standards. The surrogate is a means of assessing method
Analyte recoveries may exhibit a matrix bias. For samples performance from preparation and filtration to final chromato-
fortified at or above their original concentration, recoveries graphic measurement.
should be between 70 and 130%. For LFSM fortification at the When surrogate recovery from a sample, blank, or CCC is less
MRL, 50 to 150% recoveries are acceptable. If the accuracy of than 70% or greater than 130%, check the following: calculations to
any analyte falls outside the designated range, and the laboratory locate possible errors, standard solutions for degradation, possible
performance for that analyte is shown to be in control in the contamination, and instrument performance. If those checks do not
CCCs, the recovery is matrix-biased. Label the result for that reveal the cause of the problem, reanalyze the sample.
analyte in the unfortified sample “suspect/matrix”. If the reanalysis meets the surrogate recovery criterion, report
g. Field duplicate or laboratory-fortified sample matrix dupli- only data for the reanalyzed sample.
cate (FD or LFSMD): Within each analysis batch, analyze a If the sample reanalysis fails the 70 –130% recovery criterion,
minimum of one FD or LFSMD. Duplicates check the precision check calibration by reinjecting the most recently acceptable
CARBAMATE PESTICIDES (6610)/High-Performance Liquid Chromatographic Method S-9

calibration standard. If the calibration standard fails the criteria standard. Samples with target analyte responses that exceed the
of ¶ 5b2), recalibrate. If the calibration standard is acceptable, highest standard require dilution and reanalysis (¶ 5d).
repeat preparation (including fortifying with surrogate and fil- Adjust calculated concentrations of detected analytes to reflect
tration) and analysis of the sample if the sample is still within the initial sample volume and any dilutions performed.
holding time. If this sample reanalysis also fails the recovery Before reporting the data, review the chromatogram for any
criterion, report all data for that sample as suspect because of incorrect peak identification or poor integration.
unsatisfactory surrogate recovery. Report analyte concentrations in micrograms per liter, usually
j. Quality control sample (QCS): During the analysis of the IDC to two significant figures.
(¶ 6a), each time that new analyte standard solutions are prepared,
or at least quarterly, analyze a QCS from a source different from the 8. References
source of the calibration standards. Fortify QCS at the mid-level of
calibration in buffered reagent water and analyze in same manner as 1. GLASER, J.A., D.L. FOERST, G.D. MCKEE, S.A. QUAVE & W.L. BUDDE.
a CCC. The acceptance criteria are the same as the CCC criteria at 1981. Trace analyses for wastewaters. Environ. Sci. Technol. 15:1426.
mid-level: the calculated amount for each analyte must be ⫾ 30% 2. OCCUPATIONAL SAFETY and HEALTH ADMINISTRATION. 1976. OSHA
of the true value. If measured analyte concentrations are not of Safety and Health Standards, General Industry. OSHA 2206 29 CFR
1910, rev. Jan. 1976, Occupational Safety and Health Administration.
acceptable accuracy, check entire analytical procedure to locate and
3. Carcinogens — Working with Carcinogens. 1977. Publication No.
correct the problem source. 77–206, Dep. Health, Education, & Welfare, Public Health Service,
Center for Disease Control, National Inst. Occupational Safety &
7. Data Analysis and Calculation Health, Atlanta, Ga.
4. AMERICAN CHEMICAL SOCIETY. 1979. Safety in Academic Chemistry
Identify method analytes in the sample chromatogram by Laboratories, 3rd ed. Committee on Chemical Safety, American
comparing the retention time of the suspect peak to the retention Chemical Society, Washington, D.C.
5. U.S. ENVIRONMENTAL PROTECTION AGENCY. 2001. Method 531.2—
time of an analyte peak in a calibration standard. Confirm that
Measurement of N-methylcarbamoyloximes and N-methylcarbam-
surrogate retention times are within acceptance limits (¶ 6i), even ates in water by direct aqueous injection HPLC with postcolumn
if no target compounds are detected. derivatization. EPA 815-B-01-002, U.S. Environmental Protection
Calculate analyte concentrations using the initial calibration Agency, Off. Ground Water and Drinking Water, Cincinnati, Ohio.
curve generated as described in ¶ 5b1). Quantitate only those 6. SNYDER, L.R., J.J. KIRKLAND & J.L. GLAJCH. 1997. Practical HPLC
values that fall between the MRL and the highest calibration Method Development, 2nd ed. John Wiley & Sons, New York, N.Y.