METHOD 556. DETERMINATION OF CARBONYL COMPOUNDS IN DRINKING
WATER BY PENTAFLUOROBENZYLHYDROXYLAMINE
DERIVATIZATION AND CAPILLARY GAS CHROMATOGRAPHY
WITH ELECTRON CAPTURE DETECTION
Revision 1.0
June 1998
J.W. Munch, US EPA Office of Research and Development, NERL
DJ. Munch, US EPA, Office of Ground Water and Drinking Water
S.D. Winslow, S.C. Wendelken, B.V. Pepich, ICF Kaiser Engineers, Inc.
NATIONAL EXPOSURE RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
556-1
-------
METHOD 556
DETERMINATION OF CARBONYL COMPOUNDS IN DRINKING WATER BY
PENTAFLUORBENZYLHYROXYLAMINE DERIVATIZATION AND CAPILLARY
GAS CHROMATOGRAPHY AND ELECTRON CAPTURE DETECTION
1. SCOPE AND APPLICATION
1.1 This is a gas chromatographic method optimized for the determination of selected
carbonyl compounds in finished drinking water and raw source water. The analytes
applicable to this method are derivatized to their corresponding pentafluorobenzyl
oximes. The oxime derivatives are then extracted from the water with hexane. The
hexane extracts are analyzed by capillary gas chromatography with electron capture
detection (GC-ECD) and quantitated using procedural standard calibration. Accu-
racy, precision, and method detection limit (MDL) data have been generated for the
following compounds:
Chemical Abstract Services
Analyte Registry Number
Formaldehyde 50-00-0
Acetaldehyde 75-07-0
Propanal 123-38-6
Butanal 123-72-8
Pentanal 110-62-3
Hexanal 66-25-1
Heptanal 111-71-7
Octanal 124-13-0
Nonanal 124-19-6
Decanal 112-31-2
Cyclohexanone 108-94-1
Crotonaldehyde 123-73-9
Benzaldehyde 100-52-7
Glyoxal (ethanedial) 107-22-2
Methyl glyoxal (2-oxopropanal or pyruvic aldehyde) 78-98-8
1.2 This method applies to the determination of target analytes over the concentration
ranges typically found in drinking water. Analyte retention times are in Section 17,
Table 1. Other method performance data are presented in Sectionl7, Tables 2-6.
Experimentally determined method detection limits (MDLs) for the above listed
analytes are provided in Section 17, Table 3. The MDL is defined as the statistically
556-2
-------
calculated minimum amount that can be measured with 99% confidence that the
reported value is greater than zero.(1) However, it should be noted that background
levels of some method analytes (usually formaldehyde and acetaldehyde) are prob-
lematic. The minimum reporting level (MRL) for method analytes, for each ana-
lyst/laboratory that uses this method, will depend on their ability to control back-
ground levels (Sect. 4).
1.3 This method is restricted to use by or under the supervision of analysts skilled in
liquid-liquid extractions, derivatization procedures and the use of GC and interpreta-
tion of gas chromatograms. Each analyst must demonstrate the ability to generate
acceptable results with this method, using the procedures described in Section 9.
2.0 SUMMARY OF METHOD
2.1 A 20 mL volume of water sample is adjusted to pH 4 with potassium hydrogen
phthalate (KHP) and the analytes are derivatized at 35 °C for 2 hr with 15 mg of O-
(2,3,4,5,6-Pentafluorobenzyl)-hydroxylamine (PFBHA) reagent. The oxime deriva-
tives are extracted from the water with 4 mL hexane. The extract is processed
through an acidic wash step, and then analyzed by GC-ECD. The target analytes are
identified and quantitated by comparison to a procedural standard (Sect. 3.9). Two
chromatographic peaks will be observed for many of the target analytes. Both (E)
and (Z) isomers are formed for carbonyl compounds that are asymmetrical, and that
are not sterically hindered. However, the (E) and (Z) isomers may not be
chromatographically resolved in a few cases. Compounds where two carbonyl groups
are derivatized, such as glyoxal and methyl glyoxal, have even more possible isomers.
See Section 17, Table 1 and Figure 1 for the chromatographic peaks used for analyte
identification.
NOTE: The absolute identity of the (E) and (Z) isomers was not determined during
method development. Other researchers (2'3'4) have reported the first eluting
peak as (E), and the second peak as (Z). For convenience, this method will
follow this convention. Because more than 2 isomers are formed for
glyoxal and methyl glyoxal, the peaks used for identification are referred to
as "peak 1" and "peak 2."
2.2 All results should be confirmed on a second, dissimilar capillary GC column.
3. DEFINITIONS
3.1 LABORATORY REAGENT BLANK (LRB) - An aliquot of reagent water or other
blank matrix that is treated exactly as a sample including exposure to all glassware,
equipment, solvents and reagents, sample preservatives, internal standards, and
surrogates that are used with other samples. The LRB is used to determine if method
556-3
-------
analytes or other interferences are present in the laboratory environment, the re-
agents, or the apparatus.
3.2 FIELD REAGENT BLANK (FRB) -- An aliquot of reagent water or other blank
matrix that is placed in a sample container in the laboratory and treated as a sample in
all respects, including shipment to the sampling site, storage, preservation, and all
analytical procedures. The purpose of the FRB is to determine if method analytes or
other interferences are introduced during sample shipping or storage. For this
analysis the FRB should not be opened at the sampling site.
3.3 LABORATORY FORTIFIED BLANK (LFB) -- An aliquot of reagent water or
other blank matrix to which known quantities of the method analytes are added in the
laboratory. The LFB is analyzed exactly like a sample, and its purpose is to deter-
mine whether the methodology is in control, and whether the laboratory is capable of
making accurate and precise measurements.
3.4 LABORATORY FORTIFIED SAMPLE MATRIX (LFM) - An aliquot of an envi-
ronmental sample to which known quantities of the method analytes are added in the
laboratory. The LFM is analyzed exactly like a sample, and its purpose is to deter-
mine whether the sample matrix contributes bias to the analytical results. The
background concentrations of the analytes in the sample matrix must be determined
in a separate aliquot and the measured values in the LFM corrected for background
concentrations.
3.5 STOCK STANDARD SOLUTION (SSS) - A concentrated solution containing one
or more method analytes prepared in the laboratory using assayed reference materials
or purchased from a reputable commercial source.
3.6 PRIMARY DILUTION STANDARD SOLUTION (PDS) - A solution of several
analytes prepared in the laboratory from stock standard solutions and diluted as
needed to prepare calibration solutions and other needed analyte solutions.
3.7 CALIBRATION STANDARD (CAL) - A solution prepared from the primary
dilution standard solution and stock standard solutions and the internal standards and
surrogate analytes. The CAL solutions are used to calibrate the instrument response
with respect to analyte concentration.
3.8 QUALITY CONTROL SAMPLE (QCS) - A solution of method analytes of known
concentrations which is used to fortify an aliquot of LRB or sample matrix. The
QCS is obtained from a source external to the laboratory and different from the
source of calibration standards. It is used to check laboratory performance with
externally prepared test materials.
556-4
-------
3.9 PROCEDURAL STANDARD CALIBRATION - A calibration method where
aqueous calibration standards are prepared and processed (e.g. purged, extracted,
and/or derivatized) in exactly the same manner as a sample. All steps in the process
from addition of sampling preservatives through instrumental analyses are included in
the calibration. Using procedural standard calibration compensates for any inefficien-
cies in the processing procedure.
3.10 INTERNAL STANDARD (IS) - A pure analyte added to a sample, extract, or
standard solution in known amount(s) and used to measure the relative responses of
other method analytes and surrogates that are components of the same sample or
solution. The internal standard must be an analyte that is not a sample component.
3.11 SURROGATE ANALYTE (SUR) - A pure analyte, which is extremely unlikely to
be found in any sample, and which is added to a sample aliquot in known amount(s)
before extraction or other processing and is measured with the same procedures used
to measure other sample components. The purpose of the SA is to monitor method
performance with each sample.
3.12 METHOD DETECTION LIMIT (MDL) - The minimum concentration of an
analyte that can be identified, measured and reported with 99% confidence that the
analyte concentration is greater than zero.
3.13 MATERIAL SAFETY DATA (MSDS) - Written information provided by vendors
concerning a chemical's toxicity, health hazards, physical properties, fire, and
reactivity data including storage, spill, and handling precautions.
3.14 CONTINUING CALIBRATION CHECK (CCC) - A calibration standard contain-
ing one or more method analytes, which is analyzed periodically to verify the
accuracy of the existing calibration for those analytes.
3.15 MINIMUM REPORTING LEVEL (MRL) - The minimum concentration of an
analyte that should be reported. This concentration is determined by the background
level of the analyte in the LRBs and the sensitivity of the method to the analyte.
Ideally, the MRL will be at or near the concentration of the lowest calibration
standard.
4. INTERFERENCES
4.1 Method interferences may be caused by contaminants in laboratory air, solvents,
reagents (including reagent water), glassware, sample bottles and caps, and other
sample processing hardware that lead to discrete artifacts and/or elevated baselines in
the chromatograms. All of these materials must be routinely demonstrated to be free
from interferences (less than 1/2 the MRL) under the conditions of the analysis by
analyzing laboratory reagent blanks as described in Section 9.3. Subtracting blank
values from sample results is not permitted.
556-5
-------
4.1.1 Before attempting analyses by this method, the analyst must obtain a
source of reagent water free from carbonyl compounds and other
interferences. The most likely interferences are the presence of
formaldehyde and acetaldehyde in the reagent water. The most successful
techniques for generating aldehyde free water are (1) exposure to UV light,
or (2) distillation from permanganate.
4.1.2 Commercially available systems for generating reagent grade water have
proved adequate, if a step involving exposure to UV light is included. For
the data presented in this method, a Millipore Elix 3 reverse osmosis
system followed by a Milli-Q TOC Plus polishing unit provided reagent
water with background levels of 1 ug/L or less for each method analyte.
Other researchers have reported typical blank values of 1-3 ug/L.(3'4)
4.1.3 Distillation of reagent water from acidified potassium permanganate has
been reported as an effective method of eliminating background levels of
aldehydes.(2) Distill 500 mL of reagent water to which 64 mg potassium
permanganate and 1 mL cone, sulfuric acid have been added. In our
laboratory, this procedure reduced formaldehyde levels to approximately 3
ug/L.
4.1.4 It may be necessary to purchase reagent grade water. If acceptably clean
reagent grade water is purchased, care must also be taken to protect it
from contamination caused by contact with laboratory air.
4.2 Formaldehyde is typically present in laboratory air and smaller amounts of other
aldehydes may also be found. Care should be taken to minimize exposure of reagents
and sample water with laboratory air. Because latex is a potential aldehyde
contaminant source, protective gloves should not contain latex. Nitrile gloves, such
as N-Dex Plus, are acceptable. Bottle caps should be made of polypropylene.
Commonly used phenolic resin caps must be avoided because they can introduce
formaldehyde contamination into samples.
4.3 Reagents must also be free from contamination. Many brands of solvents may
contain trace amounts of carbonyl compounds.
4.4 Glassware must be scrupulously cleaned by detergent washing with hot water, and
rinses with tap water and distilled water. Glassware should then be drained, dried,
and heated in a laboratory oven at 130 °C for several hours before use. Solvent rinses
with methanol or acetonitrile, followed by air drying, may be substituted for the oven
heating. After cleaning, glassware should be stored in a clean environment to prevent
any accumulation of dust or other contaminants.
556-6
-------
4.5 Matrix interferences may be caused by contaminants that are coextracted from the
sample. The extent of matrix interferences will vary considerably from source to
source, depending upon the nature and diversity of the matrix being sampled.
4.6 An interferant that elutes just prior to the acetaldehyde (E) isomer peak
on the primary column is typically observed in chlorinated or
chloraminated waters. If this peak interferes with the integration of the
acetaldehyde (E) isomer peak, then acetaldehyde should be quantitated
using only the acetaldehyde (Z) isomer, or from the confirmation column
data.
5. SAFETY
5.1 The toxicity or carcinogen!city of each reagent used in this method has not been
precisely defined; however, each chemical compound should be treated as a potential
health hazard. From this viewpoint, exposure to these chemicals must be reduced to
the lowest possible level by whatever means available. The laboratory is responsible
for maintaining a current awareness file of OSHA regulations regarding the safe
handling of the chemicals specified in this method. A reference file of material safety
data sheets should also be made available to all personnel involved in the chemical
analysis. Additional references to laboratory safely are available.(5"8)
5.2 Formaldehyde and acetaldehyde have been tentatively classified as
known or suspected human or mammalian carcinogens. Glyoxal and
methyl glyoxal have been shown to be mutagenic in in-vitro tests.(2)
6. EQUIPMENT AND SUPPLIES (All specifications are suggested. Brand names and/or
catalog numbers are included for illustration only.)
6.1 SAMPLE CONTAINERS -- Grab Sample Bottle (aqueous samples) - 30 mL amber
glass, screw cap bottles and caps equipped with Teflon-faced silicone septa. Screw
caps should be polypropylene Typical phenolic resin caps should be avoided due
to the possibility of sample contamination from formaldehyde Prior to use,
wash bottles and septa according to Section 4.4.
6.2 VIALS -- 8 mL or 12 mL vials for the acid wash step (Sect. 11.1.10), and GC
autosampler vials, both types must be glass with Teflon-lined polypropylene caps.
6.3 VOLUMETRIC FLASKS - various sizes used for preparation of standards.
6.4 BALANCE - Analytical, capable of accurately weighing to the nearest 0.0001 g.
6.5 WATER BATH or HEATING BLOCK - Capable of maintaining 3 5 ± 2 °C
556-7
-------
6.6 GAS CHROMATOGRAPH -- Capillary Gas Chromatograph equipped with a
split/splitless injector, or other injector suitable for trace analysis, and an electron
capture detector.
6.6.1 Primary Column — 30 m x 0.25mm J&W DB-5ms, 0.25 um film thickness
(or equivalent). Note: The J&W DB-5 was not found to be equivalent for
this application. The surrogate analyte is not resolved from octanal with
the DB-5 column.
6.6.2 Confirmation Column - 30 m x 0.25 mm Restek Rtx- 1701, 0.25 um film
thickness (or equivalent)
7. REAGENTS AND STANDARDS
7.1 Reagent grade or better chemicals should be used in all tests. Unless otherwise
indicated, it is intended that all reagents shall conform to the specifications of the
Committee on Analytical Reagents of the American Chemical Society, where such
specifications are available. Other grades may be used, provided it is first ascertained
that the reagent is of sufficiently high purity to permit its use without lessening the
accuracy of the determination.
7.2 REAGENT WATER —. Reagent water as free as possible from interferences and
contamination is critical to the success of this method. See Section 4.1.
7.3 ACETONITRILE - High purity, demonstrated to be free of analytes and
interferences.
7.4 HEXANE — High purity, demonstrated to be free of analytes and interferences: B&J
Brand, GC2 grade or equivalent.
7.5 POTASSIUM HYDROGEN PHTHALATE (KHP) -- ACS Grade or better.
7.6 0-(2,3,5,6-PENTAFLUOROBENZYL)-HYDROXYLAMINE HYDROCHLORIDE
(PFBHA) -- 98+%, Aldrich cat# 19,448-4. (Store in a desiccator - Do not
refrigerate).
7.7 SULFURIC ACID - ACS Grade or better.
7.8 COPPER SULFATE PENTAHYDRATE - ACS Grade or better.
7.9 AMMONIUM CHLORIDE, NH4C1 or AMMONIUM SULF ATE, (NH4)2SO4.
7.10 SOLUTIONS
556-8
-------
7.10.1 PFBHA REAGENT — Prepare a fresh 15 mg/mL solution in reagent water
daily. Prepare an amount appropriate to the number of samples to be
derivatized. One mL of solution is added per sample. For example, if 14
sample vials are being extracted, prepare 15 mL of solution. For a 15 mL
volume of solution, weigh 0.225 grams of PFBHA into a dry 40 mL vial,
add 15 mL water and shake to dissolve.
7.10.2 0.2 N SULFURIC ACID -- Add 5 mL of concentrated sulfuric acid to 900
mL of reagent water.
7.11 STOCK STANDARD SOLUTIONS -- When a compound purity is assayed to be
96% or greater, the weight can be used without correction to calculate the
concentration of the stock standard.
7.11.1 INTERNAL STANDARD (IS) -- 1,2-DIBROMOPROPANE, 98+%
purity. An alternate compound may be used as the IS at the discretion of
the analyst. If an alternate is selected, an appropriate concentration will
need to be determined.
7.11.1.1 INTERNAL STANDARD STOCK SOLUTION, (10,000
ug/mL) — Accurately weigh approximately 0.1 gram to the
nearest O.OOOlg, into a tared 10 mL volumetric flask
containing hexane up to the neck. After determining weight
difference, fill to mark with hexane. Stock solutions can be
used for up to 6 months when stored at -10 °C.
7.11.1.2 INTERNAL STANDARD FORTIFIED EXTRACTION
SOLVENT, 400 ug/L in hexane — This is the solvent used to
extract the derivatized samples. The internal standard is
added to the solvent prior to performing the extraction. The
volume of this solvent to be prepared should be determined by
the sample workload. The following example illustrates
preparation of 1 L of fortified solvent. If fewer samples are to
be analyzed each month, prepare smaller batches of working
solvent. Add 40 uL of internal standard stock solution
directly to 1 L of hexane in a volumetric flask. Cap flask and
invert three times to ensure thorough mixing. Transfer to 1 L
storage bottle with Teflon lined cap. This solution can be
used up to 4 weeks. As a check, run a sample of this working
solvent on the GC before the first extraction of aqueous
samples. Have enough working solvent available to extract
all calibration and aqueous samples in each extraction set.
Never use two different batches of working solvent for one
set of extractions.
556-9
-------
7.11.2 SURROGATE (SUR) - 2',4',5' -TRIFLUOROACETOPHENONE
This compound was found to be an appropriate surrogate analyte for these
analyses. However, the chromatograms for this analysis are very crowded,
and all possible matrix interferences cannot be anticipated. An alternate
carbonyl compound may be selected as the surrogate analyte if matrix
interferences or chromatographic problems are encountered. Any
surrogate analyte selected must form an oxime derivative, because its
purpose is to monitor the derivatization process. If an alternate surrogate
is selected, its concentration may also be adjusted to meet the needs of the
laboratory.
7.11.2.1 SURROGATE STOCK SOLUTION, 10,000 ug/mL -
Accurately weigh approximately 0.1 gram SUR to the nearest
O.OOOlg, into a 10 mL tared volumetric flask containing
acetonitrile up to the neck. After determining weight
difference, fill to mark with acetronitrile. Stock solutions can
be used for up to 6 months when stored at -10 °C or less.
7.11.2.2 SURROGATE ADDITIVE SOLUTION, 20 ug/mL - Dilute
the surrogate stock solution to 20 ug/mL in acetonitrile.
This solution can be used up to 3 months when stored at 4 °C
or less.
7.11.3 STOCK STANDARD SOLUTION (SSS)
Prepare stock standard solutions for each analyte of interest at a
concentration of 1 to 10 mg/mL in acetonitrile, or purchase SSSs or
primary dilution standards (PDSs) from a reputable supplier. Method
analytes may be obtained as neat materials or as ampulized solutions from
commercial suppliers. The stock standard solutions should be stored at
-10 °C or less and protected from light. Standards prepared in this manner
were stable for at least 60 days. Standards may be used for longer periods
of time if adequate records of stability are kept. Laboratories should use
standard QC practices to determine when their standards need to be
replaced.
7.11.3.1 For analytes which are solids in their pure form, prepare stock
standard solutions by accurately weighing approximately 0.1
gram of pure material to the nearest O.OOOlg in a 10 mL
volumetric flask. Dilute to volume with acetonitrile.
7.11.3.2 Stock standard solutions for analytes which are liquid in their
pure form at room temperature can be accurately prepared in
the following manner.
556-10
-------
7.11.3.2.1 Place about 9.8 mL of acetonitrile into a 10-
mL volumetric flask. Allow the flask to stand,
unstoppered, for about 10 min. to allow
solvent film to evaporate from the inner walls
of the volumetric, and weigh to the nearest
0.0001 gram.
7.11.3.2.2 Use a 100-uL syringe and immediately add 100
uL of standard material to the flask by keeping
the syringe needle just above the surface of the
acetonitrile. Be sure the standard material falls
dropwise directly into the acetonitrile without
contacting the inner wall of the volumetric.
7.11.3.2.3 Reweigh, dilute to volume, stopper, then mix
by inverting several times. Calculate the
concentration in milligrams per milliliter from
the net gain in weight.
7.11.4 PRIMARY DILUTION STANDARD (PDS) -- The PDS for this method
should include all method analytes of interest to the analyst. The PDS is
prepared by combining and diluting stock standard solutions with
acetonitrile to a concentration of 100 ug/mL. Store at -10 °C or less and
protect from light. Standards prepared in this manner were stable for at
least 60 days. Standards may be used for longer periods of time if
adequate records of stability are kept. Laboratories should use standard
QC practices to determine when their standards need to be replaced. This
primary dilution standard is used to prepare calibration spiking solutions,
which are prepared at 5 concentration levels for each analyte, and are used
to spike reagent water to prepare the aqueous calibration standards.
7.11.5 CALIBRATION SPIKING SOLUTIONS - Five calibration spiking
solutions are prepared, each at a different concentration, and are used to
spike reagent water to prepare the calibration standards. The calibration
spiking solutions are prepared from the PDS. Store the calibration spiking
solutions at -10 °C or less and protect from light. Solutions prepared in this
manner were stable for at least 60 days. Solutions may be used for longer
periods of time if adequate records of stability are kept. Laboratories
should use standard QC practices to determine when solutions need to be
replaced. An example of how the calibration spiking solutions are prepared
is given in the following table. Modifications of this preparation scheme
may be made to meet the needs of the laboratory.
556-11
-------
PREPARATION OF CALIBRATION SPIKING SOLUTIONS
Cal.
Level
1
2
3
4
5
PDS Cone.,
ug/mL
100
100
100
100
100
Vol. PDS
Std., uL
250
500
1000
1500
2000
Final Vol.,
Cal Spike
Sol'n, mLs
5
5
5
5
5
Final Cone.,
Cal Spike
Sol'n, ug/mL
5
10
20
30
40
7.11.6 PROCEDURAL CALIBRATION STANDARDS -- A designated amount
of each calibration spiking solution is spiked into five separate 20 mL
aliquots of reagent water in a 30 mL sample container, to produce
aqueous calibration standards. The reagent water used to make the
calibration standards should contain the preservation reagents described in
Section 8.1.2 (ammonium chloride or ammonium sulfate at 500 mg/L and
copper sulfate pentahydrate at 500 mg/L). Aqueous calibration standards
are processed and analyzed according to the procedures in Section 11.
Resulting data are used to generate a calibration curve. An example of the
preparation of aqueous calibration standards is given below. The lowest
concentration calibration standard should be at or near (within 25% of)
the MRL. Modifications of this preparation scheme may be made to meet
the needs of the laboratory. Preparing aqueous calibration standards using
varying volumes of one calibration spiking solution is an acceptable
alternative to the example below.
PREPARATION OF PROCEDURAL (AQUEOUS) CALIBRATION
STANDARDS
Cal.
Level
1
2
3
4
5
Cal. Spike Sol'n
Cone., ug/mL
5
10
20
30
40
Vol. Cal.
Spike
Sol'n., uL
20
20
20
20
20
Final Vol.,
Cal Std
mL
20
20
20
20
20
Final Cone.,
Cal Std
ug/L
5
10
20
30
40
556-12
-------
8. SAMPLE COLLECTION. PRESERVATION. AND STORAGE
8.1 SAMPLE VIAL PREPARATION
8.1.1 Grab samples must be collected in accordance with conventional sampling
practices (6) using amber glass 30 mL containers with PTFE-lined screw-
caps, or caps with PTFE-faced silicon septa.
8.1.2 Prior to shipment to the field, 15 mg of copper sulfate pentahydrate must
be added to each bottle. This material acts as a biocide to inhibit
bacteriological decay of method analytes. If samples to be collected contain
free chlorine, then 15mg of ammonium chloride or ammonium sulfate must
also be added to the bottle prior to sample collection. The ammonium
compound will react with the free chlorine to form monochloramine, and
retard the formation of additional carbonyl compounds. Add these
materials as dry solids to the sample bottle. The stability of these materials
in concentrated aqueous solution has not been verified.
NOTE: Aldehydes have been demonstrated to be extremely susceptible to
microbiological decay. The use of other chlorine reducing agents
such as sodium thiosulfate or ascorbic acid, has also been shown
to produce invalid data. Proper sample collection and preser-
vation is important to obtaining valid data. The data in Sectionl?,
Table 6 illustrates the importance of proper sample preservation.
8.2 SAMPLE COLLECTION
8.2.1 Fill sample bottles to just overflowing but take care not to flush out the
sample preservation reagents. The capped sample should be head-space
free.
8.2.2 When sampling from a water tap, remove the aerator so that no air bubbles
will be trapped in the sample. Open the tap, and allow the system to flush
until the water temperature has stabilized (usually about 3-5 min). Collect
samples from the flowing system.
8.2.3 When sampling from an open body of water, fill a 1 quart wide-mouth
bottle or 1L beaker with sample from a representative area, and carefully
fill sample bottles from the container.
8.2.4 After collecting the sample, cap carefully to avoid spillage, and agitate by
hand for 1 min.
556-13
-------
8.3 SAMPLE STORAGE/HOLDING TIMES
8.3.1 Samples must be iced or refrigerated at 4 °C and maintained at these
conditions away from light until extraction. Samples must be extracted
within 7 days of sampling. However, since aldehydes are subject to decay
in stored samples, all samples should be derivatized and extracted as soon
as possible.
NOTE: A white or blue precipitate is likely to occur. This is normal and
does not indicate any problem with sample collection or storage.
8.3.2 Extracts (Sect. 11.1.11) must be stored at 4 °C or less away from light in
glass vials with Teflon-lined caps. Extracts must be analyzed within 14
days of extraction.
8.4 FIELD REAGENT BLANKS - Processing of a field reagent blank (FRB) is
required along with each sample set. A sample set is composed of the samples
collected from the same general sampling site at approximately the same time. Field
reagent blanks are prepared at the laboratory before sample vials are sent to the field.
At the laboratory, fill a sample container with reagent water (Sect. 7.2), add sample
preservatives as described in Section 8.1.2, seal and ship to the sampling site along
with the empty sample containers. FRBs should be confirmed to be free (less than
1/2 the MRL) of all method analytes prior to shipping them to the field. Return the
FRB to the laboratory with filled sample bottles. DO NOT OPEN THE FRB AT
THE SAMPLING SITE. If any of the analytes are detected at concentrations equal
to or greater than 1/2 the MRL , then all data for the problem analyte(s) should be
considered invalid for all samples in the shipping batch.
9. QUALITY CONTROL
9.1 Each laboratory that uses this method is required to operate a formal quality control
(QC) program. Minimum QC requirements are initial demonstration of laboratory
capability (which includes calculation of the MDL), analysis of laboratory reagent
blanks, laboratory fortified blanks, field reagent blanks, laboratory fortified sample
matrices, and QC samples. Additional QC practices are encouraged.
9.2 INITIAL DEMONSTRATION OF CAPABILITY (IDC) -- Requirements for the
initial demonstration of laboratory capability are described in the following sections
and summarized in Section 17, Table 7.
9.2.1 Initial demonstration of low system background. (See Sect. 9. 3)
9.2.2 Initial demonstration of precision. Prepare, derivatize, extract, and analyze
4-7 replicate LFBs fortified at 20 ug/L, or other mid-range concentration,
over a period of at least 2 days. Generating the data over a longer period
556-14
-------
of time, e.g. 4 or 5 days may produce a more realistic indication of day to
day laboratory performance. The relative standard deviation (RSD) of the
results of the replicate analyses must be less than 20%.
9.2.3 Initial demonstration of accuracy. Using the same set of replicate data
generated for Section 9.2.2, calculate average recovery. The average
recovery of the replicate values must be within ± 20% of the true value.
9.2.4 MDL(1) determination. Replicate analyses for this procedure should be
done over at least 3 days (both the sample derivatization/extraction and the
GC analyses should be done over at least 3 days). Prepare at least 7
replicate LFBs at a concentration estimated to be near the MDL. This
concentration may be estimated by selecting a concentration at 2-5X the
noise level. Analyze the seven replicates through all steps of Section 11.
Calculate the MDL
MDL=St(n.! i . alpha = 0.99)
where:
= Student's t value for the 99% confidence level with
= 0.99)
n-1 degrees of freedom
n = number of replicates
S = standard deviation of replicate analyses.
NOTE: Do not subtract blank values when performing MDL calculations.
9.2.5 Minimum Reporting Level (MRL) — Although an MDL can be calculated
for analytes that commonly occur as background contaminants, the
calculated MDLs should not be used as the MRL for each analyte.
Method analytes that are seen in the background (typically formaldehyde,
acetaldehyde) should be reported as present in field samples, only after
careful evaluation of the background levels. It is recommended that a
MRL be established at the mean LRB concentration + 3a, or three times
the mean LRB concentration, whichever is greater. This value should be
calculated over a period of time, to reflect variability in the blank
measurements. It is recommended that this value be used as a minimum
reporting level in order to avoid reporting false positive results.
9.3 LABORATORY REAGENTS BLANKS (LRB) -- Each time a set of samples is
extracted or reagents are changed, a LRB must be analyzed. If within the retention
time window of any analyte, the LRB produces a peak that would prevent the
determination of that analyte, determine the source of contamination and eliminate
the interference before processing samples. Because background contamination is a
556-15
-------
significant problem for several method analytes, it is highly recommended that the
analyst maintain a historical record of LRB data. If target analytes are detected in the
LRB at concentrations equal to or greater than 1/2 the MRL (Sect. 9.2.5), then all
data for the problem analyte(s) should be considered invalid for all samples in the
extraction batch.
9.4 CONTINUING CALIBRATION CHECK/LABORATORY FORTIFIED BLANK --
Since this methodology is based on procedural standard calibration, a LFB and the
calibration check sample (CCC) are prepared and analyzed in the same manner.
Laboratory fortified blank QC requirements are therefore omitted. Calibration
procedure options and the QC acceptance criteria associated with them are fully
described in Sect 10.3. Please refer to that section for these criteria.
9.5 INTERNAL STANDARD—The analyst must monitor the IS response peak area of
all injections during each analysis day. A mean IS response is determined from the
five point calibration curve. The IS response for any chromatographic run should not
deviate from this mean IS response by more than 30%. If a deviation greater than
30% occurs with an individual extract inject a second aliquot of that extract.
9.5.1 If the reinjected aliquot produces an acceptable internal standard response,
report results for that aliquot.
9.5.2 If a deviation of greater than 30% is obtained for the reinjected extract, the
analyst should check the calibration by analyzing the most recently
acceptable calibration standard. If the calibration standard fails the criteria
of Section 9.5, recalibration is in order per Section 10. If the calibration
standard is acceptable, extraction of the sample should be repeated
provided the sample is still within the holding time. Otherwise, report
results obtained from the reinjected extract, but annotate as suspect.
9.6 SURROGATE RECOVERY-The surrogate standard is fortified into the aqueous
portion of all calibration standards, samples, FRBs and LRBs. The surrogate is a
means of assessing method performance from derivatization to final chromatographic
measurement.
9.6.1 When surrogate recovery from a sample, blank, or CCC is <70% or
>130%, check (1) calculations to locate possible errors, (2) standard
solutions for degradation, (3) contamination, and (4) instrument
performance. If those steps do not reveal the cause of the problem,
reanalyze the extract.
9.6.2 If the extract reanalysis meets the surrogate recovery criterion, report only
data for the reanalyzed extract.
556-16
-------
9.6.3 If the extract reanalysis fails the 70-130% recovery criterion, the analyst
should check the calibration by analyzing the most recently acceptable
calibration standard. If the calibration standard fails the criteria of Section
9.6.1, recalibration is in order per Section 10. If the calibration standard is
acceptable, it may be necessary to extract another aliquot of sample if
sample holding time has not been exceeded. If the sample reextract also
fails the recovery criterion, report all data for that sample as suspect.
9.7 LABORATORY FORTIFIED SAMPLE MATRIX (LFM)
9.7.1 Within each analysis set, a minimum of one field sample is fortified as a
LFM for every 20 samples analyzed. The LFM is prepared by spiking a
sample with an appropriate amount of the calibration standard. The
concentrations 5, 10, and 20 ug/L are suggested spiking concentrations.
Select the spiking concentration that is closest to, and at least twice the
matrix background concentration. Use historical data or rotate through the
designated concentrations to select a fortifying concentration. Selecting a
duplicate vial of a sample that has already been analyzed, aids in the
selection of appropriate spiking levels.
9.7.2 Calculate the percent recovery (R) for each analyte, after correcting the
measured concentration, A, from the fortified sample for the background
concentration, B, measured in the unfortified sample, i.e.,
c
where C is the fortified concentration. Compare these values to
control limits appropriate for reagent water data collected in the
same fashion.
9.7.3 Recoveries may exhibit a matrix dependence. For samples fortified at or
above their native concentration, recoveries should range between 70 -
130%. If the accuracy of any analyte falls outside the designated range,
and the laboratory performance for that analyte is shown to be in control,
the accuracy problem encountered with the fortified sample is judged to be
matrix related, not system related. The result for that analyte in the
unfortified sample is labeled suspect/matrix to inform the data user that the
results are suspect due to matrix effects. Repeated failure to meet the
suggested recovery criteria indicates potential problems with the extraction
procedure and should be investigated.
556-17
-------
9.8 FIELD DUPLICATES — Within each analysis batch, a minimum of one field sample
should be analyzed in duplicate. Duplicate sample analyses serve as a check on
sampling and laboratory precision.
9.8.1 Calculate the relative percent difference (RPD) for duplicate measurements
(FD1 and FD2) as shown below.
RPD - FD1~FD2 .(100)
(FDl+FD2)/2
9.8.2 Relative percent differences for laboratory duplicates should fall in the
range of ± 30 %. Greater variability may be observed for target analytes
with concentrations near their MRL.
9.9 QUALITY CONTROL SAMPLE (QCS) - At least quarterly, analyze a QCS from
an external source. If measured analyte concentrations are not of acceptable
accuracy (60-140% of the expected value), check the entire analytical procedure to
locate and correct the problem source.
9.10 ASSESSING(Z/£) RATIOS -- In addition to monitoring analyte response from
CCC/LFB, the ratio of the peak areas of each isomer pair should be monitored.
When samples and standards are processed and analyzed by exactly the same
procedure, the ratio of the (Z/E) isomers produced by each method analyte will be
reproducible. This information can be used as a QC check to avoid biased results
caused by an interferant with one isomer of the pair. Calculate and record the ratio of
the peak area of the first eluting isomer (designated (E)) to the second eluting isomer
(designated (Z)~). This ratio will be used in data evaluation Section 12.4.
10. CALIBRATION AND STANDARDIZATION
10.1 Demonstration and documentation of acceptable initial calibration is required before
any samples are analyzed, and is required intermittently throughout sample analysis.
After initial calibration is successful, the analyst may choose one of two options for
maintaining on-going calibration. The first option is to verify the initial calibration
daily using a minimum of 2 calibration standards. The other option is daily
calibration of the method with all 5 calibration standards. These options are further
described in Section 10.3.
10.2 INITIAL CALIBRATION CURVE
10.2.1 Establish GC operating parameters equivalent to the suggested
specifications in Section 17, Table 1. The GC system must be calibrated
using the internal standard (IS) technique. Other GC columns or GC
conditions may be used if equivalent or better performance can be
demonstrated.
556-18
-------
10.2.2 Five calibration standards are recommended to calibrate over the range of
approximately 2-40 ug/L. The lowest level standard will depend upon the
level of blank contamination for each analyte (Sect. 7.11.6).
10.2.3 Prepare each calibration standard by the procedural standard calibration
method. Method analytes are fortified into reagent water and carried
through the entire extraction and derivatization procedure described in
Section 11.
10.2.4 Inject 1 uL of each calibration standard extract into the GC and tabulate
peak area response and concentration for each analyte and the internal
standard. NOTE: The formaldehyde peak will be much larger (for the
same concentration) than the other analyte peaks. The formaldehyde peak
may need to be attenuated on some instruments/data systems to avoid
signal saturation.
10.2.5 (Z/E) ISOMERS — Two isomers, referred to as (E) and (Z), are formed for
most asymmetrical carbonyl compounds derivatized with PFBHA.
Chromatographic resolution is usually obtained with the columns suggested
in Section 6.6 for acetaldehyde, propanal, butanal, pentanal, hexanal,
heptanal, octanal, and crotonaldehyde (see chromatograms in Fig. 1 and
Fig 2). With dicarbonyl species such as glyoxal and methyl glyoxal, (E)
and (Z) isomerism occurs from oxime formation with both carbonyl groups,
increasing the number of isomers. The demonstration data included in this
method has used two distinct isomer peaks each for glyoxal and methyl
glyoxal. Use one of the following methods for both calibration and
quantitation of each method analyte.
(a) Use the sum of the isomer peak areas for each constituent for both
calibration and quantitation.
(b) Use the peak area of each individual isomer to independently
calculate a concentration for each isomer. Then average the amount
of the two isomers to report one value for the analyte.
10.2.6 Generate a calibration curve for each analyte by plotting the area ratios
(A/A^) against the concentration ratios (Ca/Cis) of the five calibration
standards where:
Aa is the peak area of the analyte (or analyte isomer pair),
AiS is the peak area of the internal standard,
Ca is the concentration of the analyte, and
C:a is the concentration of the internal standard.
556-19
-------
10.2.7 This curve must always be forced through zero and can be defined as either
first or second order. Forcing zero allows for a better estimate of the
background level of method analytes.
10.2.8 A data system is recommended to collect the chromatographic data, to
calculate relative response factors, and calculate either linear or second
order calibration curves.
10.2.9 VERIFICATION OF CALIBRATION STANDARD MATERIALS --
Analyze a LFB prepared from standard materials from a source other than
those used to prepare the initial calibration curve (Sect. 3.8, QCS).
Calculate the concentration of this QCS from the calibration curve. The
calculated concentration of the QCS must agree within 60-140% of its true
value. This step verifies the validity of calibration standard materials and
the calibration curve prior to sample analyses.
10.3 OPTIONS FOR ON-GOING CALIBRATION
The time, temperature, pH, and PFBHA concentration will all affect the rate,
efficiency and reproducibility of the derivatization reaction. It is critical that those
parameters be controlled. Calibration frequency will depend upon the laboratory's
ability to control these parameters so that continuing calibration check standard
criteria can be met. Some laboratories may find it more productive to prepare and
analyze a calibration curve with each batch of samples. A batch of samples for this
methodology should not exceed 20 samples, including field samples, FRBs,
laboratory duplicates, and fortified sample matrices.
10.3.1 CONTINUING CALIBRATION CHECK (CCC) OPTION-The analyst
must periodically verify calibration during the analysis of samples in order
to ensure accuracy of analytical results. Prepare a minimum of one low-
level (suggested concentration 2-5 ug/L) and one mid-level (suggested
concentration 10-30 ug/L) calibration standard with each batch of samples.
Verify calibration using these two standards, prior to analyzing any of the
sample extracts from the batch. In addition, reanalyze one of these two
standard extracts after every tenth sample extract, and after the last sample
in an analysis batch to ensure instrument stability throughout the analysis
batch. Recovery must be within 70-130% of the true value for the mid-
level standard, and within 50-150% of the true value for the low-level
standard.
10.3.2 DAILY CALIBRATION OPTION - The analyst may choose to create a
new calibration curve for each batch of samples by preparing and analyzing
a standard at all five calibration concentrations, with each batch of samples.
If this option is selected, the calibration standard extracts should be
analyzed prior to the analysis of sample extracts. To ensure that sensitivity
556-20
-------
and performance of the method has not changed significantly between
sample batches , or changed since the IDC, the following performance
check is required. The response (peak area) of the internal standard,
surrogate and each method analyte in the mid-level standard (suggested
concentration 10-30 ug/L), must be within 50-150% of the mean peak area
for that analyte in the initial demonstration of precision replicates (Sect.
9.2.2). One of the calibration standard extracts must be reanalyzed after
every tenth sample extract, and after the last sample in an analysis batch to
ensure instrument stability throughout the analysis batch. Recovery must
be within 70 to 130% of the true value for mid- and high-level calibration
standards, and within 50-150% of the true value for the low-level standard
(suggested concentration 2-5 ug/L).
11 PROCEDURE
11.1 SAMPLE EXTRACTION — Once samples have been opened, process the samples
straight through to step 11.1.11. There is no known "safe" stopping point once
sample processing has begun. Samples are derivatized and extracted in the sample
bottle in which they were collected. Transferring the sample to another container for
derivatization and extraction has been shown to cause a loss of method analytes.
11.1.1 Remove the samples from storage and allow them to equilibrate to room
temperature.
11.1.2 Remove 10 mL of sample and discard. Mark the level of the remaining
sample volume on the outside of the bottle, for later sample volume
determination.
11.1.3 Add 200 mg KHP to the sample for pH adjustment.
11.1.4 Add 20 uL surrogate solution (Sect 7.11.2.2).
11.1.5 Add 1 mL of freshly prepared PFBHA Reagent as per Section 7.10.1. Cap
and swirl gently to mix.
11.1.6 Place all samples in a constant-temperature water bath set at 35 ± 2°C for
2 hrs. Remove vials and cool to room temperature for 10 min.
11.1.7 To each vial add 0.05 mL (approximate 2 to 4 drops) of concentrated
sulfuric acid. This prevents the extraction of excess reagent, which will
cause chromatographic interferences.
11.1.8 Add 4 mL of hexane that contains the internal standard (as per Sect.
7.11.1.2).
556-21
-------
11.1.9 Shake manually for 3 min. Let stand for approximately 5 min to permit
phases to separate.
11.1.10 Draw off hexane layer (top layer) using a clean disposable Pasteur pipet
for each sample into a smaller 8 mL vial containing 3 mL 0.2 N sulfuric
acid. Shake for 30 sec and let stand for 5 min for phase separation.
NOTE: This acid wash step further reduces the reagent and other
interferants from the final extract. Do not skip this step.
11.1.11 Draw off top hexane layer using another clean disposable pipet for each
sample and place in two 1.8 mL autosampler vials per sample. Store extra
autosampler vials as a backup extract. Extracts may be stored for up to 14
days at 4 °C.
11.1.12 Sample Volume Determination — Discard remaining water sample and
hexane in each sample bottle. Fill with water to the level indicated by the
mark made in Section 11.1.2. Pour the water into a 25 mL graduated
cylinder and measure the volume to the nearest mL. Record the sample
volume for each sample.
Alternately, if a laboratory has control over the brand and style of the
sample bottles being used, the exact volume of a number of bottles from
the same manufacturer and lot may be measured, and the average bottle
volume minus 10 mL may be used as the sample volume for all samples
using the same lot of sample bottles. A minimum of 10 % of the sample
bottles obtained from the same manufacturer, from the same lot should be
measured.
11.2 GAS CHROMATOGRAPHY
11.2.1 Analyze the extracts by GC/ECD. Table 1 (Sect. 17) summarizes
recommended GC operating conditions and retention times observed using
this method. Figure 1 illustrates the performance of the recommended
primary column with the method analytes. Figure 2 illustrates the
performance of the recommended confirmation column with the method
analytes. Other GC columns or chromatographic conditions may be used if
the requirements of Section 9 are met.
11.2.2 The width of the retention time window used to make identifications
should be based on measurements of actual retention time variations of
standards over the course of time. Plus or minus three times the standard
deviation of the retention time for a compound can be used to calculate a
suggested window size; however the experience of the analyst should
weigh heavily in the interpretation of chromatograms.
556-22
-------
11.2.3 If an analyte peak area exceeds the range of the calibration curve, the
extract may be diluted with the hexane extraction solvent (that contains the
internal standard) and reanalyzed. Incorporate the dilution factor into final
concentration calculations. The analyst must not extrapolate beyond the
calibration range established.
12. DATA ANALYSIS AND CALCULATIONS
12.1 Identify the method analytes in the sample chromatogram by comparing the retention
time of the suspect peak to the retention time of an analyte peak (or isomer peaks) in
a calibration standard or the laboratory fortified blank.
12.2 Calculate the analyte concentrations using the first or second order calibration
curves generated as described in Section 10.
12.3 For any analytes that are found, adjust the calculated concentration to reflect the true
sample volume determined in Section 11.1.12.
12.4 Prior to reporting the data, the chromatogram should be reviewed for any incorrect
peak identification or poor integration. If a confirmation column has been used, all
identifications should be verified using the retention time data from that analysis. In
addition, the (Z/E) isomer ratio should be within 50% of the ratio observed in
standards. If the (Z/E) ratio does not meet these criteria, it is likely that an
interferant occurred at the retention time of one of the isomer peaks. In this case,
the amount indicated by the lower of the 2 isomer peaks should be reported. (This
may require that the analyst recalculate the analyte amount using individual isomer
peaks for quantitation.) If one peak of the isomeric pair is missing, the identification
is not confirmed and should not be reported.
12.5 Analyte concentrations are reported in ug/L.
13. METHOD PERFORMANCE
13.1 Precision and accuracy data are presented in Section 17. Data are presented for
three water matrices: reagent water (Table 2), chlorinated "finished" surface water
(Table 4), and untreated "raw" surface water (Table 5). These data, as well as the
MDL data in Table 3, were generated in two laboratories. Data in Table 2 and
column A of Table 3, were generated in one laboratory, while data in column B of
Table 3 and in Tables 4 and 5 were generated in a second laboratory. Method
performance in both laboratories was similar.
13.2 DERIVATIZATION PARAMETERS - This method is a procedural standard
method that will generate accurate and precise results when used as written. The
time, temperature, pH, and PFBHA concentration will all affect the rate, efficiency
556-23
-------
and reproducibility of the derivatization reaction. It is critical that those parameters
be controlled. Calibration frequency will depend upon the laboratory's ability to
control these parameters. Some laboratories may need to prepare and analyze a
calibration curve with each batch of samples. Of all the method analytes, glyoxal,
methyl glyoxal, benzaldehyde, crotonaldehyde and cyclohexanone are the most
difficult to derivatize. Poor sensitivity for any of these compounds indicates that
there may be a problem with the reaction conditions. Measurements of nonanal,
decanal, glyoxal and methyl glyoxal appear to be less precise than the measurement
of other analytes.
13.3 The importance of low background levels of formaldehyde and acetaldehyde cannot
be overemphasized. Some laboratories or reagent waters may also contain
background amounts of other method analytes. Care must be taken to avoid
reporting false positive results that result from background contamination.
13.4 The importance of proper sample collection and preservation also cannot be
overemphasized. Holding time studies in various matrices showed better than 70%
recovery of all method analytes when samples were collected, preserved, and stored
according to Section 8, and analyzed within 7 days. There were variations in the
recovery of analytes from fortified samples from different matrices. Therefore, it is
strongly recommended that samples be analyzed as soon as possible after collection.
The data in Section 17, Table 6 illustrates the dramatic difference between a
preserved and a non-preserved sample.
13.5 Data for crotonaldehyde is not listed in Section 17, Tables 4-6, because it was not
included in the standard mixtures being used at the time that those data were
collected. However, crotonaldehyde was included in many other studies not
presented here, and its performance was similar to other method analytes.
14. POLLUTION PREVENTION
14.1 This method uses a micro-extraction procedure which requires very small quantities
of organic solvents.
14.2 For information about pollution prevention that may be applicable to laboratory
operations, consult "Less is Better: Laboratory Chemical Management for Waste
Reduction" available from the American Chemical Society's Department of
Government Relations and Science Policy, 1155 16th Street N.W., Washington,
D.C.20036.
15. WASTE MANAGEMENT
15.1 Due to the nature of this method tHere is little need for waste management. Only
small volumes of solvents are used. The matrices of concern are finished drinking
water or source water. However, the Agency requires that laboratory waste
556-24
-------
management practices be conducted consistent with all applicable rules and
regulations, and that laboratories protect the air, water, and land by minimizing and
controlling all releases from fume hoods and bench operations. Also, compliance is
required with any sewage discharge permits and regulations, particularly the
hazardous waste identification rules and land disposal restrictions. For further
information on waste management, consult "The Waste Management Manual for
Laboratory Personnel" also available from the American Chemical Society at the
address in Section 14.2.
16. REFERENCES
1. Glaser, J.A., D.L. Foerst, G.D. McKee, S.A. Quave, and W.L. Budde, "Trace
Analyses for Wastewaters," Environ. Sci. Technol. 1981, 15, 1426-1435.
2. Standard Method Number 6252B, "PFBFIA Liquid-Liquid Extraction Gas
Chromatographic Method", Standard Methods for the Examination of Water and
Wastewater. pp. 6-77 to 6-83, American Public Health Assoc., Washington, D.C.,
1995.
3. Sclimenti, M.J., S.W. Krasner, W.H. Glaze, and H.S. Weinberg,"Ozone Disinfection
By-Products: Optimization of the PFBHA Derivatization Method for the Analysis of
Aldehydes", In Advances in Water Analysis and Treatment, Proc. AWWA Water
Quality Technology Conf, 1990,pp 477-501.
4. Glaze, W.H. and H.S. Weinberg, Identification and Occurrence of Ozonation By-
Products in Drinking Water. American Water Works Assoc. Research Foundation,
Denver, CO., 1993, pp!9-22.
5. "OSHA Safety and Health Standards, General Industry," (29CRF1910).
Occupational Safety and Health Administration, OSHA 2206, (Revised, Jan. 1976).
6. ASTM Annual Book of Standards, Part II, Volume 11.01, D3370-82, "Standard
Practice for Sampling Water," American Society for Testing and Materials,
Philadelphia, PA, 1986.
7. "Carcinogens-Working with Carcinogens", Publication No. 77-206, Department of
Health, Education, and Welfare, Public Health Service, Center for Disease Control,
National Institute of Occupational Safety and Health, Atlanta, Georgia, August
1977.
8. "Safety In Academic Chemistry Laboratories", 3rd Edition, American Chemical
Society Publication, Committee on Chemical Safety, Washington, D.C., 1979.
556-25
-------
17. TABLES. DIAGRAMS. FLOWCHARTS. AND VALIDATION DATA
TABLE 1. CHROMATOGRAPHIC CONDITIONS AND RETENTION DATA FOR
ANALYTE DERIVATIVES
ANALYTE
Formaldehyde
Acetaldehyde (E)
Acetaldehyde (Z)
Propanal (E)
Propanal (Z)
Butanal (E)
Butanal (Z)
Crotonaldehyde (E)
Crotonaldehyde (Z)
Pentanal (E)
Pentanal (Z)
Hexanal (E)
Hexanal (Z)
Cyclohexanone
Heptanal (E)
Heptanal (Z)
Octanal (E)
Octanal (Z)
Benzaldehyde
Nonanal (E)
Decanal
Column A
RT (min)
10.69
14.23
14.49
17.36
17.62
20.58
20.81
22.67
23.00
23.82
24.01
26.99
27.16
29.42
30.05
30.16
33.01
33.09
33.88
35.89
38.63
Peak #, Fig 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
19
20
21
22
23
Column B
RT (min)
13.35
16.71
17.00
19.59
19.86
22.62
22.88
24.79
25.31
25.54
25.70
28.74
28.93
31.40
31.64
31.78
34.45
34.57
36.74
37.18
39.80
Peak #, Fig.2
2
3
4
5
6
7
8
not shown
not shown
9
10
11
12
13
14
15
16
18
19
20
21
556-26
-------
ANALYTE
Glyoxal (peakl)
Glyoxal (peak 2)
Methyl glyoxal
(peak 1)
Methyl glyoxal
(peak 2)
1 .2 dibromopropane
(Internal standard)
2',4',5' trifluoro-
acetophenone
(Surrogate)
Column A
RT (min)
40.87
41.09
41.22
41.88
6.14
32.84
Peak #, Fig 1
24
25
26
27
1
18
Column B
RT (min)
43.74
43.86
43.86
44.38
7.84
34.45
Peak #, Fig.2
22
23
24
25
1
17
Column A- DB-5ms, 30 m x 0.25mm i.d., 0.25 um film thickness, injector temp. 220 °C, head
pressure 15 psi, detector temp. 300 °C, splitless injection, 1 min split delay.
Temperature program: 50 °C for 1 min, program at 4°C/min to 220 °C, program at
20 °C/min to 250 °C and hold at 250 °C for 10 min. (Figure 1)
Column B- Rtx-1701, 30 m x 0.25 mm i.d., 0.25 um film thickness, injector temp. 180 °C, head
pressure 15 psi, detector temp. 300°C,splitless injection, 1 min split delay.
Temperature program: 50 °C for 1 min, program at 4°C/min to 220 °C, program at
20 °C/min to 250 °C and hold at 250 °C for 10 min. (Figure 2)
Carrier gas- Helium
Detector gas- P5 Argon/Methane
556-27
-------
TABLE 2. PRECISION AND ACCURACY IN REAGENT WATER (n=8)
ANALYTE
Formaldehyde
Acetaldehyde
Propanal
Butanal
Crotonaldehyde
Pentanal
Hexanal
Cyclohexanone
Heptanal
Octanal
Benzaldehyde
Nonanal
Decanal
Glyoxal
Methyl glyoxal
Fortified
Cone. (ug/L)
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Mean Cone.
Measured
(ug/L)
19.5
19.2
19.5
19.8
19.8
20.0
19.7
19.2
19.2
19.1
19.1
18.8
18.7
18.4
17.9
Standard
Deviation
0.792
0.784
0.873
1.12
1.11
1.10
1.25
0.729
1.37
1.32
0.362
1.09
1.20
0.571
1.36
Relative
Std Dev (%)
4.0
4.1
4.5
5.6
5.6
5.5
6.4
3.8
7.1
6.9
1.9
5.8
6.4
3.1
7.6
Mean
Accuracy
(%)
98
96
97
99
99
100
98
96
98
95
95
94
93
92
90
a- Analyzed over 2 days on the primary chromatographic column (DB-5ms).
556-28
-------
TABLE 3. METHOD DETECTION LIMITS IN REAGENT WATER (n=8)
ANALYTE
Formaldehyde
Acetaldehyde
Propanal
Butanal
Crotonaldehyde
Pentanal
Hexanal
Cyclohexanone
Heptanal
Octanal
Benzaldehyde
Nonanal
Decanal
Glyoxal
Methyl glyoxal
Fortified
Cone. (ug/L)
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
MDL(ug/L)
Column A
0.36
0.21
0.41
0.35
0.28
0.47
0.42
0.29
0.43
0.60
0.31
0.74
1.0
0.59
0.81
MDL(ug/L)
Column B
0.11
0.14
0.06
0.12
0.09
0.17
0.35
0.10
0.71
0.12
0.06
0.40
0.82
0.21
0.22
Column A= DB-5ms
Column B=Rtx-1701
556-29
-------
TABLE 4. PRECISION AND ACCURACY IN CHLORINATED TAP WATER(n=4)
ANALYTE
Formaldehyde
Acetaldehyde*
Propanal
Butanal
Pentanal
Hexanal
Cyclohexanone
Heptanal
Octanal
Benzaldehyde
Nonanal
Decanal
Glyoxal
Methyl glyoxal
Fortified
Cone. (ug/L)
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Mean Cone.
Measured
(ug/L)
21.6
19.6
19.5
20.2
20.3
20.2
20.7
20.5
20.3
21.2
20.0
19.8
22.3
21.0
Standard
Deviation
0.197
0.263
0.178
0.254
0.318
0.382
0.247
3.073
0.979
0.432
0.828
1.203
0.783
0.775
Relative
Std Dev (%)
0.9
1.3
0.9
1.3
1.6
1.9
1.2
2.5
4.8
2.0
4.2
6.1
3.5
3.7
Mean
Accuracy
(%)
108
98
98
101
101
101
103
103
102
106
100
99
112
105
a- Analyzed on the primary chromatographic column (DB-5ms).
*- Values taken from the confirmation column.
556-30
-------
TABLE 5. PRECISION AND ACCURACY IN UNTREATED SURFACE WATER (n=4)
ANALYTE
Formaldehyde
Acetaldehyde*
Propanal
Butanal
Pentanal
Hexanal
Cyclohexanone
Heptanal
Octanal
Benzaldehyde
Nonanal
Decanal
Glyoxal
Methyl glyoxal
Fortified
Cone. (ug/L)
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Mean Cone.
Measured
(ug/L)
19.4
20.0
19.5
18.9
18.7
18.2
17.3
18.0
18.3
17.2
18.4
18.1
20.5
23.3
Standard
Deviation
0.941
0.891
0.966
0.965
1.00
1.080
0.943
0.995
1.017
1.118
0.881
0.908
1.10
0.799
Relative
Std Dev (%)
4.8
4.4
5.0
5.1
5.4
6.0
5.4
5.5
5.6
6.5
4.8
5.0
5.4
3.4
Mean
Accuracy
(%)
97
100
97
94
93
91
87
90
92
86
92
90
102
116
a- Analyzed on the primary chromatographic column (DB-5ms).
*- Values taken from the confirmation column.
556-31
-------
TABLE 6. HOLDING TIME DATA FOR SAMPLES FROM AN UNTREATED
SURFACE WATER SOURCE, FORTIFIED WITH METHOD ANALYTES
AT 20 ug/L, WITH AND WITHOUT COPPER SULFATE BIOCIDE
Analyte
Formaldehyde
Acetaldehyde
Propanal
Butanal
Pentanal
Hexanal
Cyclohexanone
Heptanal
Octanal
Benzaldehyde
Nonanal
Decanal
Glyoxal
Methyl glyoxal
% Recovery without Copper
Sulfate
DayO
104
96
94
92
87
83
94
83
82
94
72
50
103
108
Day 6
144
23
21
20
19
21
99
20
18
83
15
<10
98
68
Day 14
569
21
19
18
16
17
84
16
<10
74
<10
<10
37
<10
Day 21
619
24
22
21
21
22
78
17
11
79
<10
<10
<10
<10
% Recovery with Copper Sulfate
DayO
105
98
98
99
96
93
96
96
99
98
104
107
106
111
Day 6
105
99
98
98
98
97
101
94
96
100
98
101
108
105
Day 14
109
103
103
102
100
100
98
97
96
104
92
93
106
94
Day 21
106
98
96
91
94
92
94
91
93
92
84
82
90
73
- All samples were stored headspace free at 4 °C.
- Values at all time points are the mean of 5 replicate analyses. RSDs for replicate analyses of
samples containing copper sulfate were <10%. RSDs for unpreserved samples were higher due to
the degradation process.
556-32
-------
TABLE 7. INITIAL DEMONSTRATION OF CAPABILITY REQUIREMENTS
Reference
Sect. 9.3
Sect 9.2.2
Sect. 9.2.3
Sect. 9.2.4
Sect. 9.2.5
Requirement
Initial
Demonstration of
Low System
Background
Initial
Demonstration of
Precision (IDP)
Initial
Demonstration of
Accuracy
Method
Detection Limit
(MDL)
Determination
Minimum
Reporting Levels
(MRLs)
Specification and Frequency
Analyze method blank and
determine that all target
analytes are below 1/2 the MRL
prior to performing IDC
Analyze 4-7 replicate LRBs
fortified at 20.0 ug/L (or mid
cal.) on at least 2 different days
Calculate average recovery of
IDP replicates
a) select a fortifying level at 2 -
5 x the noise level
b) analyze 7 replicates in
reagent water taken thru all
steps
c) calculate MDL via equation -
do not subtract blank
d) replicate extractions and
analyses must be conducted
over at least 3 days
MRLs should be established for
all analytes during IDC, and be
updated as additional LRB data
is available.
Acceptance Criteria
The LRB concentration
must be < 1/2 of the
intended MRL
RSD must be < 20 %
Mean recovery ± 20% of
true value
Establish the MRL for
each analyte, as the LRB
concentration + 3 a or 3
times the mean LRB
concentration, whichever
is greater.
556-33
-------
TABLE 8. QUALITY CONTROL REQUIREMENTS (SUMMARY)
Reference
Sect. 10.2
Sect. 9.3
Sect. 10.3.1
Sect. 10.3.2
Sect. 8.4
Sect. 9.5
Requirement
Initial
Calibration
Laboratory
Reagent Blank
(LRB)
Continuing
Calibration
Check (CCC)
Option
Daily
Calibration
Option
Field Reagent
Blanks (FRB)
Internal
Standard (IS)
Specification and Frequency
Use internal standard technique to
generate curve with five standards
that span the approximate range
of 2-40 ug/L. First or second
order curves must be forced
through zero. Either sum E/Z
isomer areas or average the
amount of each isomer.
Calculate E/Z ratios for analytes.
Run QCS.
Include LRB with each extraction
batch (up to 20 samples).
Analyze prior to analyzing
samples and determine to be free
of interferences.
Verify initial calibration by
running CCCs prior to analyzing
samples, after 10 samples, and
after the last sample.
Calibrate daily, but verify that
sensitivity and performance have
not changed significantly since
IDC.
1 per shipping batch
1,2-Dibromopropane is added to
all samples, blanks and standards
Acceptance Criteria
QCS must agree within
60-140 %.
Lowest concentration
should be near MRL.
All analytes < 1/2 MRL
Recovery for mid-level
CCC must be 70- 130%
of the true value,
recovery for low level
must be 50-150% of the
true value.
Peak areas for IS, SUR,
and method analytes for
mid-level CAL std must
be +/- 50% of the peak
areas obtained for that
CAL std during IDC.
All analytes < 1/2 MRL
IS area counts must be
70 - 130% of the average
initial calibration area
counts
556-34
-------
Reference
Sect. 9.6
Sect. 9.7
Sect. 9.8
Sect. 9.9
Sect.9.10,
Sect. 10.2.5
and
Sect. 12.4
Sect. 8.3.1
Sect. 8.3.2
Requirement
Surrogate
Standard
(SUR)
Laboratory
Fortified
Sample Matrix
(LFM)
Field
Duplicates
Quality
Control
Sample (QCS)
E/Z Isomer
Ratio
Agreement
Sample
Holding Time
Extract
Holding Time
Specification and Frequency
2',4',5' -Trifluoroacetophenone is
added samples, blanks and
standards
Fortify at least one sample per
analysis batch (20 samples or less)
at a concentration close to that in
the native sample.
Extract and analyze at least one
duplicate sample with every
analysis batch (20 samples or less)
Analyze a QCS at least quarterly
from an external/second source.
Calculate the E/Z isomer ratio for
target analytes and compare to
E/Z ratio in initial calibration
Properly preserved samples may
be stored in the dark at 4 ° C for 7
days.
Extracts may be stored in the dark
at 4 °C for 14 days.
Acceptance Criteria
Surrogate recovery must
be 70- 130% of the true
value.
Recoveries not within
70-130% may indicate
matrix effect
Suggested RPD < 30 %
QCS must agree within
60-140 %.
E/Z ratio in standards,
blanks, and samples must
be within ±50% of E/Z
ratio in initial calibration.
Do not report value if
one isomer is missing.
Do not report data for
samples that have
exceeded their holding
time, or that have not
been properly preserved
or stored.
Do not report data for
extracts that have
exceeded their holding
time.
556-35
-------
20000 -
17500 -
15000 -
12500 -
10000 -
7500-
5000-
2500-
l
Column: J&W DBS-MS
0.250 mmx30 m, film thickness 0.25um
Carrier Gas: Helium
Injection: 1 uL Splitless, 220 °C
Inlet Pressure: 15 psi
Program: 50 "C, Hold 1 rain
4°C/minto2200C
20 "C/min to 250 "C, Hold 10 min.
li
10
15
20
25
30
35
40 «ii»
Figure 1
556-36
-------
300-
250-
200-
150-
100-
50-
0-
Column: J&W 1701
0.250 mm x 30 in, film thickness 0.25um
Carrier Gas: Helium
Injection: 1 uL Splitkss, 220 t
Inlet Pressure: 15 psi
Program: 50 °C, Hold 1 min
4°C/minto220DC
20 'C/inin to 250°C, Hold 10 min.
I! !! ]ll II ll
10 12
10 IB za zz
! in |ii iijiiii n ii[ nil ii iij nil
ZO ZB 3D 32
Figure 2
30 3B 40 42 ^4
min
556-37
------- |