EPA/600/R-15/258 I October 2015 I www2.epa.gov/research
United States
Environmental Protection
Agency
Extraction and Analysis of Lewisite 1,
by its Degradation Products, Using
Liquid Chromatography Tandem
Mass Spectrometry (LC-MS/MS)
STANDARD OPERATING PROCEDURE
REVISION 1
Office of Research and Development
National Homeland Security Research Center
-------
DISCLAIMER
The U.S. Environmental Protection Agency through its Office of Research and Development funded and
managed the research described here under contract DE-AC52-07NA27344 to U.S. Department of
Energy by Lawrence Livermore National Laboratory. It has been subjected to the Agency's review and
has been approved for publication. Note that approval does not signify that the contents necessarily
reflect the views of the Agency. Mention of trade names, products, or services does not convey official
EPA approval, endorsement, or recommendation.
Questions concerning this document or its application should be addressed to:
Stuart Willison, Ph.D.
U.S. Environmental Protection Agency
National Homeland Security Research Center
26 W. Martin Luther King Drive, MS NG16 Cincinnati, OH 45268
513-569-7253
Willison.Stuart@epa.gov
ii
-------
ACKNOWLEDGMENTS
We would like to acknowledge the following individuals and organization for their contributions towards
the development and/or review of this method.
United States Environmental Protection Agency (EPA)
Office of Research and Development, National Homeland Security Research Center
Stuart Willison
Romy Campisano
Lawrence Livermore National Laboratory (LLNL)
Forensic Science Center
Carolyn Koester
Deon Anex
Jeremiah Gruidl
iii
-------
EXTRACTION AND ANALYSIS OF LEWISITE 1, BY ITS DEGRADATION
PRODUCTS, USING LIQUID CHROMATOGRAPHY TANDEM MASS
SPECTROMETRY (LC-MS/MS)
TABLE OF CONTENTS
SECTION PGNO.
DISCLAIMER ii
ACKNOWLEDGMENTS iii
LIST OF TABLES AND FIGURES v
LIST OF ACRONYMS AND ABBREVIATIONS vi
1. INTRODUCTION 1
2. SCOPE AND APPLICATION 1
3. SUMMARY OF METHOD 3
4. DEFINITIONS 4
5. INTERFERENCES 5
6. HEALTH AND SAFETY 6
7. EQUIPMENT AND SUPPLIES 6
8. REAGENTS AND STANDARDS 8
9. SAMPLE COLLECTION, PRESERVATION AND STORAGE 11
10. QUALITY CONTROL 13
11. INSTRUMENT CALIBRATION AND STANDARDIZATION 17
12. ANALYTICAL PROCEDURE 19
13. DATA ANALYSIS AND CALCULATIONS 20
14. METHOD PERFORMANCE 20
15. POLLUTION PREVENTION 23
16. WASTE MANAGEMENT 23
17. REFERENCES 24
18. TABLES, FIGURES, AND VALIDATION DATA 26
iv
-------
LIST OF TABLES
Table 1. Materials tested for the Lewisite 1 degradation analysis 27
Table 2. Liquid Chromatography Gradient Conditions 28
Table 3. Recommended Concentrations for LC-MS/MS Calibration Standards 28
Table 4. Scan Segments Associated with Lewisite 1 Degradation Product Analysis by LC-MS/MS (LTQ
Orbitrap) 29
Table 5. Method Detection Limits (MDL) data for CVAOA in Various Matrices 30
Table 6. Surrogate (PAOA) Recovery Data Collected During MDL Study (from Samples in Table 5) 30
Table 7. Sample Holding Time Data for CVAA Spiked on Various Matrices and Analyzed as CVAOA 31
Table 8. PAOA Surrogate Data Collected During Sample Holding Time Study 31
Table 9. Statistical Analysis of Sample Holding Time Study Data for CVAOA (See data in Table 2.) 31
Table 10. CVAOA Concentration Changes in Sample Extracts; Extracts Were Prepared on Day 7 and Reanalyzed
on Day 16 (nine days of storage at 4 ± 2 °C) 32
LIST OF FIGURES
Figure 1. Chemical structures of analytes and surrogates 2
Figure 2. Quality data collected during MDL study 33
Figure 3. LC-MS/MS response vs. analyte concentration in various solvents. Solvents change depending on matrix:
Water, 30% H202 added; Wipe samples, 10 mM HC1 with 30% H202; Sand/soil, 10 mM HC1, 25% MeOH, and
30% II -O - 34
Figure 4. LC-MS/MS response vs. analyte concentration, additional data. Solvents change depending on matrix:
Wipe samples, 10 mM HC1 with 30% H202; Sand/soil, 10 mM HC1, 25% MeOH, and 30% H202 35
-------
LIST OF ACRONYMS AND ABBREVIATIONS
ACN Acetonitrile
AS Analyte Stock Standard (Solution)
ATL Analytical Target Level
CAL Calibration Standard
CAS® Chemical Abstracts Service
CCV Continuing Calibration Verification
CID Collisionally Induced Dissociation
CVAA Chlorovinyl Arsonous Acid
CVAOA Chloro vinyl Arsonic Acid
DQO Data Quality Objective
EPA U.S. Environmental Protection Agency
ESI (+) Electrospray Ionization in Positive Mode
GC-MS Gas Chromatography Coupled with Mass Spectrometry
IDC Initial Demonstration of Capability
IDL Instrument Detection Limit
LI Lewisite 1
LC Liquid Chromatography
LC-MS/MS Liquid Chromatography Coupled with Tandem Mass Spectrometry
LFMS Laboratory Fortified Matrix Spike
LFMSD Laboratory Fortified Matrix Spike Duplicate
LMB Laboratory Method Blank
MeOH Methanol
MDL Method Detection Limit
MRL Minimum Reporting Limit
MRM Multiple Reaction Monitoring
MS Mass Spectrometer(try)
MS/MS Tandem Mass Spectrometry
NE Nebraska Soil
NHSRC National Homeland Security Research Center
NIST National Institute of Standards and Technology
OSHA Occupational Safety and Health Administration
PAA Phenyl Arsonous Acid
PAOA Phenyl Arsonic Acid
PTFE Polytetrafluoroethylene
QC Quality Control
r2 Coefficient of Determination
REC Percent Recovery
RL Reporting Limit
RPD Relative Percent Difference
RPM Revolutions per Minute
SAM Selected Analytical Methods for Environmental Remediation and Recovery
SD Standard Deviation (of replicate analyses)
SDS Safety Data Sheet
SOP Standard Operating Procedure
vi
-------
S/N
Signal to Noise
ss
Surrogate Standard
sss
Stock Standard Solution
TOC
Total Organic Carbon
VOA
Volatile Organic Analysis
X
Average Percent Recovery
-------
Page Intentionally Left Blank
viii
-------
1. INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is responsible for developing tools and methodologies
that will enable the rapid characterization of indoor and outdoor areas and of water systems following a
deliberate/accidental release or a natural disaster. EPA's National Homeland Security Research Center
(NHRSC) published Selected Analytical Methods for Environmental Remediation and Recovery (SAM) (1),
which is a compendium of methods that informs sample collection and analysis during the response to an
all-hazards incident. Lewisite is a dangerous vesicant, which can break down into degradation products
sufficiently persistent and toxic to be of interest during site remediation after a release. If an incident were
to occur, versatile analytical procedures are needed to detect Lewisite (and its degradation products) and to
determine the spread and concentration of both in contaminated areas. For the purpose of identifying target
analytes described within this procedure, Lewisite refers to Lewisite 1 and was analyzed as Lewisite 1 and
its corresponding degradation products. This Standard Operating Procedure (SOP) addresses Lewisite
degradation products known as chlorovinyl arsonous acid (CVAA) and chlorovinyl arsonic acid (CVAOA)
in soil, wipe extracts, and water. The samples were extracted from the corresponding matrix types and
analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). LC-MS/MS affords
laboratories an enhanced capability to analyze specific environmental matrices for chemical warfare agent
degradation products while avoiding complications that may arise from a derivatization process commonly
used for gas chromatography-mass spectrometry (GC-MS) analysis. The derivatization process used for
GC-MS analysis is more involved and complicated because the derivatized Lewisite product can be
problematic when analyzing for Lewisite. For example, GC-MS columns deteriorate rapidly in the presence
of Lewisite, making the analysis technique impracticable. LC-MS/MS analysis is a straightforward process
where many of the complications associated with GC-MS analysis are resolved. Data corresponding to the
analysis of the Lewisite 1 (LI) degradation products by LC-MS/MS are limited in the literature for the
matrices investigated. Because Lewisite 1 degradation products are unique, and neither the parent or
degradation products exist in the environment, these compounds are unambiguous indicators for Lewisite
1. Preliminary testing to include other Lewisite byproducts, such as Lewisite Oxide, were included but are
not presented in this protocol. Further testing is needed to include all degradation products associated with
Lewisite, such as Lewisite Oxide, to ensure that sample analyses do not result in false negatives for Lewisite
detection.
2. SCOPE AND APPLICATION
This method has been developed to address the analysis of Lewisite 1 degradation products in
environmental matrices (water, soil, and wipes) by LC-MS/MS. Because Lewisite 1 hydrolyzes quickly in
the environment to CVAA and then oxidizes more slowly to CVAOA, the standard operating procedure
that has been developed extracts Lewisite 1 degradation products from the matrices of interest and converts
them to CVAOA prior to analysis. The presence of CVAOA and/or CVAA is indicative of Lewisite 1
contamination, as there are no known natural sources for these compounds. Phenyl arsonous acid (PAA),
which can be oxidized to phenyl arsonic acid (PAOA), is used as a surrogate to show that both extraction
and oxidation processes have been successfully implemented. Lewisite composition typically consists of
Lewisite 1, 2, and 3 (90%, 9%, and 1% composition, respectively). Since Lewisite 1 is the major product,
only Lewisite 1 was studied. While not tested by this method, a similar analytical strategy should also be
applicable to the analysis of Lewisite 2 and Lewisite 3. The following analytes have been determined using
this procedure:
1
-------
Analvte
Lewisite I (LI)
Chlorovinyl arsonous acid (CVAA)
Chlorovinyl arsonic acid (CVAOA)
CAS Registry Number®
541-25-3
85090-33-1
64038-44-4
The chemical structures of the analytes and surrogate compounds for this method are presented in
Figure 1.
The protocol was developed for the conversion of Lewisite 1 to its degradation products by oxidation in
water, wipe, and soil extracts and analyzed by LC-MS/MS. Because the degradation products can be formed
only when Lewisite 1 degrades, their presence is consistent with Lewisite 1 being used, thus, indirect
detection for Lewisite 1.
Cl\ /Cl
As
H\
C H
CI
Lewisite I
HO ^ ^OH
As
H\
C H
CI
chlorovinyl arsonous acid
(CVAA)
o
HO As OH
I
CI
chlorovinyl arsonic acid
(CVAOA)
HO^ ,OH
As
HO As OH
phenyl arsonous acid (PAA) phenyl arsonic acid (PAOA)
Figure 1. Chemical structures of analytes and surrogates.
2.1. The method detection limit (MDL) metrics are presented using EPA conventions (2-3).
The detection limit is defined as the statistically calculated minimum concentration that
can be measured with 99% confidence that the reported value is greater than zero (4). The
MDL is compound-dependent and reliant on sample preparation, sample matrix,
concentration of agent in the samples used for MDL determinations, and instrument
performance. MDL studies are performed as an initial demonstration of capability (IDC)
and ongoing demonstration of capability to perform the procedure, including changes in
instrumentation and operating conditions. These studies evaluate whether the reporting
limits (RLs) and calibration standard (CAL) concentrations are appropriate.
2
-------
2.2. This procedure is intended for use by analysts skilled in the operation of LC-MS/MS
instrumentation and the interpretation of the associated data. Due to the inherent
complexities of LC-MS/MS analysis, including the need to relate sample characteristics to
analytical performance, laboratories should update their initial estimates of performance and
should strive to tighten their quality control limits as more experience is gained with this
particular procedure.
2.3. METHOD FLEXIBILITY
Many variants of liquid chromatography (LC) and Tandem Mass Spectrometry (MS/MS)
technology are currently in operation. In addition, variability exists in the sources of
investigated matrices, including wipe materials and composition, soil types, and water
sources. This procedure was developed using an Orbitrap LC-MS/MS system, with
optimized LC conditions. The procedure has been verified using only the specified
equipment and conditions. Other types of LC-MS/MS instrumentation, LC and/or
electrospray ionization (ESI) in positive mode MS/MS conditions, sample collection and
processing steps, and materials can be used for analysis as long as similar performance is
demonstrated, and the quality control measures outlined in this report are implemented.
3. SUMMARY OF METHOD
3.1. The degradation products of Lewisite 1, CVAA and CVAOA, are extracted from soil and
wipes using aqueous-based solvents, oxidized to CVAOA using hydrogen peroxide, and
detected by LC-MS/MS. It is assumed that any residual Lewisite 1 in samples would be
hydrolyzed and oxidized to CVAOA during the extraction process (6). This is an acceptable
analytical strategy because there are no known environmental sources for the Lewisite 1
degradation products (i.e., the presence ofthese compounds is an unambiguous indicator for
the presence of Lewisite 1). LC approaches offer the opportunity to detect the Lewisite 1
hydrolysis and oxidation products directly (5, 6), without the added complication of a
derivatization step, which is required for gas chromatography-mass spectrometry (GC-MS)
analyses (7, 8). The expected rapid hydrolysis of Lewisite 1 in aqueous LC mobile phases
precludes the direct detection of Lewisite 1 itself, but species such as CVAA and CVAOA
are readily detectable by LC-MS/MS. (See section 13.1.1, for data supporting conversion of
Lewisite to its degradation products.) The use of LC-MS/MS provides analytical specificity,
by which degradation products of Lewisite 1 can be more easily measured in the presence
of interfering compounds. In addition, LC-MS/MS is expected to provide better instrument
detection limits (IDLs) (e.g., LC-MS/MS IDL of -0.01 ng/|_iL versus an expected GC-MS
IDL of ~0.5 ng/(iL (7)).
3.2. The target analyte is separated chromatographically and identified by retention time.
Comparison of the sample primary multiple reaction monitoring (MRM) transition to the
known standard MRM transition from reference spectra under identical LC-MS/MS
conditions. Stock standard solutions, made by weighing known masses of
analytes/surrogates into known volumes of solvent, are diluted to concentrations appropriate
to construct (linear) calibration curves. Then, the concentrations of analytes/surrogates
present in the samples are calculated considering the responses of the calibration standards
(i.e., the concentration of each analyte is determined by the instrumentation software using
external calibration). Surrogate analytes are added to samples to monitor extraction
efficiency of the method analytes during the extraction process.
3
-------
4. DEFINITIONS
4.1. ANALYSIS BATCH - A set of samples analyzed on the same instrument within a 24-hour
period and including no more than 20 field samples, beginning and ending with the analysis
of the appropriate continuing calibration verification (CCV) standards. Additional CCVs
may be required depending on the number of samples (excluding quality control (QC)
samples) in the analysis batch and/or the number of field samples.
4.2. CALIBRATION STANDARD (CAL) - A solution prepared from the analyte stock
standard solution (AS) and the surrogate/internal standard(s). The CAL solutions are used
to calibrate the instrument response with respect to analyte concentration.
4.3. COLLISIONALLY INDUCED DISSOCIATION (CID) - The process of converting the
translational energy of the precursor ion into internal energy by collisions with neutral gas
molecules to bring about dissociation into product ions.
4.4. CONTINUING CALIBRATION VERIFICATION (CCV) - A calibration standard
containing the method analytes and surrogate standard(s). The CCV is analyzed
periodically to verify the accuracy of the existing calibration for those analytes at or near
the mid-level concentrations. Low calibration concentrations can be added, in addition to
mid-level concentrations, for further accuracy, but are not required.
4.5. EXTRACTION BATCH - A set of up to twenty field samples (excluding QC samples)
extracted together using the same solvents and surrogate (s).
4.6. LABORATORY FORTIFIED MATRIX SPIKE (LFMS) - A field sample to which known
quantities of the method analytes are added in the laboratory. The laboratory fortified
matrix spike (LFMS) is processed and analyzed exactly like a sample, and its purpose is to
determine 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 sample.
4.7. LABORATORY FORTIFIED SAMPLE MATRIX DUPLICATE (LFMSD) - A
laboratory fortified sample matrix duplicate (LFMSD) of the field sample used to prepare
the LFMS. The LFMSD is fortified and analyzed identically to the LFMS. The LFMSD is
used to assess method precision when the observed concentrations of method analytes are
low.
4.8. LABORATORY METHOD BLANK (LMB) - A blank matrix that is treated exactly the
same as a sample including exposure to all glassware, equipment, solvents and reagents,
and surrogate standards that are used in the analysis batch. The laboratory method blank
(LMB) is used to determine if method analytes or other interferences are present in the
laboratory environment, the reagents, or the apparatus.
4.9. 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.
4
-------
4.10. MINIMUM REPORTING LEVEL (MRL) - The minimum concentration that can be
reported as a quantitated value for a method analyte in a sample following analysis. This
defined concentration can be no lower than the concentration of the lowest calibration
standard for that analyte and can be used only if acceptable QC criteria for this standard
are met.
4.11. PRECURSOR ION - For the purpose of this method, the precursor ion is the protonated
molecule ([M+H]+) or adduct ion of the method analyte. In MS/MS, the precursor ion is
mass-selected and fragmented by collisionally induced dissociation (CID) to produce
distinctive product ions of lower mass.
4.12. PRODUCT ION - For the purpose of this method, a product ion is one of the fragment
ions produced in MS/MS by CID of the precursor ion.
4.13. SAFETY DATA SHEET (SDS) - Written information provided by vendors concerning a
chemical's toxicity, health hazards, physical properties, fire, and reactivity data including
storage, spill, and handling precautions.
4.14. SURROGATE STANDARD (SS) - A pure chemical(s) added to a standard solution in a
known amount(s) and used to measure the relative response of other method analytes that
are components of the same solution. The surrogate standard must be a chemical that is
structurally similar to the method analytes, has no potential to be present in samples, and
is not a method analyte.
4.15. 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.
5. INTERFERENCES
Procedural interferences can be caused by contaminants in solvents, reagents, glassware and other apparatus
that lead to discrete artifacts or elevated baselines in the selected ion current profiles. All of these materials
must routinely be demonstrated to be free from interferences by analyzing Laboratory Method Blanks
(LMBs) under the same conditions as the samples. Subtraction of blank values from sample results is not
performed.
5.1. All reagents and solvents should be of pesticide grade purity or higher to minimize
interference problems. All glassware should be cleaned and demonstrated to be free from
interferences.
5.2. Matrix interferences may be caused by contaminants from the sample matrix, sampling
devices or storage containers. The extent of matrix interferences will vary considerably
from sample source to sample source, depending upon variations in the sample matrix.
Matrix interferences and contaminants are likely to be present and may have an effect on
the recoveries for the analytical procedure. These interferences lead to elevated baselines
and artifacts that may be interpreted as false positives. Wipes were used as received and
analyzed to ensure that no interferences were present. Any materials containing
interferences with the analytes of interest were not used.
5
-------
5.3. Matrix effects are known phenomena of electrospray ionization-mass spectrometry (ESI-
MS) techniques, especially for coeluting compounds. Managing the unpredictable
ionization suppression and enhancement caused by these effects is recognized as an
integral part of the performance and verification of an ESI-MS procedure. The data
presented in this procedure were designed to demonstrate that the procedure is capable of
functioning with realistic samples. Each analyst is encouraged to observe appropriate
precautions and follow the described QC procedures to help minimize the influence of ESI-
MS matrix effects on the data reported. Matrix effects include ion
suppression/enhancement, high background and improper ion ratios.
6. HEALTH AND SAFETY
The toxicity and carcinogenicity of each reagent used in this method have not been precisely defined.
However, each chemical compound was treated as a health hazard. Lewisite 1 and its degradation products
are vesicants. Exposure to these chemicals should be reduced to the lowest possible level and proper
protective equipment should be worn for skin, eyes, etc. Each laboratory is responsible for maintaining an
awareness of Occupational Safety and Health Administration (OSHA) regulations regarding the safe
handling of chemicals used in this method. A reference file of safety data sheets (SDSs) that address the
safe handling of the chemicals should be made available to all personnel involved in the chemical analyses
or subject to potential exposure. Additional references are available (9-12).
Personnel shall wear personal protective equipment, which includes nitrile gloves, laboratory coats, and
safety glasses with side shields or goggles. Nitrile gloves should be changed frequently, between each
operation or after known or suspected contact with hazardous material. All work shall be performed in
chemical fume hoods. Sample manipulations should be performed in secondary containment (e.g.,
phototrays) to allow quick cleanup in the event of a spill. Vial trays should be used to hold vials and
minimize the potential for tipping.
Hydrogen peroxide (H2O2 (30%)) is a strong oxidizer; it is corrosive, explosive, and can cause severe burns.
Contents of a bottle may develop pressure upon prolonged storage. Refrigerated storage is recommended.
Manage wastes from Lewisite extraction and analysis separately from all other laboratory wastes. Pressure
may develop in waste containers and require periodic venting. Do not mix wastes with bleach or any
material that might react with H2O2, if waste excluding Lewisite was used.
7. EQUIPMENT AND SUPPLIES
References to specific brands of equipment and catalog numbers are provided solely as examples and do
not constitute an endorsement of the use of such products or suppliers. Glassware, reagents, supplies,
equipment, and settings other than those listed in this report may be employed, provided that method
performance has been demonstrated and documented for the intended application. Analyses may be
performed with any system capable of performing LC-MS/MS as long as performance criteria described
within are met. Tested matrix types are listed in Table 1. (All Tables are found at the end of the document.)
6
-------
7.1 LC-MS/MS INSTRUMENT
7.1.1 LIQUID CHROMATOGRAPHY (LC) SYSTEM - An analytical system complete
with a programmable temperature and solvent delivery system (Surveyor LC,
ThermoFisher Scientific, Waltham, MA) and all required accessories including
syringes, solvent degasser, and autosampler.
7.1.2 ANALYTICAL COLUMN - Atlantis® T3, 100 A, 3 (.un particle size, 2.1 mm x
150 mm (Waters, Milford, MA, Catalog # 186003719; reverse-phase, C18), or
equivalent.
7.1.3 TANDEM MASS SPECTROMETER (MS/MS) SYSTEM - A Thermo Scientific
LTQ Orbitrap (ThermoFisher Scientific, Waltham, MA) mass spectrometer was
used in the development of this method. The LC-MS/MS should be tuned and
calibrated, as needed, per the vendor's instructions and specifications. A mass
spectrometer capable of MRM analysis with the capability to obtain at least 10
scans over a peak with adequate sensitivity is required.
7.1.4 DATA SYSTEM - The LC-MS/MS should be controlled by software that allows
the continuous acquisition and storage on machine-readable media of all mass
spectra obtained throughout the duration of the chromatographic program.
Xcaliber 2.0 SR2 (or similar software) is used for all quantitative analysis for data
generated from the LC-MS unit.
7.2 EXTRACTION DEVICE
7.2.1 A shaker table was used for the preparation of soil and wipes samples (Glas-Col®
Digital Pulse Mixer, model no. 099ADPM12, Terre Haute, IN; or equivalent). For
water samples, a vortex mixer (VWR Mini Vortexer, product, no. 58816-121, or
equivalent) was used.
7.3 GLASSWARE AND MISCELLANEOUS SUPPLIES
7.3.1 AUTOSAMPLER VIALS - 2-mL, screw-cap, autosampler vials (Agilent, Santa
Clara, CA, product no. 5183-2083), or equivalent.
7.3.2 AUTO PIPETTES - 10.0 mL, 1000 |iL. 100 |iL and 10 |_iL ± 1% accuracy.
7.3.3 DESOLVATION GAS - Nitrogen gas generator or equivalent nitrogen gas supply
(e.g., nitrogen from liquid nitrogen boil-off) aids in the generation of an aerosol of
the ESI liquid spray and should meet or exceed instrument manufacturer's
specifications.
7.3.4 COLLISION GAS - Helium, nitrogen, or argon gas used in the collision cell in
MS/MS instruments and should meet or exceed instrument manufacturer's
specifications.
7.3.5 ANALYTICAL BALANCE - accurate to 0.1 mg; reference weights traceable to
Class S or S-l weights.
7
-------
7.3.6 THERMOMETER - National Institute of Standards and Technology (NIST)-
traceable thermometer.
7.3.7 STANDARD SOLUTION FLASKS - Class A volumetric glassware.
7.3.8 WIPES - 3 in. x 3 in. gauze wipes, 12-ply (Kendall-Curity, product no. 1903), or
equivalent. Use as received.
7.3.9 SAMPLE COLLECTION CONTAINERS - 20- or 40-mL volatile organic
analysis (VOA) vials with Polytetrafluoroethylene (PTFE) screw caps
(Environmental Sampling Supply, product no. 0040-0310-PC, San Leandro, CA),
or equivalent.
7.3.10 SAND - purified (JT Baker, Inc., product no. 3382-05; CAS no. 14808-60-7).
7.3.11 VIRGINIA SOIL - (VA soil, obtained from National Exposure Research
Laboratory, US EPA, Las Vegas, NV), with composition of 64.5% sand, 28% silt,
7.5% clay, 2.6% Total Organic Carbon (TOC), and pH4.1 when measured in a 1:1
soil: water mix.
7.3.12 NEBRASKA AGLANDS AP SOIL - (NE soil, obtained from National Exposure
Research Laboratory, US EPA, Las Vegas, NV), with composition of 5.1% sand,
57.5% silt, 31.7% clay, 1.9% TOC, and pH 5.5 when measured in a 1:1 soil: water
mix.
7.3.13 GEORGIA BT2 SOIL - (GA soil, obtained from National Exposure Research
Laboratory, US EPA, Las Vegas, NV), with composition of 46% sand, 22% silt,
32% clay, 0.2% TOC, and pH 5.0 when measured in 1:1 soil: water mix.
7.3.14 GLASS BEADS - glass beads ~ 5 mm diameter (Sigma-Aldrich, product no.
18406) or equivalent. Use as received.
7.3.15 pH INDICATING PAPER - pH range 0-14 (Sigma-Aldrich product no.
WHA103 62000) or equivalent.
7.3.16 CENTRIFUGE - (Fisher Scientific AccuSpin™ 400 with rotor, product no.
75005195, Fischer Scientific, Pittsburgh, PA) or equivalent.
8 REAGENTS AND STANDARDS
8.1 REAGENTS AND STANDARDS
Laboratories should follow standard QC procedures to determine when the standards
should be replaced. Label all standards and verify the correct grade of solvents. Reagent-
grade chemicals should be used, unless otherwise indicated. Traceability of standards is
established by the manufacturer's specifications provided at time of purchase.
8.1.1 SOLVENTS, REAGENTS and GASES - Water (resistivity of 18.2 MQ cm at 25
°C, Millipore Milli-Q®, Advantage® A-10 purification system, Millipore, Billerica,
8
-------
MA) or equivalent, demonstrated to be free of analytes and interferences.
Hydrogen peroxide (H2O2), 30 wt% in H2O, Reagent Grade (Sigma-Aldrich,
product no. 216763-lOOmL). Hydrochloric acid (HC1), 10 mmol, diluted from 5N
HC1 (JT Baker, Baker Analyzed® Reagent, product no. 5618-02) by adding 1.5
mL of 5N HC1 to 750 mL laboratory-grade water. Dichloromethane (Fluka brand
from Sigma-Aldrich, product no. 34488). Methanol (MeOH), Fluka brand, LC-MS
Chromasolv®, >99.9% (Sigma-Aldrich product number 34966-4L). Acetonitrile
(ACN) (Optima™ LC-MS grade, Fisher Scientific, product number A955-212).
Formic acid (Optima™ LC-MS grade, Fisher Scientific, product number A117-
50). Nitrogen is used for the generation of aerosol of the ESI liquid spray, and
purity should meet instrument manufacturer's specifications. Helium is used as the
collision gas in Orbitrap MS/MS applications, and purity should meet instrument
manufacturer's specifications. If nitrogen or argon is to be used as a collision gas
(e.g., in quadrupole-based tandem mass spectrometry systems), its purity should
meet instrument manufacturer's specifications.
8.1.2 MOBILE PHASE A - Solution A (Table 2) consisted of LC-MS grade water with
0.1% formic acid. To prepare 0.5 L, add 0.5 mL of formic acid and dilute to 0.5 L
mark with water. This solvent system is prone to some microbial growth and
should be replaced regularly.
8.1.3 MOBILE PHASE B - Solution B (Table 2) was comprised of acetonitrile with
0.1% formic acid. To prepare 0.5 L, add 0.5 mL of formic acid and dilute to 0.5 L
mark with acetonitrile.
8.1.4 TARGET ANALYTES - Lewisite 1 (LI) was synthesized in-house. The Lewisite
purity, generated from the synthesis procedure, was 91% purity (determined by
NMR and GC/MS analysis). A Lewisite 1 stock solution (2.03 mg/mL) was first
diluted in dichloromethane, fresh from a new bottle, and stored at -20 °C. CVAA
was made by diluting 246 (.iL of 2.03 mg/mL LI stock solution in dichloromethane
to 15.00 mL laboratory-grade water for a 33.3 ng/(.iL stock solution. LC-MS
analysis confirmed that LI was quantitatively converted to CVAA, and LI was not
present.
8.1.5 SURROGATE ANALYTES - Phenyl arsonous acid (PAA) was made from
phenylarsine oxide by hydrolysis. Phenylarsine oxide was obtained from Sigma-
Aldrich (product no. 402494, CAS no. 637-03-6; note: current Sigma-Aldrich
product no. P3075-1G). Direct dissolution of phenylarsine oxide in water is
difficult and requires preparation of an initial concentrated solution in acetonitrile.
8.2 STANDARD SOLUTIONS
Stock standards and all subsequent solutions should be replaced when analyzed solution
concentrations deviate more than ± 20% from the prepared concentration. Standards are
stored protected from light (amber flasks) and at - 20 °C (± 2 °C). Laboratories should
utilize QC practices to determine when standards should be replaced.
9
-------
8.2.1
SURROGATE STOCK STANDARD SOLUTION (Surrogate SSS)
While phenylarsine oxide was easier to dissolve in dimethylformamide, this
solvent was found to co-elute with the analytes of interest, causing ionization
suppression effects during analysis by LC-MS/MS. For this reason, initial
phenylarsine oxide solutions were made in acetonitrile. A stock solution of 1.00
mg/mL PAA was made in acetonitrile. It was necessary to sonicate this solution
for several hours and to let the solution stand in the hood overnight to obtain
dissolution of the PAA. Once dissolved, phenylarsine oxide was kept at ~ 4 °C and
could be used for at least two months (i.e., no degradation/oxidation was observed
over a two-month period). Subsequent dilutions of phenylarsine oxide were made
in laboratory grade water (which generated the surrogate compound PAA by
hydrolysis) immediately prior to conducting experiments.
8.2.2 ANALYTE STOCK STANDARD SOLUTION (AS)
A 33.3 ng/(.iL CVAA stock solution was made by diluting 246 |_iL of 2.03 mg/mL
Lewisite stock solution in dichloromethane to 15.00 mL using laboratory-grade
water. A 10 ng/(.iL CVAA stock solution was made by diluting 1.00 mL of 33.3
ng/(.iL CVAA to 3.33 mL with laboratory-grade water.
8.2.3 CALIBRATION STANDARD SOLUTION (CAL)
8.2.3.1 Initial Calibration Standards
The calibration range of the LC-MS/MS is defined by analyzing standard
solutions representing a range of concentrations of the analytes of interest.
Six calibration levels were used to generate a linear calibration curve for
CVAOA and the correlation coefficient is used to evaluate the quality of the
calibration curve (e.g., r2>0.99 indicates a good fit of the calibration curve
with the data points generated by the analysis of the standards). All
calibration standards are prepared in a solvent corresponding to the solvent
system used for extraction of the relevant matrix (see footnotes of Table 3).
The presence of H2O2 in the solvent is expected to oxidize all analytes (e.g.,
only CVAOA and PAOA are present, and no CVAA or PAA remains). To
confirm that all analytes were oxidized, CVAA and PAA were monitored
but were not detected at tested levels during this investigation.
Recommended initial calibration levels are listed in Table 3.
8.2.3.2 Continuing Calibration Verification Standards (CCV)
The continuing calibration verification (CCV) is used to check the continued
validity of the initial calibration. The CCV is a mid-range calibration
standard and the acceptance criteria is ±35% of the expected value(s) for all
analytes. If the CCV does not meet the acceptance criteria, it may be
reanalyzed. If after reanalysis the ±35% criteria for the CCV are not met, a
new calibration curve must be prepared and used. The CCVs consist of clean
solvent that is fortified with a specific concentration of CVAOA and PAOA.
A recommended CCV consists of 0.1 ng/(.iL CVAOA and 0.1 ng/(.iL PAOA.
A CCV check should be performed at a minimum frequency of once every
8 hours, preferably after every 10 field samples.
10
-------
8.2.3.3
Surrogate Standard
PAA is used as a surrogate standard to provide assurance that the
extraction and analysis procedure is in control. PAA is oxidized to PAOA;
exhibiting chemical behavior similar to and mass spectral features of
CVAA/CVAOA. It is recommended to adjust the concentration of PAA
such that its concentration in the sample extract to be analyzed is 0.1 ng/(.iL
(detected as PAOA), assuming 100% recovery. This concentration is
comparable to the concentration of the PAOA in recommended CCVs.
Alternatively, one could lower the concentration of PAA to demonstrate
that a specified low limit of detection is achieved.
9 SAMPLE COLLECTION, PRESERVATION AND STORAGE
The following sections describe procedures for extracting Lewisite 1 degradation products from three
matrices. Per the sample holding time study, it is recommended that water and wipe samples be extracted
and analyzed within two weeks and that soil samples should be extracted and analyzed as soon as possible
after collection (see Section 9.2).
9.1 SAMPLE COLLECTION
9.1.1 VOA vials with PTFE screw caps were used for sample collection for each matrix
type. Other containers may be used to collect samples as long as they are tested
and verified to ensure they do not contain any interfering compounds.
9.1.2 WATER EXTRACTION
9.1.2.1.1 From the original (VOA) sample container, sample, with a pipette, a drop
of water and test its pH with indicating paper. Record the pH. While this
method does not require the adjustment of pH to a specific value, pH is
recorded as good laboratory practice and to identify samples that may be
outside the normal range of pH values associated with natural waters. As
this method is applied to more matrices, sample pH might be found to be
an important parameter.
9.1.2.2 Measure 500 |o,L water sample into 2 mL autosampler vial. Spike each
sample with 10 |o,L of 10 ng/|o,L PAA solution (yielding a 0.1 ng/ |o,L
PAOA in sample extract after dilution). Mix thoroughly. Add 500 |o,L of
30% H2O2 to each sample. Cap and mix solution. Analyze as soon as
possible by LC-MS/MS using the conditions listed in Tables 2, 3, and 4.
9.1.2.3 Preparation of blank sample(s): Place laboratory-grade water into a VOA
vial used to collect the water samples. Treat as a water sample, as
described in sections 9.1.2.1 and 9.1.2.2.
11
-------
9.1.2.4 Preparation of Matrix Spike/Matrix Spike Duplicate sample(s): Measure
19.6 mL laboratory-grade water into a VOA vial. Spike with 120 |_iL of
33.3 |J.g/mL CVAA solution (which will yield a final extract concentration
of 0.1 ng/(.iL CVAOA). Use the vortex mixer to ensure that the CVAA
solution is mixed thoroughly into the water. Treat as a water sample, as
described in sections 9.1.2.1 and 9.1.2.2.
9.1.3 SOIL EXTRACTION
9.1.3.1 The sample collection (VOA) container contains 5.00 g of soil/sand. Add
200 |_iL of a solution containing 10 ng/(.iL PAA to each sample (which,
assuming 100% recovery, will yield an extract concentration of 0.1 ng/(.iL
PA OA). Mix with vortex mixer for approximately 30 seconds to distribute
the PAA. Add 5-10 glass beads, new from the bottle, to each sample. Add
10.0 mL 50/50 (v/v) 10 mM HC1 /methanol to each sample and cap tightly.
Place on shaker table for 30 minutes at 600 revolutions per minute (rpm)
(horizontal orientation). Using pH paper, record the approximate pH of the
sample extract. Allow the sample extracts to settle by gravity so that 0.5
mL of extract may be collected; it may take ~1 hour for the extracts to
settle. If needed, centrifuge the sample extracts for 5 minutes at 500 rpm.
(Note: Do not exceed 500 rpm or the vial might break). Transfer 500 |aL
of sample extract into an autosampler vial. Add 500 |_iL 30% H2O2 to each
sample. Cap and mix solution. Analyze as soon as possible by LC-MS/MS
using the conditions listed in Tables 2, 3, and 4.
9.1.3.2 Preparation of the Blank Sample(s): Weigh 5.00 g of clean sand/soil into
a 40-mL VOA vial and cap. Treat as a soil sample described in section
9.1.3.1.
9.1.3.3 Preparation of the Matrix Spike/Matrix Spike Duplicate Sample(s): Weigh
5.00 g of clean sand/soil into a 40-mL VOA vial. Spike with CVAA and
mix with a vortex mixer for -30 seconds to distribute the solution.
Recommended spike amount is 2 (.ig CVAA (which, assuming 100%
recovery, will yield an extract concentration of 0.1 ng/|_iL CVAOA). Treat
as a soil sample described in section 9.1.3.1.
9.1.4 WIPE EXTRACTION
9.1.4.1 The sample collection (VOA) container contains the wipe samples. Add
400 (.iL of a solution containing 10 ng/|_iL PAA (surrogate) to each sample
(assuming 100% recovery, final concentration will be 0.1 ng/(.iL PAOA in
the sample extract). Add 20 mL of 10 mM HC1 to each vial. Extract for 30
minutes on shaker table (horizontal orientation) at 600 rpm. Transfer 500
(.iL sample extract to autosampler vial. Add 500 (.iL 30% H2O2 to each
12
-------
sample. Cap and mix. Analyze as soon as possible by LC-MS/MS using
the conditions listed in Tables 2, 3, and 4.
9.1.4.2 Preparation of the Blank Sample(s): Place unused wipe into a 40-mL vial
with PTFE screw cap. Treat as a wipe sample, as described in section
9.1.4.1.
9.1.4.3 Preparation of the Matrix Spike/Matrix Spike Duplicate Sample(s): Place
unused wipe into a 40-mL VOA vial. Spike with 4 (.ig CVAA (which,
assuming 100% recovery, will yield an extract concentration of 0.1 ng/(.iL
CVAOA). Treat as a wipe sample, as described in in section 9.1.4.1.
9 2 SAMPLE STORAGE AND HOLDING TIMES
9.1.5 Samples should be extracted as soon as possible after collection but water and wipe
extract samples must be extracted within 14 days of collection. Soil samples should
be extracted immediately. Samples not immediately analyzed from a particular site
should be carefully characterized to ensure there is no interaction to cause
interferences or degradation of the analytes. Detailed information is described in
Section 14.3.
QUALITY CONTROL
10.1 Quality control (QC) requirements include the performance of an initial demonstration of
capability (IDC) and ongoing QC requirements that must be met to generate data of
acceptable quality when preparing and analyzing samples. This section describes the QC
parameters, their required frequencies and performance criteria. A Method Detection Limit
(MDL) study (Tables 5 and 6, and Section 10.3) were performed to demonstrate laboratory
capability. Laboratories are encouraged to institute additional QC practices to meet their
specific needs.
10.2 INITIAL DEMONSTRATION OF CAPABILITY
The IDC must be performed successfully prior to the initiation of analysis of field samples.
Prior to conducting an IDC, an acceptable initial calibration must be generated as outlined
in Section 11.1.
10.2.1 INITIAL DEMONSTRATION OF LOW SYSTEM BACKGROUND
Any time new solvents, reagents, filters and autosampler vials are used, the LMB
must be demonstrated to be reasonably free of contamination (i.e., that the criteria
are met as stipulated in Section 10.4.1). The LMB is used to ensure that analytes
of interest or other interferences are not present in the laboratory environment, the
solvent, or the apparatus.
NOTE: Good laboratory practices indicate the use of a blank before and after analyzing
a calibration curve for an instrument to ensure that no carryover will occur. If the
required criteria are not met and samples were not free of contamination, then the source
of the contamination should be identified and eliminated before the performance of
any analysis.
13
-------
10.2.2 MINIMUM REPORTING LEVEL (MRL)
Establish a target concentration for the minimum reporting limit (MRL) based on
the intended use of the method. Establish an Initial Calibration (Section 11.1). The
lowest calibration (CAL) standard used to establish the initial calibration must be
at or below the MRL concentration. If the MRL concentration is too low, ongoing
QC requirements may fail repeatedly, and the MRL must be determined again
at a higher concentration. The MRL reported in this study is the lowest
calibration level.
10.2.3 CONTINUING CALIBRATION VERIFICATION
The CCV is used to check the continued validity of the initial calibration. The CCV
is a mid-range calibration standard and the acceptance criterion is ± 35% of the
expected value(s) for all analytes. If the CCV does not meet the acceptance criteria,
it may be reanalyzed. If after reanalysis the ± 35% criteria for the CCV are not
met, a new calibration curve must be made and used. The CCVs consist of clean
solvent that is fortified with a specific concentration of CVAOA and PAOA. A
recommended CCV consists of 0.1 ng/(.iL CVAOA and 0.1 ng/(.iL PAOA. A CCV
check should be done at a minimum frequency of once every 8 hours; preferably
after every 10 field samples.
10.3 METHOD DETECTION LIMITS
The procedure for the determination of the laboratory detection and quantitation limits for
the EPA approach follows 40 CFR Part 136, Appendix B. MDLs represent the minimum
concentration at which there is a high degree of statistical confidence that, when the method
reports that an analyte is present, that analyte is actually present (i.e., a low risk of false
positives).
10.3.1 DETERMINATION OF LABORATORY INSTRUMENT DETECTION
LIMITS
The laboratory instrument detection limit (IDL) can be used to establish an
estimate of the initial spiking concentration used for determination of the MDL,
although other approaches for determining the initial spiking concentration may
be used. The laboratory IDL is determined for each analyte as a concentration that
produced an average signal-to-noise (S/N) ratio in the range of 3:1 - 5:1 for at least
three replicate injections. For example, successively lower concentrations of the
analytes are injected until the S/N ratio is in the range of 3:1 - 5:1. Replicates are
then injected at that target concentration to ensure that the average S/N of the
replicates was within the 3:1 - 5:1 range. Note that since linearity of S/N ratio with
increasing or decreasing concentration cannot be assumed, the concentrations
determined via this procedure are necessarily approximate.
10.3.2 DETERMINATION OF LABORATORY METHOD DETECTION LIMIT
MDLs represent the optimal detection achieved by a laboratory in a matrix of
interest. The analyte spiking solution, containing all target analytes, was added to
14
-------
the matrix. The 40 CFR Part 136, Appendix B procedure is followed, particularly
with regard to spike levels used. Replicate reference matrix samples are spiked at
a level between 1-5 times the estimated detection level (e.g., suggested by the IDL
procedure in 10.3.1). The resulting MDL must be within 10 times the spike level
used, or the MDL determination would be repeated using a more appropriate spike
level. Full method sample preparation procedures to prepare and analyze at least
seven replicates of the spiked clean matrix of interest are used. Apply the following
equation to the analytical results (Student's Z-factor is dependent on the number of
replicates used; the value 3.14 assumes seven replicates):
MDL = t (n-l, l-a = 0.99) X SD
where
MDL = method detection limit,
t = Student's t value for the 99% confidence level with n-l degrees of
(n-l,l-a = 0.99)
freedom (for seven replicate determinations, the Student's t value is 3.143
at a 99% confidence level),
n = number of replicates, and
SD = standard deviation (SD) of replicate analyses.
Data for MDLs are shown in Table 5.
10.4 ONGOING QC REQUIREMENTS
10.4.1 LABORATORY METHOD BLANK
Method blanks are used to determine the background of each particular matrix. An
LMB is prepared and analyzed with each extraction batch for confirmation that
there are no background contaminants interfering with the identification or
quantitation of the target analytes. If there is a contaminant within the retention
time window preventing the determination of the target analyte, the source of the
contamination should be determined and eliminated before processing samples.
The method blanks undergo the same extraction procedure as authentic samples
and are spiked with the surrogate standard; however, the method blanks do not
contain Lewisite 1/CVAA/CVAOA. One method blank is prepared for each set of
samples. The maximum number of samples in a set is 20.
10.4.2 CONTINUING CALIBRATION VERIFICATION CHECK
A CCV check should be performed at a minimum frequency of once every 8
hours; preferably after every 10 field samples. CCVs should be within ± 35% of
the expected value(s) for all analytes for the data to be considered valid. CCV
values should be specified by the sample submitter's Data Quality Objectives
(DQOs) or fulfill other QC requirements.
10.4.3 MATRIX SPIKE/LABORATORY FORTIFIED MATRIX SPIKE
A laboratory fortified matrix spike (LFMS) is analyzed to determine that spike
15
-------
accuracy for a particular sample matrix is not adversely affected by chemical
interactions between target analytes and experimental matrices (i.e., wipe
materials, sand, etc.). If a variety of sample matrices are analyzed, performance
should be established for each matrix or sample type.
10.4.4 When performing sample analyses, it is expected that LFMS and laboratory
fortified matrix spike duplicate (LFMSD) samples will be analyzed.
LFMS/LFMSDs are representative analyte-free environmental matrices that have
been fortified with CVAA and PAA. These samples are taken through the
extraction process to show that the method is capable of detecting the analytes of
interest in the relevant matrices. LFMS and LFMSD samples should be prepared
for each type of matrix. Records are maintained of the target compound spike
analyses, and the average percent recovery (X) and the standard deviation (SD)
are calculated. Analyte recoveries may exhibit bias for certain matrices.
Acceptable recoveries are 50-150% if a low-level concentration near or at the
MRL (within a factor of 3) is used. If the recovery does not fall within this range,
check with a CCV or prepare a fresh AS solution for analysis. If the recovery of
any analyte still falls outside the designated range and the laboratory
performance for that analyte is shown to be in control in the CCVs, the recovery
is judged to be matrix-biased. 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.
10.4.5 SURROGATE STANDARD
All samples (CCVs, LMBs, LFMSs, LFMSDs, and CAL standards) are spiked
with surrogate standard spiking solution as described in Section 8.2. An average
percent recovery of the surrogate compound and the standard deviation of the
percent recovery (REC) are calculated and updated regularly.
10.4.6 MATRIX SPIKE DUPLICATE
10.4.6.1 Calculate the relative percent difference (RPD) for the LFMS and LFMSD
using the equation:
| LFMS - LFMSD |
RPD = x 100
(LFMS - LFMSD)/2
RPDs for duplicate LFMSs should be < 35% for each analyte. Greater
variability may be observed when the matrix is fortified at analyte
concentrations at or near the MRL (within a factor of two times the MRL
concentration). LFMSs at these concentrations must have RPDs that are <
50%. If the RPD of an analyte falls outside the designated range and the
laboratory performance for the analyte is shown to be in control in the
CCV and in the LMB, the precision is judged matrix-influenced. Report
the result for the corresponding analyte in the unfortified sample as
"suspect/matrix."
16
-------
11 INSTRUMENT CALIBRATION AND STANDARDIZATION
All laboratory equipment should be calibrated according to manufacturer's protocols. Demonstration and
documentation of acceptable mass spectrometer (MS) tuning and initial calibration is necessary prior to
sample analysis. Verification of the tuning of the MS must be repeated each time instrument
modification/maintenance is performed and prior to analyte calibration. After initial calibration is
successful, a CCV (at the appropriate concentration described in section 10.4.2) should be performed at the
beginning and end of each analysis batch.
11.1 INITIAL CALIBRATION FOR ANALYTES
11.1.1 ESI positive mode was used for this method. Optimize the [M+H]+ ion in ESI
positive mode for each analyte by infusing an appropriate calibration solution at a
flow rate similar to the flow rate used for the LC separation. Adjust MS parameters
(voltages, temperatures, gas flows, etc.) until optimal analyte responses are
achieved. Optimize the product ion by adjusting the collision energy. Ensure that
there are at least 10 scans across the peak for optimal precision. ESI-MS and
MS/MS parameters utilized during development of this method are presented in
Table 4.
11.1.2 Establish LC operating conditions that will optimize peak resolution and shape.
Suggested LC conditions (listed in Table 2) may not be optimal for all LC systems.
11.1.3 The initial calibration contains a six-point curve using the analyte concentrations
prepared in section 8.2.3 and shown in Table 3. The lowest calibration curve
standard is at the MRL. The calibration curve and all samples should be analyzed
in a low to high concentration regimen so carryover is less of a concern in case the
LC cleaning cycle does not clean the system adequately between injections. Verify
that all analytes have been properly identified and quantified using software
programs. Integrate manually, if necessary, in accordance with laboratory quality
assurance plans. Depending on the instrument, sensitivity and calibration curve
responses may vary. If the polynomial type excludes the point of origin, use a fit
weighting of 1/X to give more weighting to the lower concentrations. The
coefficient of determination (r2) of the linear fit should be greater than or equal to
0.98. If one of the calibration standards other than the high or low standard causes
the r2 to be <0.98, this point must be re-injected or a new calibration curve must
be analyzed. The r2 of the quadratic curve should be greater than or equal to 0.99.
If one of the calibration standards other than the high or low standards causes the
r2 to be <0.99, follow the same procedure given above for a linear fit. A calibration
curve and an instrument blank will be analyzed at the beginning of each batch or
daily to ensure instrument stability. When quantitated, each calibration point for
each analyte should calculate to be within 70-130% of its true value. The lowest
CAL standard should calculate to be within 50-150% of its true value. A new curve
will be generated daily. The calibration method is used to quantify all samples.
Note that, because of solvent differences, different calibration curves must be made
for water samples and soil/wipe samples; see Figure 2. (Remaining Figures are at
the end of the SOP.)
17
-------
11.2 QUANTITATION OF ANALYTES
The quantitation of the target analytes is accomplished with quantitation software as it relates
to each specific instrument. For data collected with the Orbitrap, Xcaliber 2.0 SR2 software
was used for quantification. Peak areas associated with the MRM transitions for each analyte
were compared to those of calibration standards. An external calibration (linear) is used along
with monitoring PAA surrogate recoveries. A calibration range of 0.02-0.2 (ig/mL is
suggested. Refer to Table 4 for the MRM transitions and retention times utilized during the
development of this method.
18
-------
ANALYTICAL PROCEDURE
12.1 SAMPLE PREPARATION
12.1.1 Samples were collected and stored as described in Section 9.2. The surrogate
(PAA) is added first, then the appropriate solvent is added to the VOA vial.
12.1.2 After extraction, transfer resulting sample extract (via pipette) to a standard, 2-mL
autosampler vial.
NOTE: Calibration standards are not filtered. If alternate filtering is incorporated, the
filters should be subjected to QC requirements to ensure they do not introduce interferences
or retain the target analytes.
12.2 SAMPLE ANALYSIS/ANALYTICAL SEQUENCE
12.2.1 Use the same Liquid Chromatography/Mass Spectrometry conditions established
per guidance described in Section 11 and summarized in Tables 2, 3, and 4.
12.2.2 Prepare an analytical batch that includes all QC samples and field samples. The
first sample to be analyzed is a 10 (iL injection of a blank solvent on column
followed by the calibration curve.
12.2.3 Update the calibration file and print a calibration report. Review the report for
calibration outliers and make area corrections by manual integration, if necessary
and appropriate. If corrections have been made, update the calibration file, noting
the changes, and regenerate a calibration report. Alternatively, re-analyze
"nonconforming" calibration level(s) and repeat the above procedures.
12.2.4 The first sample analyzed after the calibration curve is an additional blank to
ensure there is no carryover. If the initial calibration data are acceptable, begin
analyzing samples, including QC and blank samples, at their appropriate frequency
injecting the same size aliquots (10 |o,L) under the same conditions used to analyze
CAL standards. The ending CCV must have each analyte concentration within
3 5 % of the calculated true concentration or the affected analytes from that run must
be qualified as estimates or the samples must be re-analyzed with passing criteria
to remove the qualification.
12.2.5 If the absolute amount of a target compound exceeds the working range of the LC-
MS system (see Level 6 in CAL standards), the prepared sample is diluted with
the appropriate solvent and re-analyzed along with additional samples that may
have run after the sample known to exceed the calibration range, because of the
possibility of carryover. Care must be taken to ensure that there is no carryover of
the analyte that has exceeded the calibration range. If the amount of analyte
exceeds the calibration range, a blank sample should be analyzed afterward to
demonstrate no carryover will occur.
12.2.6 At the conclusion of the data acquisition, use the same software that is used in the
calibration procedure to identify peaks of interest from the predetermined retention
time windows. Use the data software to examine the ion abundances of the peaks
19
-------
in the chromatogram to identify and compare retention times in the sample
chromatogram with the retention time of the corresponding analyte peak in an
analyte standard.
13 DATA ANALYSIS AND CALCULATIONS
13.1 QUALITATIVE AND QUANTITATIVE ANALYSIS
13.1.1 Complete chromatographic resolution is not needed for accurate and precise
measurements of analyte concentrations when using MS/MS. An external
calibration is used when monitoring the MRM transitions of each analyte.
Quantitation software is utilized to conduct the quantitation of the target analytes
and surrogate standard. The MRM transitions of each analyte are used for
quantitation and confirmation. The MRM transition serves as a confirmation by
isolating the precursor ion, fragmenting the precursor ion to the product ion, and
relating the transition to the retention time in the calibration standard. Under these
conditions, the elutiontimes of CVAOA and PAOA are approximately 3.7 minutes
and 7.0 minutes, respectively. The elution times for CVAA and PAA (the absence
of which provides assurances that complete oxidation has occurred) are
approximately 6.7 minutes and 12.5 minutes, respectively.
13.1.2 Computer programs used for analysis of data include instrumentation and
quantitation software. Manual integration may be necessary for some peak areas if
the peak area is not integrated properly (i.e., the integration for the peak is not fully
performed by the instrument's software, which will be noticeable by visual
inspection of each peak). Inspect all integrated peaks for visible integration errors
and manually integrate as necessary to ensure consistent integration of other peaks
and/or known calibration peaks. Any manual integration should be carried out by
a qualified analyst, noted, and checked against quality control procedures (sections
10 and 11.3).
13.2 Prior to reporting data, the chromatogram should be reviewed for any incorrect peak
identifications. The retention time window of the MRM transitions must be within 5% of
the retention time of the analyte standard. If this is not true, the calibration curve needs to
be re-analyzed to see if there was a shift in retention times during the analysis and the
sample needs to be re-injected. If the retention time is still incorrect in the sample, the
analyte is referred to as an unknown. If peaks need to be manually adjusted due to incorrect
integration by the program, clarification of where professional judgment was used to alter
the peaks should be documented during the data reduction and verification process.
14 METHOD PERFORMANCE
14.1 DETECTION LIMITS
14.1.1 Detection limit results for a single laboratory study are presented in Tables 5 and
6. MDLs were compared to analytical target levels (ATLs) to establish the
method's ability to detect target analytes at health-based and environmentally
relevant levels. ATLs are based on chemical agent health-based standards and
guidelines established by U.S. Army Public Health Command (13). U.S. Army
20
-------
levels were used because civilian health-based guidelines have not been
established and are not known to exist. The surface (wipe) ATL was estimated
because it was not provided within the U.S. Army document and is based on total
arsenic concentration, which may not be an accurate representation for Lewisite
(14).
14.2 RECOVERIES AND PRECISION FOR MATRIX TYPES
14.2.1 Section 18 lists recoveries and precision of target analytes for all tested matrices.
14.3 STORAGE STABILITY STUDY
14.3.1 A sample holding time study was initiated to determine the stability of Lewisite
(deposited as CVAA) on matrices, including laboratory water, wipes, sand, and
Nebraska soil, over a two week period. The samples were spiked on Day 0 and
extracted and analyzed immediately or stored at ~ 4 ± 2 °C until being extracted
and analyzed on Days 2, 7, or 14. Samples were prepared, using materials and
matrices previously described in this procedure. All analyses were performed using
the procedures of Sections 9 and 12.
Water samples were prepared by placing 400 mL of laboratory water in a 1-L (one
liter), pre-cleaned, amber bottle, and adding 80 |_iL of 2.03 mg/mL Lewisite in
dichloromethane. This solution was shaken vigorously for several minutes and
allowed to equilibrate for approximately two hours before placing aliquots of this
solution into 20-mL VOA vials (with no headspace) for immediate analysis or
storage at 4 ± 2 °C. Three replicate samples in VOA vials were prepared for
analysis at each time point. A total of 14 samples, including two extra samples,
were prepared for the holding time study. (NOTE: The extra samples were
collected as a precautionary step in the event additional samples were needed. For
this investigation, the additional samples were not needed and were not analyzed)
In addition, four blank samples (laboratory water placed in a 20-mL VOA vial)
were prepared on Day 0. These were stored under the same conditions as the spiked
samples and a single blank was analyzed on each day that data were collected (i.e.
one blank sample was analyzed on Days 0, 2, 7, and 14) following sample
preparation procedures (section 9.1.2).
Wipe samples were prepared by placing a single wipe in 40-mL VOA vial. Each
wipe was spiked with 120 (.iL of 33.3 (ig/mL CVAA solution. A total of 14 wipe
samples were prepared; three replicate samples were analyzed immediately (Day
0) and the remaining samples were stored at 4 ± 2 °C until their extraction and
analysis on Days 2, 7, and 14. In addition, four blank samples (clean wipe placed
in a 40-mL VOA vial) were prepared on Day 0. These blank samples were stored
under the same conditions as the spiked samples, and a single blank was analyzed
on each day that data were collected (i.e., one blank sample was analyzed on Days
0, 2, 7, and 14) following sample preparation procedures (section 9.1.4).
Sand/soil samples were prepared by placing 5.00 g of sand/soil in 40-mL VOA
vial. Each sand/soil was spiked with 60 (iL of 33.3 (ig/mL CVAA solution and
mixed, using a vortex mixer, for approximately 30 seconds. A total of 14 samples
were prepared; three replicate samples were analyzed immediately (Day 0) and the
remaining samples were stored at 4 ± 2 °C until their extraction and analysis on
21
-------
Days 2, 7, and 14. In addition, four blank samples (clean sand/soil placed in a 40-
mL VOA vial) were prepared on Day 0. These blank samples were stored under
the same conditions as the spiked samples, and a single blank of sand/soil was
analyzed on each day that data were collected (i.e., one blank sample was analyzed
on Days 0, 2, 7, and 14) following sample preparation procedures (section 9.1.3).
CVAOA concentrations measured for the holding time study are found in Table 7;
surrogate information is found in Table 8. Statistical analysis of sample holding
time data for CVAOA is presented in Table 9. A statistical software program called
"R" (15) was used to determine if differences in CVAOA concentrations as a
function of time could be discerned using statistical analyses. Statistical analyses
were performed using the data collected by LC-MS/MS, as described in this
standard operating procedure.
Dunnett's Test was performed separately for each experimental condition, to
compare the CVAOA concentrations measured at each time point (i.e., t > 0) with
the initial measured CVAOA concentration (t = 0). The null hypothesis was that
the average CVAOA concentrations at the later times are greater than or equal to
the initial CVAOA concentration; the alternative hypothesis was that one or more
average CVAOA concentration at a later time is less than the initial CVAOA
concentration (a one-sided test). At a significance level (a) of 0.01 (a conservative
value of a = 0.01 was chosen over the commonly used value of a = 0.05 to
compensate for the increased rate of statistical false positives resulting from
multiple applications of Dunnett's Test), no statistical differences in CVAOA
concentrations were observed in water or wipe samples. However, sand and
Nebraska soil samples showed statistically significant losses after two days of
storage. It is unclear as to whether these decreases are caused by loss of analyte or
irreversible binding of the analyte to the soil. Further investigation is needed.
A limited extract holding time study was performed by re-analyzing Day 7 sample
extracts nine days after the initial Day 7 analysis (i.e., Day 7 was considered from
the initial time for the extract holding study and nine days after the initial Day 7
analysis is considered Day 16). Note that for these experiments, the wipes, sand,
and soil were extracted in their storage vials and the resulting sample extracts were
not transferred to a new vials prior to being replaced in refrigerated storage for 9
days (i.e., extracts remained in contact with the sample matrices to minimize
sample handling). On Day 16 of the study, a 0.5-mL aliquot was removed from
each Day 7 analysis vial (still containing the original sample matrix), placed in an
autosampler vial to which 0.5 mL of 30% H2O2 was added, and analyzed by LC-
MS/MS. The results of this analysis are presented in Table 10. The limited data set
suggests that wipe extracts and Nebraska soil extracts are relatively stable; sand
extracts are not. However, these should be considered preliminary data.
Figures 3 and 4 represent the CVAOA and PAOA responses obtained during
calibration; standards were made according to specification in Section 8.2.3, Table
3. Each data point represents one measurement, and all data were collected on the
same day. Data suggest that calibration curves are best made in solvents matched
to those used for extraction of the relevant sample matrix; however, a single
calibration curve could be used for wipes and soils (see also data in Figure 3). For
the data shown in Figures 3 and 4, calibration standards for water were diluted in
30% H2O2; calibration standards for wipes were diluted in 10 mM HC1 and 30%
22
-------
H2O2; and calibration standards for soils were diluted in 10 mM HC1, 25% MeOH,
and 30% H2O2. Calibration curves were evaluated after each data collection. Data
suggest that one calibration curve is sufficient for both wipes and soils, if needed;
however, calibration curves should be made as often as necessary for each matrix.
14.4 PROBLEM ANALYTES AND SURFACES
14.4.1 TARGET ANALYTES IN SOIL
Sand and Nebraska soil present challenging matrices for extracting the target
analytes. As a result, presence/absence of the target analyte may be best for these
particular matrices.
15 POLLUTION PREVENTION
15.1 This method utilizes small volumes of organic solvent and small quantities of pure
analytes, thereby minimizing the potential hazards to both analyst and environment.
15.2 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 or on-line at
http://www.acs.org/content/dam/acsorg/about/governance/committees/chemicalsafety/pu
blications/less-is-better.pdf (accessed August 15, 2013).
16 WASTE MANAGEMENT
The analytical procedures described within generate relatively small amounts of waste since only small
amounts of reagents and solvents are used. Laboratory waste management practices must be conducted
consistent with all applicable rules and regulations, and laboratories should protect the air, water, and land
by minimizing and controlling all releases from fume hoods and bench operations. Also, compliance with
any sewage discharge permits and regulations is required, particularly the hazardous waste identification
rules and land disposal restrictions.
16.1 Each laboratory should determine with federal and local officials how to safely dispose of
field and QC samples. Waste containers should be properly labeled to identify the contents.
Remember to attach the appropriate chemical waste label, date the beginning of collection
before using the container and follow all appropriate federal and local waste disposal
requirements.
23
-------
17 REFERENCES
1. U.S. Environmental Protection Agency (EPA), 2012. Selected Analytical Methods for
Environmental Remediation and Recovery (SAM). EPA/600/R-12/555 July 2012. Cincinnati,
Ohio: United States Environmental Protection Agency, Office of Research and Development,
National Homeland Security Research Center.
2. Code of Federal Regulations, 40 CFR Part 136, Appendix B. Definition and Procedure for the
Determination of the Method Detection Limit - Revision 1.11.
3. Federal Advisory Committee on Detection and Quantitation Approaches and Uses in Clean Water
Act Programs. Final Report. (Submitted to U.S. Environmental Protection Agency.) December 28,
2007.
4. Glaser, J.A., D.L. Foerst, G.D. McKee, S.A. Quave, W.L. Budde, "Trace Analyses for
Wastewaters." Environ. Sci. Technol. 1981, 15, 1426-1435.
5. Rodin, I., Braun, A., Stavrianidi, A., Shpigun, O. and Rybalchenko, I (2011), "Lewisite Metabolites
Detection in Urine by Liquid Chromatography-Tandem Mass Spectrometry", J. Chromatogr. B,
879(32), 3788-3796.
6. Wada, T., Nagasawa, E., Hanaoka, S. (2006), "Simultaneous Determination of Degradation
Products Related to Chemical Warfare Agents by High-Performance Liquid Chromatography/Mass
Spectrometry," Applied. Organomet. Chem. 20(9), 573-579.
7. Black, R.M., Muir, B. "Derivatisation reactions in the chromatographic analysis of chemical
warfare agents and their degradation products," J. Chromatogra. A, 1000 (2003) 253-281.
8. Tomes, J.A., Opstad, A.M., Johnsen, B.A. "Determination of organoarsenic warfare agents in
sediment samples from Skagerrak by gas chromatography-mass spectrometry,' Science of the Total
Environment, 356 (2006) 235- 246.
9. OSHA Safety and Health Standards, General Industry (29CRF1910). Occupational Safety and
Health Administration, OSHA 2206 (Revised, July 1, 2001).
10. Carcinogens-Working with Carcinogens, Publication No. 77-206. Atlanta, Georgia: Department of
Health, Education, and Welfare, Public Health Service, Center for Disease Control, National
Institute of Occupational Safety and Health, August 1977.
11. Safety in Academic Chemistry Laboratories, American Chemical Society Publication, Committee
on Chemical Safety, 7th Edition.
12. "Prudent Practices in the Laboratory: Handling and Disposal of Chemicals," National Academies
Press (1995), available at http://www.nap.edu (accessed August 15, 2013).
13. U. S. Army Public Health Command (2011) "Chemical Agent Health-Based Standards and
Guidelines Summary Table 2: Criteria for Water, Soil, Waste as of July 2011".
24
-------
14. Contaminants of Potential Concern (COPC) Committee of the World Trade Center Indoor Air Task
Force Working Group (2003) "World Trade Center Indoor Environment Assessment: Selecting
Contaminants of Potential Concern and Setting Health-Based Benchmarks", available at
http://www.epa.gov/wtc/reports/contaminants of concern benchmark studv.pdf. .
15. R Core Team. (2012) "R: A language and environment for statistical computing." R Foundation
for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0. URL http://www.R-project.org/.
25
-------
18 TABLES, FIGURES AND VALIDATION DATA
Item Title
Table 1. Materials tested for the Lewisite 1 degradation analysis 27
Table 2. Liquid Chromatography Gradient Conditions 28
Table 3. Recommended concentrations forLC-MS/MS calibration standards 28
Table 4. Scan segments associated with Lewisite 1 degradation product analysis by LC-MS/MS (LTQ Orbitrap) ..29
Table 5. Method detection limita (MDL) data for CVAOA in various matrices 30
Table 6. Surrogate (PAOA) recovery data collected during MDL study (from samples in Table 5) 30
Table 7. Sample holding time data for CVAA spiked on various matrices and analyzed as CVAOA 31
Table 8. PAOA surrogate data collected during sample holding time study 31
Table 9 . Statistical analysisa of sample holding time study data for CVAOA (See data in Table 2 .) 31
Table 10. CVAOA concentration changes in sample extracts; extracts were prepared on Day 7 and reanalyzed on
Day 16 (nine days of storage at 4 ± 2 °C) 32
Figure 1. Chemical structures of analytes and surrogates 2
Figure 2. Quality data collected during MDL study 33
Figure 3. LC-MS/MS response vs. analyte concentration in various solvents. Solvents change depending on matrix:
Water, 30% H2O2 added; Wipe samples, 10 mM HC1 with 30% H2O2; Sand/soil, 10 mM HC1, 25% MeOH, and
30% HO- 34
Figure 4. LC-MS/MS response vs. analyte concentration, additional data. Solvents change depending on matrix:
Wipe samples, 10 mM HC1 with 30% H2O2; Sand/soil, 10 mM HC1, 25% MeOH, and 30% H2O2 35
26
-------
Table 1. Materials tested for the Lewisite 1 degradation analysis
Material
Manufacturer^ endor
Water
(Millipore Milli-Q, Advantage A-10 system,
resistivity of 18.2 MQ.cm at 25 °C),
Millipore, Cincinnati, OH
Wipe
3" x 3" 12-ply gauze (Kendall Curity)
Sand, purified
JT Baker, Inc., Phillipsburg, NJ
Virginia Soil (VA soil)
Composition (64.5% sand, 28% silt, 7.5% clay,
2.6% TOC, pH 4.1 in 1:1 soil: water
mix)
Nebraska Aglands Ap Soil (NE soil)
Composition (5.1% sand, 57.5% silt, 31.7%
clay, 1.9% TOC, pH 5.5 in 1:1 soil:
water mix)
Georgia Bt2 soil (GA soil)
Composition (46% sand, 22% silt, 32% clay,
0.2% TOC, pH 5.0 in 1:1 soil: water
mix)
27
-------
Instrumental Conditions
Under these conditions, the elution times of CVAOA and PAOA are approximately 3.7 min and 7.0 min,
respectively. The elution times for CVAA and PAA (the absence of which provides assurances that
complete oxidation has occurred) are approximately 6.7 min and 12.5 min, respectively.
Table 2. Liquid Chromatography Gradient Conditions
Time
Flow
%
%
(min)
(jiL/min)
Solution A+
Solution B++
0
200
5
95
5
200
5
95
15
200
80
20
25
200
80
20
28
200
5
95
38
200
5
95
+A: Acetonitrile (0.1% Formic Acid)
+ +B: Water (0.1% Formic Acid)
*Injection volume - 10 |iL(rccommcndcd)
* Column Temperature: 30 ° C
*Equilibration time: 10 minutes
*Column: 100 A, 150 mm x 2.1mm, 3|_un particle size
Table 3. Recommended Concentrations for LC-MS/MS Calibration Standards
Concentration of Analyte in
Calibration Standard (ng/jiL)
Add Together the Following Volumes (jiL) to
Make Calibration Standard
Calibration
Level
CVAA
PAA
10 ng/jiL
CVAA
solution
10 ng/jiL
PAA
solution
Solvent3
1
0.02
0.02
4
4
2000
2
0.05
0.05
10
10
2000
3
0.08
0.08
16
16
2000
4
0.1
0.1
20
20
2000
5
0.15
0.15
30
30
2000
6
0.2
0.2
40
40
2000
a Solvents change depending on matrix: Water, 15% H2O2 added; Wipe samples, 5 mM HC1 with
15% H2O2; Sand/soil, 2.5 mM HC1, 25% MeOH, and 15% H202.
28
-------
Mass Spectrometer Conditions
Ionization mode: Positive electrospray
Capillary temperature: 300°C
Scan segments: The chromatographic run is divided into six scan segments, as summarized in
Table 2. This table describes parameters that are specific to the LTQ Orbitrap system used in
method development.
Table 4. Scan Segments Associated with Lewisite 1 Degradation Product Analysis by LC-MS/MS
(LTQ Orbitrap)
Segment
Retention Time
(min)
Scan Events
Type
Comment
1
0.00-3.15
100-750m/z
full scan
2
3.15-5.50
100-750m/z
full scan
[M+H]+ CVAOA
(187 m/z—>169 m/z)
187—>50-300 m/z
MS/MS3
(3 sequential scans),
CEb= 31%
187—>50-300 m/z
187—>50-300 m/z
3
5.50-6.86
100-750m/z
full scan
[M-H20+H]+ cvaa
(153 m/z—>127 m/z)
153—>50-300 m/z
MS/MS3
(3 sequential scans),
CEb = 25%
153—>50-300 m/z
153—>50-300 m/z
4
6.86-9.87
100-750m/z
full scan
[M+H]+ PAOA
(203 m/z—>185 m/z)
203—>55-300 m/z
MS/MS3
(3 sequential scans),
CEb= 30%
203—>55-300 m/z
203—>55-300 m/z
5
9.87-13.26
100-750m/z
full scan
169—>50-300 m/z
MS/MS3
(3 sequential scans),
CEb= 30%
[M-H20+H]+ paa
(169 m/z—>91 m/z)
169—>50-300 m/z
169—>50-300 m/z
6
13.26-30.00
100-750m/z
full scan
a for all MS/
VIS scans, Activation Q = 0.25, Activation Time = 30 ms, and Isolation Width =1.4 m/z
b Normalized Collision Energy
29
-------
Method Detection Limit Data
Table 5. Method Detection Limit3 (MDL) data
or CVAOA in Various Matrices
Matrix
Spiked
CVAOA
Concentratio
n
Measured
CVAOA
Concentration
(avg ± std)a
Average
Recovery
(%)
MDLb for
CVAOA
Spike:
MDL
ATLd
Water
0.20 mg/L
0.22 ±0.01 mg/L
110
0.041
mg/L
4.9
0.027 mg/Ld
Wipe
3.00 Mg
3.72 ± 0.14 (ig
124
0.44 (ig
8.5
4 (ige
Wipec f
3.00 Mg
3.04 ± 0.12 (ig
101
0.38 Mg
8.1
4 (ige
Sand
0.20 jj.g/g
0.17 ± 0.02 jj.g/g
85
0.073
Mg/g
2.3
0.3 jj.g/g d
Nebraska
(NE) Soil
0.20 jj.g/g
0.22 ±0.01 jag/g
112
0.032
Mg/g
7.1
0.3 jj.g/g d
Virginia
(VA) Soil
0.40 jj.g/g
0.17 ±0.01 jag/g
43
0.028
Mg/g
6.1
0.3 jj.g/g d
Georgia
(GA) Soil
0.40 jj.g/g
0.32 ± 0.02 jag/g
80
0.055
Mg/g
5.8
0.3 jj.g/g d
a Numbers are the average of seven independent samples plus/minus the standard deviation of
the measurements.
b Statistically determined MDL = sxt a=o .01, where s is the standard deviation of seven
measurements and t is the Student t value for 6 degrees of freedom (3.143).
c Data corrected for small amount of CVAOA in blank for second set of wipe data.
d Analyte target level (ATL), derived from U.S. Army Public Health Command (13).
e Analyte target level (ATL) was estimated, based on screening level for arsenic at ~ 400 (.ig/nr = 0.04
(.ig/ni2, which assuming a wipe sample from 100 cm2, yields 4 (ig (14).
fTwo sets of wipe data were collected at different times.
Table 6. Surrogate (PAOA) Recovery Data Collected During MDL Study (from Samples in Table
5)
Matrix
Spiked PAOA
Concentration
Measured PAOA
Concentration
(avg ± std)a
Average
Recovery
(%)
Water
0.20 mg/L
0.23 ±0.01 mg/L
110
Wipe
4.0 (ig
4.08 ±0.07 ng
102
Sand
0.40 jj.g/g
0.35 ± 0.03 jj.g/g
87
NE Soil
0.40 jj.g/g
0.34 ±0.01 jag/g
86
VA Soil
0.40 jj.g/g
0.11 ±0.01 jag/g
28
GA Soil
0.40 jj.g/g
0.30 ±0.01 jag/g
74
a Numbers are the average of seven independent samples plus/minus the
standard deviation of the measurements.
30
-------
Holding Time Study Data
Table 7. Sample Holding Time Data for CVAA Spiked on Various Matrices and Analyzed as
CVAOA
CVAOA Concentrations3
Matrix
Units
Spiked
Day 0
Day 2
Day 7
Day 14
Water
mg/L
0.40
0.41 ±0.01
0.39 ±0.01
0.38 ±0.02
0.40 ± 0.02
Wipe
Mg
4.00
5.09 ±0.12
4.71 ±0.19
4.85 ±0.12
4.80 ±0.20
Sand
mg/kg
0.40
0.39 ±0.02
0.29 ±0.02
0.22 ±0.04
0.16 ±0.02
NE soil
mg/kg
0.40
0.40 ± 0.02
0.24 ±0.01
0.12 ±0.00
0.11 ±0.01
a Numbers are the average of three independent samples plus/minus the standard deviation of the
measurements.
Table 8. PAOA Surrogate Data Collected During Sample Holding Time Study
PAOA Concentrations3
Matrix
Units
Spiked
Day 0
Day 2
Day 7
Day 14
Water
mg/L
0.20
0.20 ±0.01
0.18 ±0.00
0.19 ±0.00
0.20 ±0.00
Wipe
Mg
4.00
4.19 ± 0.10
3.93 ±0.06
3.95 ±0.05
4.29 ±0.09
Sand
mg/kg
0.40
0.38 ±0.01
0.36 ±0.01
0.36 ±0.01
0.65 ± 0.02b
NE soil
mg/kg
0.40
0.35 ±0.01
0.33 ±0.00
0.31 ±0.01
0.57 ± 0.03b
a Numbers are the average of three independent samples plus/minus the standard deviation of the
measurements.
b Incorrect spiking at 0.80 mg/kg (used same pipet as for wipes).
Table 9 . Statistical Analysis" of Sample Holding Time Study Data for CVAOA (See data in Table
2.)
Significance Leve
Matrix
t2 " to
ti - to
tl4 " to
Water
0.107
0.016
0.331
Wipe
0.025
0.105
0.068
Sand
<0.010
<0.010
<0.010
NE Soil
<0.010
<0.010
<0.010
a Significance levels (p values) for comparison of CVAOA concentrations at time t (days) and time
0; /;<().01 indicates statistically significant concentration decrease.
31
-------
Additional Holding Time Data
Table 10. CVAOA Concentration Changes in Sample Extracts; Extracts Were Prepared on Day 7
and F
.eanalyzed on Day 16 (nine days of storage at 4±2 °C)
CVAOA Concentrations
anc
n Sample Extracts Produced on Day 7
Reanalyzed on Day 16a
Matrix
Units
Spiked
Day 7
(t=0 days)
Day 16
(t=9 days)
Wipe
w?
4.00
4.85 ±0.12
4.75 ±0.25
Sand
mg/kg
0.40
0.22 ±0.04
ND
NE soil
mg/kg
0.40
0.12 ±0.00
0.10 ±0.00
a Numbers are the average of three independent samples p
us/minus the standard c
eviation of the
measurements.
32
-------
Quality Control Data Collected During MDL Study
0.16
J*
0.14
60
0.12
3
£
0.10
O
4->
0.08
ro
4->
0.06
£
a>
0
0.04
£
0.02
O
U
0.00
0.1 ug/mL QA Samples
~CVAOA
¦ PAOA
0 10 20 30 40
Time (hrs)
Figure 2. Quality data collected during MDL study.
The above data represent replicate analyses of the same continuing calibration verification (CCV) sample,
which contained 0.1 (ig/mL each of CVAOA and PAOA in 30% H2O2 in laboratory-grade water. These
data were collected during the period when data for the wipe, sand, and soil samples were collected (see
Tables 3 and 4). These data were collected overthe course of a sequence that contained a total of 65 samples
run over a period of approximately 40 hours. During this time, all CCVs remained within 35% of their
expected values.
33
-------
Calibration Data for Various Matrices
CVAOA Calibration
a>
cc
10000
8000
6000
4000
2000
0
0.00 0.05 0.10 0.15 0.20
Concentration (ug/mL)
~ Water
¦ Wipe
Soil
0.25
40000
35000
30000
25000
20000
15000
10000
5000
0
PAOA Calibration
--
~ Water
¦ Wipe
Soil
0.00 0.05 0.10 0.15 0.20
Concentration (ug/mL)
0.25
Figure 3. LC-MS/MS response vs. analyte concentration in various solvents. Solvents change
depending on matrix: Water, 30% H2O2 added; Wipe samples, 10 mM HC1 with 30% H2O2;
Sand/soil, 10 mM HC1, 25% MeOH, and 30% H202.
34
-------
IS)
+->
c
3
O
u
(0
O)
O)
(/)
c
o
Q.
(/)
O)
0£
CVAOA Calibration, 7/9/14
20000
15000
10000
5000
0
~ Wipe
¦ Soil
—i—
0.1
0.05 0.1 0.15 0.2
Concentration (ug/mL)
0.25
(/)
+->
C
3
O
u
ro
(U
(U
«/)
c
o
Q.
(/)
(U
DC
140000
120000
100000
80000
60000
40000
20000
0
PAOA Calibration, 7/9/14
~ Wipe
¦ Soil
0.05 0.1 0.15 0.2
Concentration (ug/mL)
0.25
Figure 4. LC-MS/MS response vs. analyte concentration, additional data. Solvents change
depending on matrix: Wipe samples, 10 mM HC1 with 30% H2O2; Sand/soil, 10 mM HC1, 25%
MeOH, and 30% H202.
35
-------
&EPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |