EPA/600/R-16/173 I October 2016
www.epa.gov/homeland-security-research
United States
Environmental Protection
Agency
&EPA
Fate and Transport of Chemical
Warfare Agents VX and HD across a
Permeable Layer of Paint or Sealant
into Porous Subsurfaces
Office of Research and Development
Homeland Security Research Program
-------
This page left Intentionally Blank
-------
EPA/600/R-16/173
October 2016
Fate and Transport of Chemical Warfare
Agents VX and HD across a Permeable
Layer Paint or Sealant into Porous
Subsurfaces
U.S. Environmental Protection Agency
Office of Research and Development
National Homeland Security Research Center
Research Triangle Park, NC 27711
-------
This page left Intentionally Blank
-------
Disclaimer
The U.S. Environmental Protection Agency, through its Office of Research and Development's
National Homeland Security Research Center funded and managed the research described here
under contract EP-C-11-038, Task Order 29 with Battelle. This report has been subjected to the
Agency's peer and administrative 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:
Lukas Oudejans, Ph.D.
Decontamination and Consequence Management Division
National Homeland Security Research Center
Office of Research and Development
U.S. Environmental Protection Agency (MD-E343-06)
109 T.W. Alexander Drive
Research Triangle Park, NC 27711
Phone: 919-541-2973
Fax: 919-541-0496
E-mail: Oudeians.Lukas@epa.gov
-------
Acknowledgments
The following individuals and organizations are acknowledged for their role in conducting the
research, preparation of the draft and review of this document:
Project team members (US Environmental Protection Agency (EPA)):
Office of Research and Development (ORD), National Homeland Security Research Center
(NHSRC):
Lukas Oudejans (Principal Investigator);
Stuart Willison;
Matthew Magnuson; and
Sang Don Lee
Office of Emergency Management (OEM), CBRN Consequence Management Advisory Division
(CM AD):
Terry Smith and
Larry Kaelin
EPA Peer Review
Romy Campisano, ORD/NHSRC;
Elise Jakabhazy, OEM/CMAD; and
Dan Stout, ORD/NERL
Battelle
ii
-------
Executive Summary
Under the U.S. Environmental Protection Agency's (EPA) Homeland Security Research
Program (HSRP), EPA develops procedures and methods to contain and mitigate chemical,
biological, and radiological (CBR) contamination and to remediate the environment following
public health and environmental incidents and disasters. EPA's goal is to protect human health
and restore the use of contaminated areas as efficiently as possible. Within the HSRP, EPA's
National Homeland Security Research Center (NHSRC) is tasked with planning for and
responding to, among others, releases of chemicals into the environment, including the deliberate
release of chemical warfare agents (CWAs). A CWA release could result in the deposition of
CWA onto painted or sealed/coated surfaces. The intent of this investigation was to study the
fate and transport of CWA applied to painted/sealed materials including the potential partitioning
of CWA into permeable paints/sealants and subsequently into underlying porous materials. Over
time, CWAs might evaporate from painted/sealed surfaces, remain in the permeable paint/sealant
layers, or diffuse into the underlying porous or absorptive materials. Information on CWA
persistence and partitioning into painted and sealed surfaces may aid in development of more
appropriate response and remediation actions. Two relatively persistent CWAs, VX (O-ethyl S-
[2-(diisopropylamino)ethyl] methylphosphonothioate) and sulfur mustard [HD, bis(2-
chloroethyl)sulfide], were investigated.
The primary project objectives and the approach were to investigate the fate and transport,
primarily persistence and partitioning, of VX and HD across painted or sealed surfaces. Briefly,
VX or HD was spiked onto painted/sealed stainless steel, freestanding (FS) paint films or
sealants placed over a porous material (solid phase extraction [SPE] disks), and unpainted
stainless steel, which served as a reference material. SPE disks were used as a representative
substance for a porous, absorbing material instead of the material itself as it is inherently
problematic to isolate a paint layer from a porous or uneven subsurface.
Three types of paint (latex flat, latex semi-gloss, and oil gloss) and two types of sealant (epoxy
and polyurethane) were used to paint or seal the stainless steel and create FS paint or sealant
films. The spiked materials were held at ambient laboratory conditions for four (paint) or three
(sealant) dwell times ranging from 3 hours to 48 hours (HD) or 72 hours (VX). Longer dwell
times were considered for VX compared to HD based on the lower volatility of VX compared to
HD volatility. Following the holding times, the materials (approximately 10 square centimeters
surface area) were wipe-sampled for VX or HD, and then the entire material coupon was
extracted and analyzed for VX or HD. For tests with SPE disks beneath FS paint or sealant films,
the SPE disks and the FS paint/sealant films were extracted and analyzed separately, but only the
FS paint/sealant films were wipe-sampled. The VX and HD recoveries were used to determine
the amount of CWA readily accessible on the material surfaces (wipe samples), the amount of
CWA more strongly associated with the material surfaces or embedded in the paint/sealant
(coupon samples), and the amount of CWA that migrated through the paint/sealant into porous
material beneath the paint/sealant (SPE disk samples).
Painted Surfaces Fate and Transport Results
Fate and transport results for VX associated with painted surfaces are summarized in Figure
Executive Summary (ES)-l. The amount of VX applied to the surface was 1470 |ig based on
iii
-------
recovery from VX spike controls. The amount of VX recovered from wipe sampling (as well as
the total amount of VX recovered) generally decreased over the 72-hour dwell time. VX was also
recovered directly from coupon (unpainted and painted stainless steel and FS paint films)
extraction (after wipe sampling) as well as from the porous material below the paint layer, (SPE
disk) by extraction. In Figures ES-1 through ES-4, the red "coupon" bars, refer to the amount of
CWA recovered from coupon extraction after the coupons were wipe sampled. Wipe sampling
recovered the majority of VX from unpainted stainless steel, latex flat painted stainless steel, and
oil gloss painted steel. The amount of VX recovered from coupon and SPE disk extraction
represented a larger portion of the total VX recovery for the remaining materials (i.e., latex semi-
gloss painted stainless steel [no SPE] and all three of the FS paint films overlaying SPE disks).
The amount of VX recovered from the SPE disks tended to increase with time, especially at the
24-hour dwell time. At 24 hours, the mean VX mass levels recovered from the SPE disks
differed significantly between the oil gloss paint film and the latex flat paint films, while at 48
hours, all three paint films differed significantly from each other (at the 95% confidence level).
These observations indicate that VX permeation into and through a paint layer is likely to occur.
The permeated amount appears to be paint type specific.
1200
_ 1000
I
ฆC
a>
*_
a>
>
o
u
a>
cc
x
>
ro
a>
800
600
400
200
SPE
I Coupon
I Wipe
3 6 2472
Unpainted
Stainless
Steel
3 6 2472
Painted
Stainless
Steel:
Latex Flat
3 6 2472
Painted
Stainless
Steel: Latex
Semi-Gloss
3 6 2472
Painted
Stainless
Steel: Oil
Gloss
3 6 2472
FS Paint
Film: Latex
Flat and
SPE
Dwell Time (hours) and Material
3 6 2472
FS Paint
Film: Latex
Semi-Gloss
and SPE
3 6 2472
FS Paint
Film: Oil
Gloss and
SPE
Figure ES-1. Painted surfaces - VX recovered from wipes, coupons (after wipe), and SPE
disks.
Fate and transport results for HD associated with painted surfaces are summarized in Figure ES-
2. The amount of HD applied to the surface was 2500 jag based on the recovery from HD spike
controls. The amount of HD recovered via wipe sampling decreased rapidly over the 48-hour
dwell time. However, total HD recoveries remained relatively high (>1500 micrograms [|ig])
over the 48-hour dwell time (except for unpainted stainless steel and oil gloss painted stainless
steel) due to substantial amounts of HD recovered from the coupons and SPE disks. The amount
iv
-------
of HD recovered from the porous material below the paint (SPE disks) was noticeably greater
after the 24- and 48-hour dwell times than after the 3- and 6-hour dwell times for each paint film
type. At the 6- and 48-hour dwell times, the mean HD mass levels recovered from the SPE disks
differed significantly among all three paint films (each at the 0.05 level), while at 24 hours, a
significant difference was noted among the oil gloss paint film and the other two paint film
types. These observations show that HD permeation into and through a paint layer is likely to
occur while the amount of HD permeated was found to be paint type specific.
2500
a
ฆa
ai
*_
a>
>
o
u
a>
cc
a
x
ro
a>
2000
1500
1000
500
SPE
I Coupon
I Wipe
3 6 2448
Unpainted
Stainless
Steel
3 6 2448
Painted
Stainless
Steel: Latex
Flat
3 6 2448
Painted
Stainless
Steel: Latex
Semi-Gloss
3 6 2448
Painted
Stainless
Steel: Oil
Gloss
3 6 2448
FS Paint
Film: Latex
Flat and
SPE
3 6 2448
FS Paint
Film: Latex
Semi-Gloss
and SPE
3 6 2448
FS Paint
Film: Oil
Gloss and
SPE
Dwell Time (hours) and Materials
Figure ES-2. Painted surfaces - HD recovered from wipes, coupons (after wipe), and SPE
disks.
Sealed Surfaces Fate and Transport Results
Fate and transport results for VX associated with sealed surfaces are summarized in Figure ES-3.
The amount of VX applied to the surface was 2100 jag based on recovery from VX spike
controls. For sealed surfaces, the amount of VX recovered from coupon extraction represented
relatively large proportions of the VX recovered compared to wipe sampling, especially as the
dwell time increased. Smaller proportions of VX were also found to penetrate through FS
polyurethane sealant into the SPE disk. Any VX mass levels that could penetrate through FS
epoxy sealant into the SPE disks were below the quantitation limits at each tested dwell time.
Fate and transport results for HD associated with sealed surfaces are summarized in Figure ES-4.
The amount of HD applied to the surface was 2700 jag based on the recovery from HD spike
controls. As observed with VX on sealed surfaces, the amount of HD recovered from coupon
v
-------
extraction represented relatively large proportions of the HD recovered compared to wipe
sampling, especially as the dwell time increased. In contrast to the results associated with VX
and sealed surfaces, relatively large proportions of HD were also found to penetrate through FS
epoxy sealant and FS polyurethane sealant into the SPE disks at the 24- and 48-hour dwell times.
The HD mass levels penetrating through sealant to the SPE disks differed significantly between
the two sealant types at these two dwell times, with higher levels seen with the polyurethane
sealant.
1800
1500
ฃ 1200
900
600
300
6 24 72
Unsealed
Stainless Steel
6 24 72
Sealed
Stainless Steel:
Epoxy
6 24 72
Sealed
Stainless Steel:
Polyurethane
SPE
I Coupon
I Wipe
6 24 72
FS Sealant:
Epoxy and SPE
6 24 72
FS Sealant:
Polyurethane
Dwell Time (hours) and Materials
Figure ES-3. Sealed surfaces - VX recovered from wipes, coupons (after wipe), and SPE
disks.
vi
-------
2800
DO
T3
ai
ai
>
o
2400
2000
1600
1200
800
400
6 24 48
Unsealed
Stainless Steel
6 24 48
Sealed
Stainless Steel:
Epoxy
6 24 48
Sealed
Stainless Steel:
Polyurethane
6 24 48
FS Sealant:
Epoxy and SPE
Dwell Time (hours) and Materials
SPE
I Coupon
I Wipe
6 24 48
FS Sealant:
Polyurethane
and SPE
Figure ES-4. Sealed surfaces - HD recovered from wipes, coupons (after wipe), and SPE
disks.
Impact of the Study:
Based on the results obtained from this investigation, VX and HD have the ability to permeate
into paints and sealants, including in some cases the underlying porous materials. It is likely that
other permeable materials besides paints and sealants may also show similar behavior. Wipe
sampling may capture only a fraction of the VX and HD retained in paint/sealants and/or
underlying porous materials. In fact, no HD was detected by wipe sampling of the paint after 24
and 48 h for all three paint types in the presence of the porous material (SPE disk); such would
result in false negatives that would indicate that no HD is present. The extraction of the
underlying porous material indicated a potential reservoir that allowed HD to persist at levels
much higher than would be indicated by wipe sampling and coupon (paint/sealant) extraction. As
such, decontamination procedures should address porous materials underlying painted/sealed
surfaces that may harbor CWA mass. This study did not address whether and how much of these
chemical agents diffuse back from the porous material below the paint or sealant into the paint
and back to the surface. Such process would result in a continuous exposure hazard, either by
contact or inhalation, even after the initial surface decontamination.
vii
-------
Table of Contents
Disclaimer i
Acknowledgments ii
Executive Summary iii
Table of Contents viii
List of Figures x
List of Tables xi
Acronyms and Abbreviations xii
1.0 Introduction 1
1.1 Purpose 1
1.2 Project Objectives 1
1.3 Test Facility Description 2
2.0 Experimental Methods 3
2.1 Sample Process Design 3
2.2 Experimental Design 3
2.2.1 Painted Surfaces 3
2.2.2 Sealed Surfaces 4
2.3 Method Development and Demonstration 5
2.3.1 Painted Surfaces - Wipe Sampling of VX and HD from Coupons 5
2.3.2 Painted Surfaces - Extraction of VX and HD from Coupons and SPE Disks 7
2.3.3 Sealed Surfaces - Wipe Sampling of VX and HD from Coupons 8
2.4 Test Matrices 8
2.4.1 Painted Surfaces 8
2.4.2 Sealed Surfaces 10
2.5 Test Materials 12
2.5.1 Unpainted and Unsealed Stainless Steel 12
2.5.2 Painted and Sealed Stainless Steel 12
2.5.3 FS Paint Films and FS Sealants 14
2.5.4 SPE Disks and LVAP Method 14
2.6 Chemical Agent, Coupon Spiking, and Dwell Conditions 17
2.7 Extraction of VX and HD 17
2.8 Analytical Methods for VX and HD 18
3.0 Test Results 22
3.1 Painted and Sealed Surfaces - Thickness 22
3.2 Method Development Results 22
3.2.1 Painted Surfaces - Wipe Method Development and Demonstration Results 22
viii
-------
3.2.2 Painted Surfaces - Extraction Method Development Results 26
3.2.3 Sealed Surfaces - Wipe Method Demonstration Results 28
3.3 Fate and Transport Results 30
3.3.1 Painted Surfaces - VX 30
3.3.2 Painted Surfaces - HD 35
3.3.3 Sealed Surfaces - VX 39
3.3.4 Sealed Surfaces - HD 42
4.0 Quality Assurance/Quality Control 46
4.1 Control of Monitoring and Measuring Devices 46
4.2 Equipment Calibrations 48
4.3 Technical Systems Audit 49
4.4 Performance Evaluation Audits 50
4.5 Data Quality Audit 51
5.0 Summary 52
5.1 Painted Surfaces 52
5.2 Sealed Surfaces 55
6.0 References 58
Appendix A 59
ix
-------
List of Figures
Figure ES-1. Painted surfaces - VX recovered from wipes, coupons (after wipe), and SPE disks
Figure ES-2. Painted surfaces - HD recovered from wipes, coupons (after wipe), and SPE disks
Figure ES-3. Sealed surfaces - VX recovered from wipes, coupons (after wipe), and SPE disks
Figure ES-4. Sealed surfaces - HD recovered from wipes, coupons (after wipe), and SPE disks v:
Figure 1. Low Volatility Agent Permeation (LVAP) test cell diagram 1.
Figure 2. Low Volatility Agent Permeation (LVAP) test cell assembly, stepwise 1
Figure 3. VX and HD recoveries from unpainted stainless steel for various wipe combinations
(method development) 2
Figure 4. VX and HD wipe recoveries from unpainted stainless steel and FS paint films (method
demonstration) 2>
Figure 5. VX and HD extraction recoveries from unpainted stainless steel, FS paint films, and SPE
disk 2
Figure 6. VX and HD extraction recoveries from unpainted stainless steel, FS paint films, and SPE
disk 2'
Figure 7. Photograph of a FS paint film (lifted) showing hole in the center of the coupon 3'
Figure 8. Mean VX mass recovered from unpainted and painted stainless steel (wipe and coupon
[after wipe] samples) 3
Figure 9. Mean VX mass recovered from FS latex flat paint films and underlying SPE disks (wipe,
coupon [after wipe], and SPE disk samples) 3
Figure 10. Mean VX mass recovered from FS latex semi-gloss paint films and underlying SPE disks
(wipe, coupon [after wipe], and SPE disk samples) 3
Figure 11. Mean VX mass recovered from FS oil gloss paint films and underlying SPE disks (wipe,
coupon [after wipe], and SPE disk samples) 3
Figure 12. Mean HD mass recovered from unpainted and painted stainless steel (wipe and coupon
samples) 3
Figure 13. Mean HD mass recovered from FS latex flat paint films and underlying SPE disks (wipe,
coupon [after wipe], and SPE disk samples) 3
Figure 14. Mean HD mass recovered from FS latex semi-gloss paint films and underlying SPE disks
(wipe, coupon [after wipe], and SPE disk samples) 3
Figure 15. Mean HD mass recovered from FS oil gloss paint films and underlying SPE disks (wipe,
coupon [after wipe], and SPE disk samples) 3
Figure 16. Mean VX mass recovered from unsealed stainless steel (wipes) and sealed stainless steel
(wipe and coupon samples) 4
Figure 17. Mean VX mass recovered from FS sealants (epoxy and polyurethane) and underlying
SPE disks (wipe, coupon [after wipe], and SPE disk samples) 4
Figure 18. Mean HD mass recovered from unsealed stainless steel (wipes) and sealed stainless steel
(wipe and coupon samples) 4
Figure 19. Mean HD mass recovered from FS sealants (epoxy and polyurethane) and SPE disks
(wipe, coupon [after wipe], and SPE disk samples) 4
-------
List of Tables
Table 1. Painted Surfaces - Wipe Sampling Method Development and Demonstration 6
Table 2. Painted Surfaces Test Matrix for Fate and Transport Investigation 10
Table 3. Sealed Surfaces Test Matrix for Fate and Transport Investigation 11
Table 4. Paint and Sealant Information 13
Table 5. Gas Chromatography/Mass Spectrometry Conditions for VX Analysis 19
Table 6. Gas Chromatography/Mass Spectrometry Conditions for HD Analysis 19
Table 7. Dry Paint and Sealant Film Thicknesses 22
Table 8. VX and HD Recoveries from Unpainted Stainless Steel for Various Wipe Combinations (Method
Development) 24
Table 9. VX and HD Wipe Recoveries from Unpainted Stainless Steel and FS Paint Films (Method
Demonstration) 25
Table 10. VX and HD Extraction Recoveries from Unpainted Stainless Steel, FS Paint Films and SPE
Disk 27
Table 11. DIC Stabilizer Evaluation Results 28
Table 12. VX and HD Wipe Recoveries from Sealed Stainless Steel 29
Table 13. Painted Surfaces - VX Recoveries from Wipes, Coupons, and SPE Disks1, 31
Table 14. Painted Surfaces - HD Recoveries from Wipes, Coupons, and SPE Disks1 36
Table 15. Sealed Surfaces - VX Recoveries from Wipes, Coupons, and SPE Disks'1 40
Table 16. Sealed Surfaces - HD Recoveries from Wipes, Coupons, and SPE Disks1 43
Table 17. Data Quality Indicators and Results 46
Table 18. Equipment Calibration Schedule 48
Table 19. Gas Chromatography Performance Parameters and Acceptance Criteria 49
Table 20. Performance Evaluation Results 50
Table 21. Percent VX Recoveries from the Painted Surfaces Investigation 53
Table 22. Percent HD Recoveries from the Painted Surfaces Investigation 54
Table 23. Percent VX Recoveries from the Sealed Surfaces Investigation 56
Table 24. Percent HD Recoveries from the Sealed Surfaces Investigation 57
Table A-l. Environmental conditions during the method development / demonstration and subsequent
fate and transport tests of VX and HD on paints and sealants 58
xi
-------
Acronyms and Abbreviations
ฐc
degrees Celsius
CBR
chemical, biological, and radiological
CCV
continuing calibration verification
cm
centimeter(s)
cm2
square centimeter(s)
cm3
cubic centimeter(s)
CWA
chemical warfare agent
DIC
N,N'-diisopropylcarbodiimide
DoD
U.S. Department of Defense
DRC
double reinforced crepe
EPA
U.S. Environmental Protection Agency
ES
Executive Summary
FID
flame ionization detector
FOD
frequency of detection
FS
Freestanding
GC
gas chromatography
HD
sulfur mustard, bis(2-chloroethyl) sulfide
HMRC
Hazardous Materials Research Center
HPLC
high performance liquid chromatography
HSRP
Homeland Security Research Program
IPA
isopropyl alcohol
IS
internal standard
kHz
kiloHertz
LVAP
Low Volatility Agent Permeation
Hg
microgram(s)
|iL
microliter(s)
|im
micrometer(s)
mil
a thousandth of an inch
min
minute(s)
mL
milliliter(s)
mm
millimeter(s)
MS
mass spectrometry
NHSRC
National Homeland Security Research Center
NIST
National Institute of Standards and Technology
ORD
Office of Research and Development
PE
performance evaluation
PTFE
polytetrafluoroethylene
QA
quality assurance
QAPP
quality assurance project plan
r2
coefficient of determination
RH
relative humidity
Rpm
rotations per minute
SD
standard deviation
SPE
solid phase extraction
TSA
technical systems audit
xii
-------
O-ethyl S-(2-[diisopropylamino]ethyl) methylphosphonothioate
xiii
-------
1.0 Introduction
Under the U.S. Environmental Protection Agency's (EPA) Homeland Security Research
Program (HSRP), EPA develops procedures and methods to contain and mitigate chemical,
biological, and radiological (CBR) contamination and to remediate the environment following
public health and environmental incidents and disasters. EPA's goal is to protect human health
and restore the use of contaminated areas as efficiently as possible. Within the HSRP, EPA's
National Homeland Security Research Center (NHSRC) is tasked with planning for and
responding to, among others, releases of chemicals into the environment, including the deliberate
release of chemical warfare agents (CWAs). Stakeholders to the HSRP identified a need to better
understand the fate and transport of CWAs applied to porous or permeable surfaces such as
painted/sealed surfaces, which could be contaminated following a CWA release. Over time,
CWAs might evaporate from painted/sealed surfaces, remain in the permeable paint/sealant
layers, or diffuse into the underlying porous or absorptive materials. Information on CWA
persistence and partitioning into painted and sealed surfaces may aid in development of more
appropriate response actions.
1.1 Purpose
The purpose of this project was to investigate the fate and transport of two CWAs: VX (O-ethyl
S-[2-(diisopropylamino)ethyl] methylphosphonothioate) and sulfur mustard [HD, bis(2-
chloroethyl) sulfide] applied to painted surfaces and sealed surfaces. In particular, data were
collected to determine the amount of VX and HD that remains amenable to wipe sampling, the
amount of VX and HD recovered by extraction of the materials (i.e., the amount of VX and HD
more strongly bound to the materials and/or embedded within the paint and sealant), and the
amount of VX and HD that penetrates into porous materials beneath the paints/sealants.
Sampling was conducted at three to four time points to allow persistence and changes in
partitioning to be monitored at different dwell times (<72 hours).
1.2 Project Objectives
The project objectives were to investigate the fate and transport of VX and HD on unpainted
stainless steel, painted stainless steel, and freestanding (FS) paint films placed over solid phase
extraction (SPE) disks. SPE disks were used as a representative substance for a porous,
absorbing material instead of the material itself. It is intrinsically difficult to separate paint from
a porous or uneven surface. Mechanical removal would likely leave some paint on the material
while chemical removal would lead to difficulties in analysis and the likelihood that the chemical
agent would react leading to poor recoveries. Hence the approach to mimic the transport into a
porous material using a FS paint layer in contact with a SPE disk.
Similarly, the project objectives for the sealants were to investigate the fate and transport of VX
and HD on unsealed stainless steel, sealed stainless steel, and FS sealants placed over SPE disks.
SPE disks were used as a representative substance for a porous, absorbing material instead of the
material itself.
1
-------
1.3 Test Facility Description
All testing was performed at the Battelle Hazardous Materials Research Center (HMRC) located
in West Jefferson, Ohio. The HMRC is certified to work with chemical surety material through
its Bailment Agreement W911SR-10-H-0001 with the U.S. Department of the Army.
2
-------
2.0 Experimental Methods
2.1 Sample Process Design
For this project, most samples consisted of coupons (small representative pieces of
approximately 10 square centimeters [cm2]) of various materials that were spiked with VX or
HD. Coupons were comprised of unpainted/unsealed stainless steel, painted stainless steel,
sealed stainless steel, FS paint films, and FS sealants as these materials were spiked with CWA.
While SPE disks were also utilized on this project, they are not referred to as coupons because
they were not directly spiked with VX or HD during the fate and transport testing; the SPE disks
were directly spiked with CWA only during the extraction efficiency method demonstration.
SPE disks were not wipe-sampled but were only analyzed by direct extraction. The spiked
coupons were protected from air flow and held under ambient laboratory conditions for specified
dwell times. At designated times, the samples were wiped and/or extracted and then analyzed to
determine the amount of VX or HD remaining on the sample.
2.2 Experimental Design
2.2.1 Painted Surfaces
A multiple group time-series experimental design was used. The same design was replicated for
VX and HD and for three paint types (latex flat, latex semi-gloss, and oil gloss). The
experimental design is represented as:
unpainted stainless steel, wipe, n:
Oi
Oi
O3
04
unpainted stainless steel, extraction, n:
Oi
O2
O3
04
painted stainless steel, wipe, n:
Oi
O2
O3
04
painted stainless steel, extraction, n:
Oi
O2
O3
04
FS paint film, wipe, n:
Oi
O2
O3
04
FS paint film, extraction, n:
Oi
O2
O3
04
SPE disk, extraction, n:
Oi
O2
O3
04
For this experimental design, time passes from left to right. The experimental condition (n)
represents the combination of CWA, paint type, material type, and sampling method (wipe or
extraction) being tested. The material coupons were spiked with VX or HD (at time zero) and
randomly assigned to groups that were extracted at each of four dwell times (t). The mean
masses of VX or HD recovered via wiping or extraction after a given time (t) were the
experimental results. Symbolically, the results for an experimental condition (n) were specified
as an observation (O) at time (t); Ot. Note the times allowed for partitioning were different for
VX and HD because of the higher volatility of HD.
3
-------
Unpainted stainless steel served as a reference material for the three paint types. CWA spiked
onto unpainted stainless steel was not expected to penetrate or react with this nonporous
material. Any loss of CWA from unpainted steel was therefore attributed to evaporation from
unpainted steel. The SPE disks are absorbent porous materials into which CWA may diffuse.
CWA might evaporate or partition into the SPE disk or remain in the FS paint film. CWAs may
also degrade at the painted surface although this process is expected to be of lower relevance.
To address the study objectives, the following two statistical hypothesis tests were formulated:
Test#l (one-tailed test):
Null hypothesis: The mean mass of CWA diffused into the porous layer (SPE disk)
underlying an FS paint film is equal to 2.5 micrograms (|ig) (quantitation limit).
Alternative hypothesis: The mean mass of CWA diffused into the porous layer (SPE
disk) underlying a FS paint film exceeds 2.5 |ig.
Test #2 (two-tailed test):
Null hypothesis: The mean mass of CWA diffused into the porous layer (SPE disk)
underlying a FS paint film is equal among the two specified paint types.
Alternative hypothesis: The mean mass of CWA diffused into the porous layer (SPE
disk) underlying a FS paint film differs among the two specified paint types.
A Student-t test was applied at the 0.05 significance level to the project data to perform Test #1,
while a Student's two-sample t-test was applied at the 0.05 significance level to perform Test #2.
In addition to presenting the results of the hypothesis tests, this report presents descriptive
statistics to show the CWA recovered from the unpainted stainless steel, painted stainless steel,
FS paint film, and SPE disk over time for each combination of paint and CWA.
2.2.2 Sealed Surfaces
For tests with sealants, a multiple group time-series experimental design was also used. The
same design was replicated for VX and HD and the two sealant types (polyurethane and epoxy).
The experimental design is represented below, with time passing from left to right:
unsealed stainless steel, wipe, n:
Oi
Oi
O3
sealed stainless steel, wipe, n:
Oi
O2
O3
sealed stainless steel, extraction, n:
Oi
O2
O3
FS sealant, wipe, n:
Oi
O2
O3
FS sealant, extraction, n:
Oi
O2
O3
SPE, extraction, n:
Oi
O2
O3
4
-------
The experimental condition (n) represents the combination of CWA, sealant type, material type,
and sampling method (wipe and/or extraction) being tested. The unsealed stainless steel was not
extracted following the wiping of the surface. The material coupons were spiked with VX or HD
(at time zero) and randomly assigned to groups that were extracted at each of three dwell times
(t). The mean masses of VX or HD recovered via wiping or extraction after a given time (t) were
the experimental results. Symbolically, the results for an experimental condition (n) were
specified as an observation (O) at time (t); Ot. Note the times allowed for partitioning were
different for VX and HD because of the higher volatility of HD.
To address the study objectives, the following two statistical hypothesis tests were formulated:
Test #1 (one-tailed test):
Null hypothesis: The mean mass of CWA diffused into the porous layer (SPE disk)
underlying a FS sealant is equal to 2.5 |ig (quantitation limit).
Alternative hypothesis: The mean mass of CWA diffused into the porous layer (SPE
disk) underlying a FS sealant exceeds 2.5 |ig.
Test #2 (two-tailed):
Null hypothesis: The mean mass of CWA diffused into the porous layer (SPE disk)
underlying a FS sealant is equal between the two specified sealant types.
Alternative hypothesis: The mean mass of CWA diffused into the porous layer (SPE
disk) underlying a FS sealant differs between the two specified sealant types.
A Student's two-sample t-test was applied at the 0.05 significance level to the project data to
address Test #2.
In addition to presenting the results of the hypothesis tests, descriptive statistics are presented to
show the CWA recovered from the unsealed stainless steel, sealed stainless steel, FS sealant, and
SPE over time for each combination of sealant and CWA.
2.3 Method Development and Demonstration
Coupons for the method development and demonstration were prepared as described in Section
2.5. Method development and demonstration work was primarily conducted with coupons
mimicking painted surfaces. The methods used for painted surfaces were then demonstrated on
sealed surfaces.
2.3.1 Painted Surfaces - Wipe Sampling of VX and HD from Coupons
The wipe sampling was performed as described in Table 1. The first row under the header row
describes the method development for evaluating alternative wipe types, solvents, and solvent
volumes for recovering VX and HD from unpainted stainless steel coupons. Unpainted stainless
steel coupons were spiked with VX or HD as described in Section 2.6.
After holding for 60 minutes (min) (dwell time), the test coupons were sampled using alternative
wipes (two types), solvents (three types), and solvent volumes (two volumes). One wipe was a
lint-free 5 centimeters (cm) x 5 cm four-ply rayon/polyester blend sponge (22-037-921, Fisher
5
-------
Scientific, Pittsburgh, PA). This wipe is referred to as gauze. The second wipe was a double
reinforced crepe (DRC) wipe (3052, Ace-Tex Enterprises, Detroit, MI), which was cut to similar
dimensions as the gauze (5 cm x 5 cm). Both of these wipe types were used as received without
pre-cleaning and have been successfully used in similar CWA testing.
Table 1. Painted Surfaces - Wipe Sampling Method Development and Demonstration
Coupon Types
Events
Coupons Per
Event*
Coupons Required (Agent
Types x Coupons Per Event)
1 (unpainted stainless
steel)
3 solvents x 2 wipe types x 2
solvent volumes =12 events
3 test coupons;
1 procedural blank
coupon^
1 evaporation control
coupon
120 unpainted stainless steel
coupons
4 (unpainted stainless
steel and 3 types of FS
paint films: latex flat,
latex semi-gloss, and oil
gloss)
2 solvents/wipe type/solvent
volume combinations = 2
events per CWA (although
onlv 1 event was run for 1
CWA)#
3 test coupons;
1 procedural blank
coupon^
1 evaporation control
coupon
60 coupons (unpainted stainless
steel [15], latex flat [15], latex
semi-gloss [15], and oil gloss
[15])
* Each combination of agent, solvent, wipe type, and solvent volume.
' Procedural blank coupons were not spiked with CWA but are handled similarly to the test coupons.
* Evaporation control coupons were spiked with CWA and handled similarly to the test coupons but were directly
extracted for CWA rather than being wipe-sampled.
* During testing with VX, one solvent (acetone) was found to be incompatible with paint (specifically latex flat
paint). Therefore, testing was not conducted with acetone and HD.
The three solvents were n-hexane (>95%, H306-K4, Fisher Scientific, Pittsburgh, PA), acetone
(99.9% A929SK-4, Fisher Scientific, Pittsburgh, PA), and isopropyl alcohol (IPA; 99.9%,
A464SK-4, Fisher Scientific, Pittsburgh, PA). Hexane was selected because it has been used
routinely for both VX and HD extraction in previous studies [Stone 2016], and acetone was
selected for testing because it has been routinely used to extract CWA from SPE disks. IPA was
selected because it has been used for both VX and HD wipe sampling as a wetting solvent (EPA,
2007). The solvents were high performance liquid chromatography (HPLC)-grade or better to
minimize water and impurity content.
Two volumes of solvent to be applied to the wipes were determined: an amount that was
approximately saturating for the wipe; and an amount that was approximately half-saturating for
the wipe. These two volumes were determined by weighing three wipes before and after soaking
the wipes in the selected solvent. The amount of solvent remaining on the wipe after 30 seconds
after immersion in solvent and hanging vertically to allow excess solvent to drip off served as the
first volume. Half the amount remaining on the wipes after 30 seconds hanging vertically served
as the second volume.
The sampling method included the following steps:
The wipes were wetted with the specified volume of solvent prior to use.
Each coupon was wiped using four horizontal and four vertical strokes. Given the
small surface area of the coupons, strokes were short (coupon length) and overlapped
6
-------
each other. The wipes were folded in half for an approximate 2.5 cm x 5 cm wiping
contact area, as necessary, for manageability during wiping.
Wipes were extracted in solvent (the same solvent used to wet the wipe) as described
in Section 2.7.
Extracts were analyzed for VX and HD by gas chromatography (GC)/mass
spectrometry (MS) as described in Section 2.8.
The wipe types, solvents, and solvent volume combinations that were used for the method
demonstration were based on the data from the evaluation of alternative wipe methods on
unpainted stainless steel. The selected methods were demonstrated as described in the last row of
Table 1. Three types of FS paint film coupons as well as unpainted stainless steel coupons were
spiked as described in Section 2.6 and held for 60 min (dwell time). The coupons were then
wipe-sampled using the selected methods, and the wipes were extracted and analyzed as
described in Sections 2.7 and 2.8, respectively. The solvents selected for wiping were also used
during the wipe extraction process. The data from the method demonstration resulted in the wipe
type, solvent, and solvent volume combination that was used for this investigation (based on
CWA recoveries and solvent compatibility with the paint).
2.3.2 Painted Surfaces - Extraction of VX and HD from Coupons and SPE Disks
Extraction recovery efficiency was demonstrated for VX and HD from unpainted stainless steel
coupons, isolated FS paint film coupons (latex flat, latex semi-gloss, and oil gloss), and SPE
disks. Coupons and SPE disks, in triplicate, were spiked with 2 |iL of neat VX or HD as
described in Section 2.6 and protected from air flow during a 60-min dwell time. After 60 min,
the materials were extracted using hexane for the unpainted stainless steel and FS paint film
coupons and using acetone for the SPE disks. The extracts were analyzed as described in Section
2.8.
Three spike controls (a spike of equal amounts of VX or HD, as appropriate, directly applied into
hexane) and a single laboratory blank per material were also included. The laboratory blank was
a coupon of each material that was handled in the same way as the test coupons, except not
exposed to VX or HD (i.e., the laboratory blanks were never placed inside the agent hood). One
procedural blank (an unspiked coupon handled similarly to the test coupons) and one evaporation
control (a spiked coupon handled similarly to the test coupons but directly extracted for CWA
rather than being wipe-sampled) were included per paint and CWA type).
The extraction method was considered acceptable for use in the subsequent fate and transport
investigation if the mean extraction efficiency from unpainted stainless steel coupons was 65%
to 120% with a coefficient of variance between triplicates of less than 30%. The extraction
method was also required to be compatible with the FS paint film coupons.
The Low Volatility Agent Permeation (LVAP) setup (described in Section 2.5) was used to
assess the retention of CWA in the SPE disks and to assess the potential for propagation of CWA
out of the SPE disks into the latex gasket of the LVAP. Briefly, five LVAP cells for each CWA
were used in this testing. The SPE disks were spiked with a known amount of CWA mass (using
dilute solutions). For VX, the SPE disks were spiked with 8.4 |ig of VX, 10 |iL of an 840
|ig/milliliter (mL) solution. For HD, the SPE disks were spiked with 51.3 |ig of HD, 10 |iL of a
7
-------
5,126 |ig/mL solution. These spike amounts and concentrations created relatively high CWA
mass loads without putting too much solvent on the SPE disks. An unspiked FS latex flat paint
film coupon was placed on top of the latex gasket/SPE disk. After 72 hours for VX (the longest
VX dwell time) or 48 hours for HD (the longest HD dwell time), the SPE disks were extracted
and analyzed for CWA. The HD spiked coupons were held for shorter dwell times because HD is
more volatile than VX. In addition to the spiked SPE disks, the unspiked latex gaskets and
unspiked FS latex flat paint film coupons were also extracted and analyzed to evaluate the
potential propagation of CWA from the SPE disks.
2.3.3 Sealed Surfaces - Wipe Sampling of VX and HD from Coupons
Based on the method development and demonstration work with painted surfaces, only limited
demonstration testing was conducted with sealed coupons using the same wipe and extraction
methods selected for use with painted coupons. Testing was conducted with stainless steel sealed
with epoxy and stainless steel sealed with polyurethane. Three test coupons (per type of sealant
and type of CWA) were spiked with 2 |iL of neat VX or HD as described in Section 2.6 and held
for a 60-min dwell time. The coupons were then wipe-sampled using the selected methods for
painted surfaces, and the wipes were extracted and analyzed as described in Sections 2.7 and 2.8,
respectively. One procedural blank (an unspiked coupon handled similarly to the test coupons)
and one evaporation control (a spiked coupon handled similarly to the test coupons but directly
extracted for CWA rather than being wipe-sampled) were included per sealant and CWA type).
The outcome of the wipe sampling method development (in short, the use of the gauze wetted
with half saturating amounts of hexane; see Section 3.2.1) was used for collection of both CWAs
from painted and sealed surfaces during the fate and transport studies.
2.4 Test Matrices
The test matrices are described separately for painted surfaces and sealed surfaces. In reality, the
testing associated with both types of coatings is similar. Presenting the methods and results
associated with these coatings is not intended to imply conceptual differences but simply reflects
the fact that the testing associated with this project was conducted in two phases with painted
surfaces tested before testing with the sealed surfaces. Presenting the painted surfaces and sealed
surfaces separately simply allows relatively minor differences in the tests to be documented
clearly.
2.4.1 Painted Surfaces
The test matrix for the fate and transport investigation associated with painted surfaces is shown
in Table 2, including the number of test samples and dwell times. Seven types of coupons
(described in Section 2.5.1) were required (denoted as columns in Table 2). The wipes and
extractions were completed within ฑ10 min of the target dwell time. The laboratory blanks were
comprised of each of the seven coupon types plus the SPE disk. Only unpainted stainless steel
and FS paint film coupons were used for procedural blanks because of the limited value of
replicating the controls with the same paints on stainless steel.
The approach to this investigation was as follows:
8
-------
On each day of testing, 12 coupons of each material type (except SPE disks, which were
placed underneath the FS paint film coupons) were spiked with either 2 |iL of neat VX or
HD as a single droplet as described in Section 2.6. In addition, 2 |iL of VX or HD were
spiked directly into extraction solvent (three replicates) at time zero to serve as spike
controls. The spike controls were spread evenly throughout the time zero spiking
operation, i.e., one spike control prior to spiking samples, one at the midpoint, and the
last one following the spiking of all samples.
Unpainted and painted stainless steel coupons were placed into 60 mm x 15 mm
polystyrene Petri dishes (AS4052, Fisher Scientific, Pittsburgh, PA) during testing and
covered. FS paint film coupons were tested using LVAP cells (as described in Section
2.5).
Three coupons of each coupon type (except SPE disks) were wipe-sampled at four dwell
times shown in Table 2. The wipes were extracted as described in Section 2.7.
Following wiping, the coupons (unpainted stainless steel, painted stainless steel, or FS
paint films) were extracted in solvent as described in Section 2.7. SPE disks underneath
the FS paint film coupons were solvent-extracted as well.
Extracts were analyzed for VX or HD using GC/MS operated in full scan mode as
described in Section 2.8.
In addition to the test coupons, procedural blank coupons (a coupon in the agent hood
that is not spiked with CWA) were extracted and analyzed along with the test coupons.
For each CWA fate and transport investigation, one procedural blank was used for
unpainted stainless steel and the three types of FS paint films. Procedural blanks were
tested using the same approach as the associated spiked test coupons (that is, FS
procedural blank coupons were run in LVAP cells with SPE disks underneath the
coupons). The procedural blanks were extracted and analyzed after the longest dwell
times (72 hours for VX and 48 hours for HD). Since HD is more volatile than VX, the
HD spiked coupons were held for a shorter dwell time than the VX spiked coupons.
One laboratory blank coupon (a coupon kept outside the agent hood that was not spiked
with CWA) of each material type (unpainted stainless steel, the three types of painted
stainless steel, the three types of FS paint films, and SPE disk) was extracted and
analyzed on the day that test coupons were spiked with CWA.
9
-------
Table 2. Painted Surfaces Test Matrix for Fate and Transport Investigation
Agent,
and
Spike
Amount
Dwell
Unpainted
Painted Stainless Steel
FS Paint Film and SPE
Total
Time
(hours)
Method
Stainless
Steel
Latex
flat
Latex
semi gloss
Oil
gloss
Latex
flat
Latex
semi gloss
Oil
gloss
Test
Samples
3
Wipe
3
3
3
3
3
3
3
21
Extraction
3
3
3
3
3+3
3+3
3+3
30
6
Wipe
3
3
3
3
3
3
3
21
VX,
Extraction
3
3
3
3
3+3
3+3
3+3
30
2 nL
24
Wipe
3
3
3
3
3
3
3
21
Extraction
3
3
3
3
3+3
3+3
3+3
30
72
Wipe
3
3
3
3
3
3
3
21
Extraction
3
3
3
3
3+3
3+3
3+3
30
3
Wipe
3
3
3
3
3
3
3
21
Extraction
3
3
3
3
3+3
3+3
3+3
30
6
Wipe
3
3
3
3
3
3
3
21
HD,
Extraction
3
3
3
3
3+3
3+3
3+3
30
2 nL
24
Wipe
3
3
3
3
3
3
3
21
Extraction
3
3
3
3
3+3
3+3
3+3
30
48
Wipe
3
3
3
3
3
3
3
21
Extraction
3
3
3
3
3+3
3+3
3+3
30
3+3 = extraction of agent from FS paint film (3) and underlying SPE disk (3).
t Spike controls (CWA spiked directly into the extraction solvent), procedural blanks (unspiked coupons handled
similarly to test coupons including being placed inside the testing hood), and laboratory blanks (unspiked coupons
that are maintained outside the testing hood) were also included in the testing.
2.4.2 Sealed Surfaces
The test matrix for the fate and transport investigation associated with sealed surfaces is shown
in Table 3, including the number of test samples and dwell times. Five types of coupons
(described in Section 2.5.2) were required (denoted by columns in Table 3). The wipes and
extractions were completed within ฑ10 min of the target dwell time. The laboratory blanks were
comprised of each of the five coupon types plus the SPE disk. Only FS sealants were used for
procedural blanks because of the limited value of replicating the controls with the same sealants
on stainless steel. Fate and transport tests associated with sealed surfaces were limited to three
dwell times with an emphasis on the longer dwell times. The 3 h dwell time was not included in
the sealed surfaces fate and transport tests.
The approach to this investigation was as follows:
On each day of testing, nine coupons of each material type (except SPE disks, which
were placed underneath the FS sealant coupons) were spiked with either 2 |iL of neat VX
or HD as a single droplet as described in Section 2.6. Three additional sealed stainless
steel coupons per coating were spiked with the same amount of VX or HD and wipe-
sampled after one hour (i.e., this is the wipe method demonstration for sealed surfaces).
For logistical reasons, these were spiked and wipe-sampled on Day 2 of the testing. These
two sets of sealed stainless steel coupons served to demonstrate efficacy of the wipe
methods developed for painted surfaces. In addition, 2 |iL of VX or HD were spiked
10
-------
directly into extraction solvent (three replicates; at start, middle and end of coupon
spiking) at time zero to serve as spike controls.
As with the investigation of painted surfaces, unsealed and sealed stainless steel coupons
were placed into Petri dishes during testing. FS sealants were tested using LVAP cells (as
described in Section 2.5).
The LVAP method used for FS paint films was also used for FS sealants.
Three coupons of each coupon type (except SPE disks) were wipe-sampled at three dwell
times shown in Table 3. The wipes were extracted as described in Section 2.7.
Following wiping, the coupons (sealed stainless steel and FS sealants) were extracted in
solvent as described in Section 2.7. SPE disks underneath the FS sealant coupons were
solvent extracted as well. The unsealed stainless steel coupons were not extracted as these
were only wipe-sampled based on results for unpainted stainless steel coupons that were
part of the paint fate and transport tests.
Extracts were analyzed for VX or HD using GC/MS operated in full scan mode as
described in Section 2.8.
In addition to the test coupons, a single procedural blank coupon (a coupon in the agent
hood that was not spiked with the CWAs) and a single laboratory blank coupon (a
coupon kept outside the agent hood that was not spiked with the CWAs) of each material
type (unsealed stainless steel, epoxy sealed stainless steel, polyurethane sealed stainless
steel, FS epoxy sealant, and FS polyurethane sealant) were extracted and analyzed.
Procedural blanks were tested using the same approach as the associated spiked test
coupons (that is, FS procedural blank coupons were run in LVAP cells with SPE disks
underneath the coupons).
The laboratory blanks were extracted on the spike day, and the procedural blanks were
extracted after the longest dwell time (72 hours).
Table 3. Sealed Surfaces Test Matrix for Fate and Transport Investigation
Agent,
and
Dwell
Unsealed
Sealed Stainless Steel
FS Sealant and SPE
Total
Time
(hours)
Method
Stainless
Steel
Test
Samplest
Spike
Amount
Epoxy
Polyurethane
Epoxy
Polyurethane
6
Wipe
3
3
3
3
3
15
Extraction
3
3
3+3
3+3
18
VX,
24
Wipe
3
3
3
3
3
15
2 nL
Extraction
3
3
3+3
3+3
18
72
Wipe
3
3
3
3
3
15
Extraction
3
3
3+3
3+3
18
6
Wipe
3
3
3
3
3
15
Extraction
3
3
3+3
3+3
18
HD,
24
Wipe
3
3
3
3
3
15
2 nL
Extraction
3
3
3+3
3+3
18
48
Wipe
3
3
3
3
3
15
Extraction
--
3
3
3+3
3+3
18
3+3 = extraction of agent from FS sealant (3) and underlying SPE disk (3).
t Spike controls (CWAs spiked directly into the extraction solvent), procedural blanks (unspiked coupons handled
similarly to test coupons including being placed inside the testing hood), and laboratory blanks (unspiked coupons
that are maintained outside the testing hood) were also included in the testing.
11
-------
2.5 Test Materials
The method development and demonstration and the fate and transport investigation associated
with painted surfaces were conducted with the following materials: unpainted stainless steel,
latex flat painted stainless steel, latex semi-gloss painted stainless steel, oil gloss painted
stainless steel, FS latex flat paint film, FS latex semi-gloss paint film, FS oil gloss paint film, and
SPE disk.
For testing with sealed surfaces, unsealed stainless steel, stainless steel sealed with epoxy,
stainless steel sealed with polyurethane, FS epoxy sealant, FS polyurethane sealant, and SPE
disks were used. Additional details on the test materials used in this study are provided in the
following subsections.
2.5.1 Unpainted and Unsealed Stainless Steel
For testing with painted surfaces, the stainless steel (24 gauge 304 stainless steel) was procured
as a custom part from Adept Products Inc., West Jefferson, OH. For testing with sealed surfaces,
a similar stainless steel (304 stainless steel, 0.024 inches thick, 3368T321) was obtained from
McMaster-Carr, Aurora, OH.
Unpainted and unsealed stainless steel coupons were cut to a uniform length (4.0 cm) and width
(2.5 cm) from larger pieces of material. These coupon dimensions enabled the coupons to fit
inside the 125 mL bottle that was used for the extraction. Unpainted and unsealed stainless steel
coupons were cleaned with dry air to remove dust prior to use in the tests.
2.5.2 Painted and Sealed Stainless Steel
Stainless steel sheets were cleaned to ensure good paint and sealant adhesion by washing with
Dawnฎ dishwashing liquid (Procter and Gamble Co., Cincinnati, OH) and water, rinsing with
deionized water, and wiping with a Kimwipe wipe (Kimberly-Clark, Dallas, TX) saturated
with isopropanol. The stainless steel was then air dried and painted with latex flat, latex semi-
gloss, and oil gloss paint, or sealed with a polyurethane or epoxy coating. See Table 4 for details.
12
-------
Table 4. Paint and Sealant Information
Paint Name
Type /Manufacturer
Part Number
Vendor
Latex Flat
Belirฎ Premium Plus Ultra Pure White
105001
Home Depot,
Flat Zero VOC Interior Paint
Reynoldsburg, OH
Latex Semi-Gloss
Belirฎ Premium Plus Ultra Pure White
Semi-Gloss Zero VOC Interior Paint
305001
Home Depot,
Reynoldsburg, OH
Oil Gloss
Rust-Oleumฎ Professional High
Performance White Gloss Oil-Based
35475
Lowe's, Columbus,
OH
Enamel Interior/Exterior Paint
Sealant Name
Polyurethane
coating
Rust-Oleumฎ 6711 System Water-
Based Polyurethane, Clear
4MG61
Grainger, Chicago, IL
Epoxy coating
Rust-Oleumฎ 5300 System Water-
Based Epoxy in White, Gloss Finish
6A414
Grainger, Chicago, IL
Epoxy coating
activator
Rust-Oleumฎ 5300 System Epoxy
Coating Activator
6A403
Grainger, Chicago, IL
All paints and polyurethane coating were applied from the container after mixing for 10 min on a
twin arm paint shaker. The epoxy coating preparation required the mixing of the base material
and epoxy activator in a 7;1 volume ratio as per the manufacturer's instructions. After addition of
the activator, the coating was mixed for 1 min at 3,000 rpm to combine and degas the mixture.
The mixture was allowed to sit for the 30 min induction period prior to applying on the
substrates.
Paints and sealants were applied to the stainless steel substrate using ASTM D823 "Standard
Practices for Producing Films of Uniform Thickness of Paint, Varnish, and Related Products on
Test Panels". Practice E (Hand-Held Blade Film Application) of ASTM D823 is a standard
method for application of film coatings of uniform thickness on a flat panel using a doctor blade.
More specifically for this project, the Universal Blade Applicator (AP-G08, Paul N. Gardner
Company, Pompano Beach, FL) was used. This blade was fixed between two side support plates
and adjustable to control vertical clearance. A flat panel was secured on a firm horizontal
surface. An ample amount of coating was poured across one end of the panel, and the doctor
blade was placed behind the coating. The blade was then drawn uniformly along the length of the
panel toward the operator to apply a uniform film. The dry paint thickness obtained was
dependent on the combination of the bar height used, the volume of solids in the coating, and the
speed of the drawdown motion.
Paint and sealant thickness on stainless steel was measured with an eddy current gauge
(PosiTectorฎ 6000, DeFelsko Corporation, Ogdensburg, NY) per ASTM E376. The target wet
application thickness for these samples was 0.25 millimeters (mm) to 0.51 mm (10 mils to 20
mils; one mil equals one thousandth of an inch). The painted and sealed stainless steel coupons
were then allowed to dry/cure for a minimum of 14 days in a room at constant temperature (24
degrees Celsius [ฐC]) and 50% relative humidity (RH) prior to testing, a standard procedure for
air dry coatings.
The painted and sealed stainless steel sheets were cut into 4.0 cm x 2.5 cm coupons. These
coupons were then cleaned using dry air to remove dust and debris prior to use in tests.
13
-------
2.5.3 FS Paint Films and FS Sealants
FS paint films and FS sealants were made using the same with latex flat, latex semi-gloss, and oil
gloss paint, or sealed with a polyurethane or epoxy coating.
FS paint and sealant sheets were prepared using ASTM D823 "Standard Practices for Producing
Films of Uniform Thickness of Paint, Varnish, and Related Products on Test Panels" as
described in Section 2.5.2, except instead of stainless steel serving as the panel substrate,
silicone-coated paper was used for FS paint films and lethicin-coated glass was used for FS
sealants. The lethicin-coated glass was prepared by diluting lethicin in chloroform to 5% by
weight and applying to plate glass.
Thickness of the FS paint film and FS sealant sheets was measured as per ASTM E376 described
earlier for the paint film thickness measurements. The target wet application thickness for these
samples was 0.25 millimeters (mm) to 0.51 mm (10 mils to 20 mils; a mil equals a thousandth of
an inch). The sheets of FS paint films and FS sealants were then allowed to dry/cure for a
minimum of 14 days in a room at constant temperature (24 ฐC) and 50% RH prior to testing.
After curing, the FS paint film and FS sealant sheets could be handled manually, and coupons
were cut from the larger sheets using a 5 cm diameter die. The coupons were then cleaned using
dry air to remove dust and debris prior to testing.
2.5.4 SPE Disks and LVAP Method
The 3M Empore SDB-XC SPE disk (47 mm diameter, 14-386-4, VWR, Radnor, PA), a
poly(styrenedivinylbenzene) copolymer used as a reversed phase sorbent for SPE, was used to
represent a porous material for this testing.
The LVAP method was applied to evaluate fate and transport of the CWA into a porous surface.
The LVAP method utilized SPE disks to capture CWA as it permeated through an FS paint film.
During testing, CWA was spiked onto the surface of 5 cm diameter FS paint film coupons
(prepared as described above). As CWA permeated through the FS paint film, it was absorbed
and retained by an underlying SPE disk.
LVAP testing was accomplished using a specific layered test cell assembly. For safety and
stability of the test cell setup, the LVAP assembly was placed inside an inverted 240 mL glass jar
with a polytetrafluoroethylene (PTFE)-lined cap (02-992-706, Fisher Scientific, Pittsburgh, PA).
The jar cap was inverted and the bottom portion of a 60 mm diameter x 15 mm polystyrene Petri
dish (FB0875713A, Fisher Scientific, Pittsburgh, PA) was placed inside the lid. A 5 cm diameter
PTFE disk (60179, U.S. Plastics, Lima, OH) was placed inside the Petri dish. A 47 mm diameter
SPE disk was die cut to a diameter of 36 mm to provide a 10 cm2 contact area beneath the FS
paint film coupon and centered on top of the PTFE disk. A latex gasket (NC0224577, Fisher
Scientific, Pittsburgh, PA, die cut to 36 mm inside diameter, 51 mm outside diameter) was
placed around the SPE disk. The FS paint film coupon was then placed directly on top of the
SPE disk/latex gasket and held in close contact using a steel washer (1.375 inch inside
diameter/1.875 inch outside diameter/0.125 inch thickness type 316 stainless steel washer
[97022A852, McMaster-Carr, Aurora, OH]). The exposed surface area of the FS paint film
14
-------
inside the steel washer was 9.58 cm2. The steel washer worked in conjunction with the latex
gasket around the SPE disk to isolate the outside edge of the SPE disk to prevent fugitive CWA
vapor from reaching it, thereby preventing a false positive result. CWA was spiked onto the FS
paint film coupon according to Section 2.6. A cylindrical stainless steel weight (custom
fabrication, type 316 stainless steel) was placed on top of the steel washer to compress the
assembly and ensure adequate contact between the FS paint film coupon and the underlying SPE
disk. The thickness of the steel washer prevented the steel weight from making contact with the
spiked CWA. Figure 1 provides a diagram of the LVAP test cell assembly. Figure 2 provides
photographs of the LVAP test cell assembly and a description of each cell assembly step.
Inverted
15
-------
r
\
10
Step 1: LVAP test cell jar. Lid is inverted and placed on hood floor.
Step 2: Petri dish placed inside test cell jar lid.
Step 3: PTFE disk placed inside Petri dish.
Step 4: Latex gasket placed on top of PTFE disk.
Step 5: SPE disk placed inside latex gasket.
Step 6: FS paint film coupon placed on top of latex gasket/SPE disk.
Step 7: Stainless steel washer placed on top of FS paint film coupon.
Step 8: CWA spiked in center of FS paint film coupon.
Step 9: Weight placed on top of stainless steel washer to compress assembly.
Step 10: Test cell jar inverted and screwed into lid.
Figure 2. Low Volatility Agent Permeation (LVAP) test cell assembly, stepwise.
16
-------
2.6 Chemical Agent, Coupon Spiking, and Dwell Conditions
Much of the VX and HD used for this study was obtained from EPA-owned agents, purchased
from the U.S. Department of Defense (DoD). For VX used with sealed surfaces, VX was
synthesized under the Chemical Weapons Convention program guidelines, with accountability
through the United States Army Edgewood Chemical Biological Center. Agent purity was
measured during this project by GC/flame ionization detector (FID). The agent (source), analysis
date, and purity were:
HD (from DoD) analyzed on 5/21/2015: 96.1%
VX (from DoD) analyzed on 5/21/2015: 84.2%
VX (from DoD) analyzed on 5/27/2015: 79.1%
VX (from DoD) analyzed on 7/22/2015: 69.9%
HD (from DoD) analyzed on 12/3/2015: 96.4%
VX (synthesized) analyzed on 12/3/2015: 96.2%
The test coupons were inspected visually prior to spiking with VX or HD; any coupons with
surface anomalies were not used. The VX or HD was applied as a 2 |iL single droplet using a
Hamilton repeating dispenser (PB600-1, Hamilton, Reno, NV) and a 100 |iL Hamilton syringe
(81085, Hamilton, Reno, NV).
Following application of CWA during fate and transport investigations, the contaminated
coupons were held for 3 hours to 72 hours. During these dwell times, the coupons were subjected
to the ambient laboratory conditions within the chemical agent hood (i.e., temperature and RH of
the coupon environment were not controlled) but either covered with a 60 mm diameter Petri
dish or contained within an LVAP cell to protect from air currents.
2.7 Extraction of VX and HD
After the specified dwell times, test coupons (unpainted/unsealed stainless steel, painted stainless
steel, sealed stainless steel, FS paint films, and FS sealants) were wipe-sampled. The wipes and
coupons (after being wipe-sampled) were transferred separately into bottles of solvent for
extraction. SPE disks were not wipe-sampled but were also transferred into separate bottles of
solvent for extraction. This process was repeated for each time point in the test matrices. All
coupons in the test matrices and the blank coupons were extracted by placing each into a separate
125 mL glass bottle (05-719-57, Fisher Scientific, Pittsburgh, PA) that contained 25 mL of
extraction solvent containing internal standard (IS; 2.5 |ag/mL naphthalene-dx [AC 17496-0010,
Fisher Scientific, Pittsburgh, PA]; solvent with IS was prepared in four liter batches prior to
filling individual extraction bottles to ensure a consistent IS concentration in each sample).
Hexane was the extraction solvent used for all materials, except for SPE disks, which were
extracted using acetone similar to the approach used by Malloy (2012). All extractions took
place in 25 mL of solvent (hexane or acetone). A stabilizer (N,N'-diisopropylcarbodiimide, DIC)
was added to the extraction solvent at 1% by volume to help improve the sensitivity of the
GC/MS analysis of VX samples. Addition of a stabilizer to prevent VX degradation was
recommended by the authors of a stability study for ultra-dilute CWA standards (EPA 2013).
DIC is one of the stabilizers that is used in nerve agent munitions. The unintended benefit of the
stabilizer was an improved sensitivity and calibration linearity.
17
-------
The following approach was used prior to testing to verify that using the DIC stabilizer in solvent
was acceptable:
1) Three clean solvent (hexane) samples (25 mL each) containing IS and DIC (at the
concentrations identified in the previous paragraph) were post-spiked with 20 |iL of a
dilute VX and HD solution (625 |ag/m L each in hexane) such that the final solution
concentration was 0.5 |ag/m L for each CWA, nominally.
2) Three paint coupon extracts were prepared by extracting FS latex flat paint films in 25
mL of hexane and sonicating at 40-60 kilohertz (kHz) for 10 min to create an extract
matrix representative of the test samples. The hexane included IS and DIC as described
in the previous paragraph, and the extracts were post-spiked (following sonication).
3) Three samples were prepared as in 1) above, but the DIC stabilizer was excluded.
4) Three samples were prepared as in 2) above, but the DIC stabilizer was excluded.
The above extracts were then analyzed by GC/MS (calibration standards used during analysis
also included the DIC stabilizer). The GC/MS results were then used to evaluate whether the
DIC was providing increased VX analysis sensitivity (and, conversely, no impact to HD
analysis) and whether the extracted coupon matrix had an effect on the effectiveness and/or
repeatability of any sensitivity increase observed. Results are discussed in Section 3.3. Based on
the results, DIC stabilizer was included in the solvent (hexane and acetone) used to extract all
samples (coupons and SPE disks).
The stainless steel coupons (unpainted/unsealed or painted/sealed with dimensions of 2.5 cm x 4
cm) and the round FS paint film and FS sealant (5 cm diameter) fit lying flat within the inside
diameter of the bottles. When liquid was added to the bottles, 25 mL reached a height of
approximately 1.3 cm, which was sufficient to submerge all coupon types fully. The bottles were
swirled by hand for approximately 5-10 seconds and placed into a sonicator. The extraction
bottles were sonicated at 40-60 kHz for 10 min. Sonication of samples was performed
continuously for each set (i.e., samples were spiked on a two-min stagger, thus during extraction,
jars were added to/removed from the sonicator bath every two minutes, according to the
corresponding spiking and dwell times. The temperature of the sonicator bath was monitored and
allowed to cool if a temperature increase of more than 10 ฐC occurred. Within 30 min of
sonication, approximately 1 mL from each extraction bottle was transferred to individual GC
vials (vial P/N 24670, Fisher Scientific [Restek Corp.], Hanover Park, IL) and sealed. Samples
not analyzed the same day were stored at -20 ฐC or lower.
2.8 Analytical Methods for VX and HD
The sample extracts were analyzed to quantify the amount of VX or HD remaining on each
coupon using GC/MS (6890 gas chromatograph and 5973 mass selective detector, Agilent
Technologies, Santa Clara, CA) operated in the full scan mode for compounds ranging from 40
to 500 Daltons. VX was detected with ions 114, 72, 127, and 79. HD was detected with ions 158,
109, 160, and 111. In most cases, ion 114 was used for VX quantification and ion 109 was used
for HD quantification. Occasionally, ions other than 114 for VX or 109 for HD were used if
interferents were present that prevented the use of these ions. The GC/MS parameters are
documented in Table 5 for VX and Table 6 for HD.
18
-------
Table 5. Gas Chromatography/Mass Spectrometry Conditions for VX Analysis
Parameter
Description
Instrument
Hewlett Packard Model HP 6890 Gas Chromatograph equipped with HP 5973 A
Mass Selective Detector and Model 7683 Automatic Sampler
Data system
Mass Selective Detector ChemStation
Column
Rtxฎ-5MS (cross-linked methylsilicone), 30 meters x 0.25 mm, 0.25 micrometer
(|im) film thickness (Restek Catalogue Number 13623)
Liner type
4 mm Split/Splitless
Carrier gas flow rate
1.2 mL/min
Column temperature
50 ฐC initial temperature, hold 1 min, 30 ฐC/min to 280 ฐC, hold 0 min
Injection volume
1.0 nL
Injection temperature
250 ฐC
MS quad temperature
150 ฐC
MS source temperature
230 ฐC
Solvent delay
3 min
Table 6. Gas Chromatography/Mass Spectrometry Conditions for HD Analysis
Parameter
Description
Instrument
Hewlett Packard Model HP 6890 Gas Chromatograph equipped with HP 5973 A
Mass Selective Detector and Model 7683 Automatic Sampler
Data system
Mass Selective Detector ChemStation
Column
Rtxฎ-5MS, 30 meters x 0.25 mm, 0.25 |im film thickness
Liner type
4 mm Split/Splitless
Carrier gas flow rate
1.2 mL/min
Column temperature
40 ฐC initial temperature, hold 2 min, 30 ฐC/minto 325 ฐC, hold 3 min
Injection volume
2.0 nL
Injection temperature
250 ฐC
MS quad temperature
150 ฐC
MS source temperature
230 ฐC
Solvent delay
3 min
The lowest standard used to establish the calibration curve was above, but near, the instrument
detection limit of the GC/MS as determined in method demonstration (GC/MS instrument
detection limit is 0.02 |ig/mL). Samples with results below the lower calibration level (i.e., 0.10
|ig/mL) were reported as less than the quantitation limit.
On a very limited basis (five unspiked latex gasket samples and five unspiked FS latex flat paint
film coupons associated with method development testing to determine the potential ability of
19
-------
VX to propagate out of SPE disks, described in Section 3.3), samples were analyzed via liquid
chromatography/MS (Shimadzu 20 XR HPLC with AB Sciex Triple Quad 5500 mass
spectrometer). Due to the small mass of VX spiked onto the SPE disk during this method
development testing, the GC/MS minimum detection limit (0.1 (ag/m L) could only detect VX if
at least 26% of the mass spiked was present on the latex gasket or FS latex flat paint film. To
improve the sensitivity of this analysis, liquid chromatography/MS was used which had a
detection limit of 0.01 |ig/mL.
The GC/MS analysis result for each extract sample was fitted to the calibration curve generated
for the specific GC used to analyze the sample, and CWA concentration was determined from
the ratio of the CWA peak area response to the IS peak area response. Use of an IS compensated
for potential variability in sample injection volumes as well as decreasing or increasing
instrument sensitivity. During the effort, CWA exhibited a quadratic response over the
concentration range analyzed, and thus the ratio of CWA area response to the IS area response
(y-axis) was plotted versus the ratio of CWA concentration to IS concentration (x-axis). The
quadratic CWA calibration curve that was fit to the analysis data took the following form:
(As / Ais) = [a x (Cs / Cis)2] + [b x (Cs / Cis)] + c (1)
where:
As = area response of the target analyte
Ais = area response of the internal standard
Cs = concentration of the target analyte
Cis = concentration of the internal standard
a, b, c = coefficients of quadratic curve fit.
CWA concentrations in the extracts from wipes, coupons, and SPE disks were calculated by
GC/MS instrument software and provided in units of ng/mL. These GC concentration results
(Hg/mL) were converted to total mass recovered by multiplying by the extraction solvent
volume:
Mm = Cs x Ev (2)
where:
Mm = calculated mass of CWA recovered from an individual replicate (ng)
Cs = GC concentration (ng/mL) from an individual replicate, see Equation 1
Ev = extraction solvent volume (mL).
The percent recovery of CWA recovered from an individual replicate relative to the mean mass
measured in the spike controls was calculated as follows:
%R = Mm / Msc x 100% (3)
where:
%R = percent recovery for an individual replicate
20
-------
Msc = mean calculated mass of CWA recovered from spike controls (Mg).
A separate %R calculation was made for recovery of CWA from each replicate coupon. The
mean %R was based on the mean of the %R for all applicable replicates.
The percent recovery of CWA from spike controls (versus theoretical) was calculated as follows:
%Rsc = [Mm / (D / CFi x Sv x CF2 x P)] x 100% (4)
where:
%Rsc = percent recovery for an individual spike control replicate (versus theoretical)
D = density of CWA (gram/cubic centimeter [cm3])
CFi = conversion factor 1 (1000 |j,L/cm3)
Sv = CWA spike volume (|iL)
CF2 = conversion factor 2 (1,000,000 gram/|j,g)
P = CWA purity (as a fraction)
A separate %Rsc calculation was made for recovery of CWA from each spike control replicate.
The mean %R was based on the mean of the %Rsc for all applicable replicates.
21
-------
3.0 Test Results
3.1 Painted and Sealed Surfaces - Thickness
Table 7 presents a summary of the dry paint and dry sealant film thickness measured, as well as
the number of painted/sealed panels prepared (before coupon cutting), number of thickness
measurements taken, and number of usable coupons prepared. Paint/sealant thickness on
stainless steel and FS paint films was measured with the eddy current gauge.
All dry film measurements were within ฑ0.2 mil for the each particular paint and coupon
combination. The mean thickness of latex flat films (3.4 mils) and latex semi-gloss films (2.4
mils) was the same whether painted on stainless steel or prepared as FS paint films. For oil gloss
films, the mean dry paint thickness was thinner (2.1 mils) when applied on stainless steel than
when prepared as a FS paint film (2.5 mils).
All dry sealant measurements were within ฑ0.2 mil for the each particular sealant and coupon
combination. However, the mean dry sealant thicknesses were thinner (3.3 mils for epoxy and
1.5 mils for polyurethane) when applied to stainless steel than when prepared as a FS sealant (4.3
mils for epoxy and 2.2 mils for polyurethane).
Table 7. Dry Paint and Sealant Film Thicknesses
Paint
Coupon Type
Dry Paint Film
Thickness (mils)
Painted
Panels
Thickness
Measurements
Coupons
Prepared
Minimum
Mean
Maximum
Prepared
Taken
Latex Flat
Painted Stainless Steel
3.3
3.4
3.5
1
10
49
FS Paint Film
3.3
3.4
3.6
9
90
91
Latex
Painted Stainless Steel
2.3
2.4
2.5
1
10
56
Semi-
Gloss
FS Paint Film
2.3
2.4
2.5
5
50
90
Oil Gloss
Painted Stainless Steel
1.9
2.1
2.2
1
10
54
FS Paint Film
2.4
2.5
2.7
4
40
108
Epoxy
Sealed Stainless Steel
3.2
3.3
3.4
1
10
70
FS Sealant
4.2
4.3
4.5
5
25
51
Polyureth
Sealed Stainless Steel
1.4
1.5
1.6
1
10
70
ane
FS Sealant
2.0
2.2
2.3
10
50
55
3.2 Method Development Results
3.2.1 Painted Surfaces - Wipe Method Development and Demonstration Results
Initial wipe sampling method development for VX and HD on unpainted stainless steel was
conducted using two types of wipes (gauze and DRC), three solvents (hexane, acetone, and IP A),
and two solvent volumes (saturating and half-saturating). The volume of solvent required to
saturate the wipes was approximately 3.0 mL for gauze and 5.6 mL for DRC. The VX and HD
recoveries associated with the DRC and acetone tests did not appear to result in anomalous
recoveries relative to the other wipe combinations used, although the mean VX recovery (95%)
22
-------
using DRC and acetone at a half-saturating volume resulted in a slightly higher mean VX
recovery than any of the other wipe combinations tried (Table 8).
The VX and HD percent recoveries (relative to spike control recoveries) are shown in Table 8
and Figure 3 for all wipe combinations (wipe type, solvent, and solvent volume). Table 8 also
presents the amount (|ig) of VX and HD recovered. The mean spike control recoveries ranged
from 94% to 114% for VX and ranged from 102% to 107% for HD. The spike control recoveries
were calculated based on the amount and purity of the CWA spiked. The purity of VX was
79. P/o, and the purity of HD was 96.1% for this testing. Most of the wipe combinations
recovered >80% of the VX and HD (relative to spike control recoveries). The use of IP A at
saturating volumes for both the gauze and DRC wipes resulted in the lowest mean VX and HD
recoveries (55% to 70%). Based on these results, two wipe sampling combinations "gauze,
hexane, half saturating" and "DRC, acetone, half saturating" were selected for further testing
with FS paint films. For each of these two wipe combinations, the mean VX and mean HD
recoveries were >86% and the standard deviations were <4.0%.
Selected wipe sampling methods were then demonstrated with unpainted stainless steel and three
types of FS paint films (latex flat, latex semi-gloss, and oil gloss). Two wipe combinations were
tested as noted from above ("gauze, hexane, half saturating" for VX and HD, and "DRC,
acetone, half saturating" for VX). The mean spike control recoveries for "gauze, hexane, half
saturating" were 106% for VX and 102% for HD, and the mean spike control recovery for
"DRC, acetone, half saturating was 81% for VX. The spike control recoveries were calculated
based on the amount and purity of the CWA spiked. The purity of VX was 84.2%, and the purity
of HD was 96.1%) for these tests.
The VX and HD recoveries (relative to spike controls) for the wipe sampling method
demonstration are presented in Table 9 and Figure 4. Recoveries were lower from the FS paint
films than the unpainted stainless steel. The mean VX recoveries were higher using the DRC
wipes with acetone (75% to 112%) than using the gauze wipes with hexane (60% to 89%). The
HD recoveries using the gauze and hexane ranged from 55% to 98%. Given the relatively similar
recoveries for both VX and HD, as well as solvent incompatibilities observed between acetone
and the latex flat paint, the "gauze, hexane, half saturating" wipe sampling method was selected
for collection of both CWAs during the fate and transport studies.
23
-------
Table 8. VX and HD Recoveries from Unpainted Stainless Steel for Various Wipe Combinations (Method Development)
Wipe Combination (Wipe Type, Solvent Volume, Solvent Type)
CWA
Measure of
Recovery
Gauze
Saturating
Gauze
Half Saturating
DRC
Saturating
DRC
Half Saturating
Hexane
Acetone
I PA
Hexane
Acetone
I PA
Hexane
Acetone
I PA
Hexane
Acetone
I PA
Mean (|ig)
1.4xl03
1.6xl03
1.3xl03
1.4xl03
1.5xl03
1.6xl03
1.2xl03
1.3xl03
l.OxlO3
1.2xl03
1.4xl03
1.5xl03
VX
SD (|ig)
79
133
14
15
18
84
33
66
130
51
17
147
% Recovery
83
88
70
86
84
85
75
89
58
77
95
89
SD (%)
4.7
7.6
0.76
0.88
1.0
4.6
2.1
4.4
7.6
3.3
1.2
8.6
Mean (|ig)
2.5xl03
2.3 xlO3
1.7xl03
2.4xl03
2.3xl03
2.4xl03
2.4xl03
2.3xl03
1.4xl03
2.3xl03
2.3xl03
2.0xl03
HD
SD (|xg)
3.7
58
606
99
104
69
114
130
111
40
68
298
% Recovery
101
92
68
96
93
95
94
92
55
91
94
78
SD (%)
0.15
2.3
23
4.0
4.2
2.7
4.5
5.2
4.3
1.6
2.8
11
SD = standard deviation; n=3.
Notes: all procedural blanks were non-detect for VX and HD (<3.1 |ig): for the evaporation controls, the percent recoveries were >82% for VX and >91% for HD.
24
-------
Wipe Combination (Wipe Type, Solvent, Solvent Volume)
Figure 3. VX and HD recoveries from unpainted stainless steel for various wipe
combinations (method development). Error bars equal plus one standard deviation.
Table 9. VX and HD Wipe Recoveries from Unpainted Stainless Steel and FS Paint Films
(Method Demonstration)
CWA and Wipe
Material
Combination (Wipe
Type, Solvent, and
Solvent Volume)
Measure of
Recovery
Unpainted
Stainless Steel
FS Paint Film:
Latex Flat
FS Paint Film:
Latex Semi Gloss
FS Paint Film:
Oil Gloss
VX - Gauze,
Hexane, Half
Saturating
Mean (|ig)
1.6xl03
1.3xl03
l.lxlO3
1.4xl03
SD (|ig)
80
68
33
87
% Recovery
89
74
60
76
SD (%)
4.4
3.8
1.8
4.8
Mean (|ig)
1.5xl03
1.3xl03
l.OxlO3
1.4xl03
VX - DRC, Acetone,
SD (|ig)
152
121
174
175
Half Saturating
% Recovery
112
96
75
99
SD (%)
11
8.8
13
13
HD - Gauze,
Hexane, Half
Saturating
Mean (|ig)
2.4xl03
1.4xl03
1.4xl03
2.0xl03
SD (|ig)
49
42
456
170
% Recovery
98
55
55
81
SD (%)
2.0
1.7
18
6.8
SD = standard deviation; n=3
Notes: all procedural blanks were non-detect for VX and HD (<2.8 |ig): for the evaporation controls, the percent
recoveries were >84% for VX and >93% for HD.
25
-------
120%
100%
^ c
S o
o u
u 01
s ฃ
QC Q.
l/>
80%
60%
OJ (11
a; %
40%
ai
20%
0%
IVX - Gauze, Hexane, Half Saturating
IVX- DRC, Acetone, Half Saturating
HD - Gauze, Hexane, Half Saturating
Unpainted
Stainless Steel
FS Paint Film:
Latex Flat
FS Paint Film:
Latex Semi-Gloss
FS Paint Film:
Oil Gloss
Material
Figure 4. VX and HD wipe recoveries from unpainted stainless steel and FS paint films
(method demonstration). Error bars equal plus one standard deviation.
3.2.2 Painted Surfaces - Extraction Method Development Results
The extraction efficiency of VX and HD from unpainted stainless steel, FS paint films, and the
SPE disk was tested by sonicating these spiked materials in 25 mL of solvent for 10 min. Hexane
was used as the extraction solvent for all materials except the SPE disks, which were extracted
using acetone. The VX and HD recoveries are presented in Table 10 and Figure 5. The mean
spike control recoveries were 103% for both VX and HD. The spike control recoveries were
calculated based on the amount and purity of the CWA spiked. The purity of VX was 84.2%, and
the purity of HD was 96.1% for these tests. For all material and VX and HD combinations, the
mean percent recoveries (relative to the spike controls) were >85% with standard deviations
<5.5%.
As described in Section 2.3.2, the ability of the SPE disk to retain VX and HD in the LVAP
setup was evaluated. After 72 hours, the mean recovery of VX (7.1 |ig) from the SPE disk was
75% with a 9.6% standard deviation relative to spike controls. After 48 hours, the mean recovery
of HD (49 |ig) from SPE disks was 89% with a 2.8% standard deviation relative to spike
controls. The propagation of CWA out of the SPE disk and into either the latex gasket of the
LVAP (described in Section 2.5) or the unspiked FS latex flat paint film coupons placed on top
was also assessed and neither VX (<0.25 |ig by LC/MS) nor HD (<2.5 |ig) was detected in these
samples. Based on the mass of VX (8.4 |ig) and HD (51.3 |ig) spiked onto the SPE disks, <3.0%
of the applied VX and <4.9% of the applied HD propagated out of the SPE disk into the latex
gasket or the unspiked FS latex paint films.
26
-------
Table 10. VX and HD Extraction Recoveries from Unpainted Stainless Steel, FS Paint
Films and SPE Disk
Material
CWA
Measure of
Recovery
Unpainted
FS Paint Film:
FS Paint Film:
FS Paint Film:
SPE Disk
Stainless Steel
Latex Flat
Latex Semi Gloss
Oil Gloss
Mean (|ig)
1.6xl03
1.6xl03
1.6xl03
1.5xl03
1.7xl03
VX
SD (|xg)
34
94
49
91
34
% Recovery
93
90
94
85
96
SD (%)
1.9
5.4
2.8
5.2
2.0
Mean (|ig)
2.5xl03
2.2xl03
2.4xl03
2.5xl03
2.4xl03
HD
SD (|xg)
20
106
31
109
139
% Recovery
99
89
96
98
97
SD (%)
0.79
4.2
1.2
4.3
5.5
SD = standard deviation; n=3.
Notes: all laboratory blanks were non-detect for VX and HD (<2.5 |ig).
100%
o
ฃ ฃ
2 o
o u
u
-------
with or without DIC was then spiked with VX and HD. The mean recoveries were 83% for VX
and 88% for HD when DIC was included and 97% for VX and 94% for HD when DIC was not
included. Results are summarized in Table 11. Given mean recoveries >82% for all of these VX
and HD tests, DIC was used in the extraction solvents of all other tests associated with
painted/sealed surfaces.
Table 11. DIC Stabilizer Evaluation Results
Post Spiked Extract
(jig/mL)
(jig/mL)
Hexane with IS
No
Yes
0.55
0.56
0.54
97%
0.48
86%
Hexane with IS
No
Yes
0.55
0.56
0.51
93%
0.45
80%
Hexane with IS
No
Yes
0.55
0.56
0.49
90%
0.44
79%
Hexane with IS
Flat Latex
Yes
0.55
0.56
0.49
89%
0.47
84%
Hexane with IS
Flat Latex
Yes
0.55
0.56
0.48
87%
0.46
83%
Hexane with IS
Flat Latex
Yes
0.55
0.56
0.48
87%
0.46
82%
Hexane with IS
No
No
0.55
0.56
0.50
90%
0.29
52%
Hexane with IS
No
No
0.55
0.56
0.49
89%
0.29
51%
Hexane with IS
No
No
0.55
0.56
0.50
90%
0.29
52%
Hexane with IS
Flat Latex
No
0.55
0.56
0.52
95%
0.55
97%
Hexane with IS
Flat Latex
No
0.55
0.56
0.52
95%
0.54
97%
Hexane with IS
Flat Latex
No
0.55
0.56
0.51
93%
0.54
96%
Hexane with IS
No
Yes
unspiked
unspiked
<0.10
NA
<0.10
NA
Hexane with IS
Flat Latex
Yes
unspiked
unspiked
<0.10
NA
<0.10
NA
Hexane with IS
No
No
unspiked
unspiked
<0.10
NA
<0.10
NA
Hexane with IS
Flat Latex
No
unspiked
unspiked
<0.10
NA
<0.10
NA
* Based on measured spike solution concentrations of HD and VX and a 20 |iL spike volume.
NA - not applicable.
3.2.3 Sealed Surfaces - Wipe Method Demonstration Results
Wipe sampling method demonstration for VX and HD on epoxy- and polyurethane-sealed
stainless steel was conducted using the wipe methods selected for painted surfaces (i.e., gauze
wipes wetted with hexane at half the saturating volume, 1.5 mL). The wipe extraction approach
was the same as used for painted surfaces (i.e., sonication in 25 mL of hexane with IS and DIC
for 10 min).
The VX and HD percent recoveries (relative to spike control recoveries) are shown in Table 12
and Figure 6. Table 12 also presents the amount (|ig) of VX and HD recovered. The mean spike
control recoveries were 108% for VX and 110% for HD. The spike control recoveries were
calculated based on the amount and purity of the CWA spiked. The purity of VX was 96.2%, and
the purity of HD was 96.4% for this testing. The recoveries were lower from polyurethane-sealed
stainless steel (60% to 75%) than from epoxy-sealed stainless steel (82% to 89%).
28
-------
Table 12. VX and HD Wipe Recoveries from Sealed Stainless Steel
CWA
Measure of
Recovery
Material
Sealed Stainless Steel:
Epoxy
Sealed Stainless Steel:
Polyurethane
VX
Mean (|ig)
1.9xl03
1.3xl03
SD (|xg)
62
117
% Recovery
89
60
SD (%)
3.0
5.6
HD
Mean (|ig)
2.2xl03
2.0xl03
SD (|xg)
91
43
% Recovery
82
75
SD (%)
3.4
1.6
SD = standard deviation; n=3.
Notes: all procedural blanks were non-detect for VX and HD (<2.6 |ig): for the evaporation controls, the percent
recoveries were >88% for VX and >86% for HD.
100%
Sealed Stainless Steel: Sealed Stainless Steel:
Epoxy Polyurethane
Material
Figure 6. VX and HD extraction recoveries from sealed stainless steel. Error bars equal
plus one standard deviation.
29
-------
3.3 Fate and Transport Results
3.3.1 Painted Surfaces - VX
Immediately after spiking, VX tended to form a bead on the coupons. In some cases (unpainted
stainless steel, FS oil gloss paint films, stainless steel painted with oil gloss, and FS latex semi-
gloss paint films), the beads were flattened and had a "pancake" appearance. The VX caused
blistering of the FS paint films and painted stainless steel coupons. In some cases (FS latex flat
paint films, FS latex semi-gloss paint films, and stainless steel painted with latex flat or latex
semi-gloss), holes were left in the paint surfaces after wipe sampling (holes were not apparent
prior to wiping). An example FS paint film with hole is shown in Figure 7. Stainless steel painted
with oil gloss was described as having an orange-colored stain after wiping. On unpainted
stainless steel, VX was spread out on the coupon, and a streak remained after wipe sampling.
Figure 7. Photograph of a FS paint film (lifted) showing hole in the center of the coupon
after wiping.
Mean VX mass measurements recovered from wipes, coupons (extracted after wipe-sampling),
and SPE disks at each dwell time are presented in Table 13 and Figures 8-11. The purity of VX
used was 69.9%, and the mean spike control recovery for VX for these tests was 1470 (.ig with a
standard deviation of 50 jig. The total VX recoveries (based on the cumulative recovery of VX
from the wipes, coupons, and SPE disks) consistently decreased over the four dwell times for the
materials tested. Total VX recoveries (from all materials including unpainted stainless steel)
ranged from 63% to 83% after 3 hours and ranged from 23% to 46% after 72 hours. Decreasing
VX recoveries were also generally observed over time based on the wipe sampling results,
30
-------
although for the FS latex semi-gloss paint film and SPE disk, a lower mean amount of VX (279
|ig) was recovered after 3 hours than after 6 hours (561 |ig).
Table 13. Painted Surfaces - VX Recoveries from Wipes, Coupons, and SPE Disks ^
Dwell
Wipe
Coupon (after wipe)
SPE Disl
V
Total VX Recovery
Material
Time
Mean
SD
FOD
Mean
SD
FOD
Mean
SD
FOD
Mean
SD
FOD
(hour)
(tig)
(tig)
(tig)
(tig)
(tig)
fog)
(%)
(%)
Unpainted
Stainless
Steel
3
1187
6.2
3/3
38
2.3
3/3
83
0.57
6/6
6
1090
132
3/3
50
18
3/3
78
10
6/6
24
843
22
3/3
52
50
3/3
61
3.7
6/6
72
503
72
3/3
117
18
3/3
42
6.1
6/6
Painted
Stainless
Steel: Latex
Flat
3
1035
47
3/3
141
10
3/3
80
2.6
6/6
6
958
87
3/3
120
16
3/3
73
5.2
6/6
24
617
123
3/3
215
40
3/3
57
7.1
6/6
72
274
56
3/3
153
47
3/3
29
6.8
6/6
Painted
Stainless
Steel: Latex
Semi-Gloss
3
526
64
3/3
421
19
3/3
64
5.2
6/6
6
352
15
3/3
505
66
3/3
58
3.7
6/6
24
243
79
3/3
431
10
3/3
46
5.9
6/6
72
280
69
3/3
190
16
3/3
32
5.7
6/6
Painted
Stainless
Steel: Oil
Gloss
3
928
35
3/3
136
4.1
3/3
72
2.6
6/6
6
712
14
3/3
164
3.5
3/3
60
0.74
6/6
24
385
79
3/3
121
24
3/3
34
5.3
6/6
72
266
8.0
3/3
65
65
3/3
23
4.2
6/6
FS Paint
Film: Latex
Flat and SPE
3
895
55
3/3
146
3.4
3/3
95
42
3/3
77
1.6
9/9
6
733
75
3/3
225
83
3/3
70
61
3/3
70
1.5
9/9
24
442
224
3/3
379
230
3/3
119
39
3/3
64
2.8
9/9
72
189
36
3/3
169
27
3/3
283
107
3/3
44
3.9
9/9
FS Paint
Film: Latex
Semi-Gloss
and SPE
3
279
42
3/3
727
60
3/3
27
14
3/3
70
1.3
9/9
6
561
277
3/3
363
253
3/3
83
25
3/3
68
1.1
9/9
24
260*
35*
3/3*
592
60
3/3
52
17
2/2*
60
2.2
8/8*
72
204
17
3/3
338
25
3/3
134
24
3/3
46
1.8
9/9
FS Paint
Film: Oil
Gloss and
SPE
3
761
38
3/3
166
19
3/3
<2.5
0
0/3
63
2.8
6/9
6
420
152
3/3
331
19
3/3
<2.5
0
0/3
51
10
6/9
24
223
79
3/3
261
82
3/3
25
19
3/3
35
12
9/9
72
241
86
3/3
69
27
3/3
23
14
3/3
23
6.5
9/9
-- = not applicable.
SD = standard deviation.
< = all replicate results were less than the quantitation limit.
FOD = frequency of detection (number of samples above the quantitation limit/total number of samples).
For results less the quantitation limit, the quantitation limit (i.e., a 2.5 |ig residual mass) was used for the calculation
of summary statistics.
t The mean VX spike amount (per spike control recovery) was 1470 |ig.
* For one of the wipe replicates, an SPE disk was inadvertently included in the extraction; thus, one wipe replicate
result included the recovery from the wipe and SPE disk. The associated SPE disk results were based only on two
replicates.
Notes: all laboratory blanks and procedural blanks were non-detect for VX (<2.6 |ig).
31
-------
two
TS
Q)
o
u
(D
o
u
-------
0
3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72
Dwell Time (h)
Figure 10. Mean VX mass recovered from FS latex semi-gloss paint films and underlying
SPE disks (wipe, coupon [after wipe], and SPE disk samples).
-a
ai
*_
a>
>
o
u
a>
cc
x
>
ro
a>
3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72
Dwell Time (h)
Figure 11. Mean VX mass recovered from FS oil gloss paint films and underlying SPE
disks (wipe, coupon [after wipe], and SPE disk samples).
VX was recovered from all of the coupon extractions, but consistent trends of increasing or
decreasing mean recovery mass among coupons were not apparent in Figure 8 (dashed trend
lines) across the materials tested. As seen in Table 13, mean mass of VX recovered from the
coupons after the wipe (38 |ig to 215 |ig) were low compared to the wipe samples (266 |ig to
1187 |ig) for unpainted stainless steel and for stainless steel painted with either latex flat or oil
gloss. For the other four test materials in Table 13 (stainless steel painted with latex semi-gloss,
and all three types of FS paint film coupons), somewhat higher amounts of VX were recovered
from the coupons after the wipe (69 |ig to 727 |ig) with all of these materials having at least one
dwell time with VX recoveries >331 |ig. The VX recoveries from wipes of these four test
materials (solid trend lines in Figures 8-11) ranged from 189 |ig to 895 |ig. The VX recoveries
from the coupon (paint) extracts associated with the FS paint films and SPE disks appeared to
33
-------
increase and then decrease during the 72-hour dwell time (as seen in the trend lines within
Figures 9-11).
Testing with FS paint films and SPE disks (last three sets of rows in Table 13) demonstrated the
following:
Measurable levels of VX mass were diffused into the underlying SPE disks in all tests
conducted, except when using FS oil gloss paint films after 3 and 6 hours (when masses
recovered on each of the SPE disks were below the quantitation limit of 2.5 |ig, equivalent to
<0.17% of the VX applied to the FS oil gloss paint film). At each tested dwell time, the FS oil
gloss paint films had the lowest permeability of VX mass, while the FS latex flat paint films had
the highest measured permeability among all tests at all but the 6-hour dwell time. Thus, a clear
separation in VX permeability measurements existed between the three FS paint film types at the
3-, 24-, and 72-hour dwell times. In addition, the VX recoveries from the SPE disks were
generally higher following the 24- and 72-hour dwell times compared to the 3- and 6-hour dwell
times.
When statistical test #2 from Section 2.2.1 was performed on the VX mass measurements from
the underlying SPE disks at each dwell time, and upon accounting for the different levels of
measurement variability which occur between the three FS paint films, significant differences in
mean mass measurement among the FS paint film types were observed (at the 0.05 level) at the
3-, 24-, and 72-hour dwell times. Specific outcomes from these tests are as follows:
Although there is no information on variability in the FS oil gloss paint film test
measurements at the 3-hour dwell time (as all three measurements were below
quantitation limits), there is some statistical evidence that mean VX mass recovered from
the SPE disk associated with the FS latex flat paint film (95 |ig) differs significantly from
the mean VX mass for each of the other two paint films (27 |ig for latex semi-gloss and
2.5 |ig or less for oil semi-gloss) at the 0.05 level at the 3-hour dwell time.
At both the 24- and 72-hour dwell times, significant differences were observed in mean
VX mass measurements from the SPE disks associated with the FS latex flat paint film
(119 |ig at 24 hours and 283 |ig at 72 hours) and oil gloss paint film (25 |ig at 24 hours
and 23 |ig at 72 hours) at the 0.05 level. In addition, at the 72-hour dwell time, mean VX
mass from the SPE disks associated with the FS latex semi-gloss (134 |ig) and the FS oil
gloss (23 |ig) paint film types differed significantly at the 0.05 level when a two-sample
t-test was performed on data for only those two film types.
When statistical test #1 from Section 2.2.1 was performed on the three VX mass measurements
from the underlying SPE disks at each dwell time, the mean VX mass measurement was
significantly higher than the quantitation limit at the 0.05 level in all cases but the following:
Oil gloss paint at the 24- and 72-hour dwell times
Latex flat paint at the 6-hour dwell time
Latex semi-gloss paint at the 24-hour dwell time (only two measurements were available
at this time point)
This test could not be performed on data for oil gloss at the 3- and 6-hour dwell times, as all data
were reported as being below the quantitation limit.
34
-------
3.3.2 Painted Surfaces - HP
Immediately after spiking, HD tended to form a bead on the coupons, except for unpainted
stainless steel, FS oil gloss paint films, and stainless steel painted with oil gloss, which had a
flattened or "pancaked" appearance. After three hours and six hours, the painted stainless steel
and FS paint films were described as being blistered where the HD was applied, and holes were
created in the FS latex flat paint film coupons during the wipe sampling at these dwell times.
After 24 hours to 48 hours, the painted stainless steel and FS paint film coupons were generally
described as having a dark spot or raised spot where the HD had been applied. On unpainted
stainless steel, the HD retained the "pancaked" appearance until 48 hours when the appearance
was described as a dry stain.
Mean HD mass measurements recovered from wipes, coupons (coupon extractions following
wipe sampling), and SPE disks at each dwell time are presented in Table 14 and Figures 12-15.
The purity of HD used was 96.1%, and the mean spike control recovery for HD for these tests
was 2492 |ig with a standard deviation of 49 |ig. The total HD recoveries (based on the
cumulative recovery of HD from the wipes, coupons, and SPE disks) remained relatively high
(>72%) over the four dwell times for the materials tested, except for unpainted stainless steel,
which decreased to 41% recovery after 24 hours and 0.23% recovery after 48 hours, and oil gloss
painted stainless steel, which decreased to 47% recovery after 24 hours and 25% recovery after
48 hours.
Based on wipe sampling only, HD recoveries decreased from the three-hour dwell time (432 |ig
to 2261 |ig) to the 48-hour dwell time (3.0 |ig to 51 |ig). Nearly all of the HD recovered from
unpainted stainless steel was recovered via wipes. Only a limited amount of HD (<22 |ig) was
recovered from the coupon extraction of unpainted stainless steel. Interestingly, coupon and SPE
disk extraction recovered much more HD when used on the other samples (painted stainless steel
and FS paint films and SPE disks). Coupon extraction had mean recoveries of HD ranging from
317 |ig to 2223 |ig, and SPE disk extraction had mean recoveries of 234 |ig to 2006 |ig for the
painted stainless steel and FS paint films and SPE disks. These data indicate that HD is retained
within paints to a greater extent than captured by wipe sampling. Similar to the trends observed
with VX, HD recoveries from the coupon (paint) extracts associated with the FS paint films
generally appeared to increase and then decrease during the 48-hour dwell time (Figures 12-15).
35
-------
Table 14. Painted Surfaces - HD Recoveries from Wipes, Coupons, and SPE Disks
Dwell
Wipe
Coupon (after wipe)
SPE Disk
Total HD Recovery
Material
Time
(hour)
Mean
(tig)
SD FOD
(tig)
Mean SD
. . . . FOD
(tig) (tig)
Mean SD
. . . . FOD
(tig) (tig)
SD (%) FOD
Unpainted
Stainless
Steel
3
2261
48
3/3
22
24
3/3
--
--
--
92
0.97
6/6
6
2052
141
3/3
4.8
0.14
3/3
--
--
83
5.7
6/6
24
1001
131
3/3
10
7.2
3/3
--
--
41
5.5
6/6
48
3.0
0.35
2/3
2.6
0.17
1/3
--
--
0.23
0.0078
3/6
Painted
Stainless
Steel: Latex
Flat
3
997
293
3/3
1282
167
3/3
--
--
91
5.5
6/6
6
573
297
3/3
1598
185
3/3
--
--
87
8.7
6/6
24
92
30
3/3
1892
150
3/3
--
--
80
7.1
6/6
48
43
10
3/3
1900
88
3/3
--
--
78
3.2
6/6
Painted
Stainless
Steel: Latex
Semi-Gloss
3
838
106
3/3
1552
130
3/3
--
--
96
1.3
6/6
6
334
101
3/3
2093
140
3/3
--
--
97
1.6
6/6
24
78
11
3/3
2223
125
3/3
--
--
92
5.0
6/6
48
51
14
3/3
2130
98
3/3
--
--
88
4.0
6/6
Painted
Stainless
Steel: Oil
Gloss
3
1684
176
3/3
447
28
3/3
--
--
86
6.3
6/6
6
1680
24
3/3
619
23
3/3
--
--
92
0.062
6/6
24
199
12
3/3
982
43
3/3
--
--
47
2.0
6/6
48
5.8
1.7
3/3
622
181
3/3
--
--
25
7.3
6/6
FS Paint
Film: Latex
Flat and SPE
3
755
193
3/3
761
9.0
3/3
877
261
3/3
96
3.8
9/9
6
582
85
3/3
988
147
3/3
814
93
3/3
96
3.6
9/9
24
18
4.6
3/3
827
78
3/3
1523
133
3/3
95
7.2
9/9
48
9.7
0.92
3/3
618
84
3/3
1708
81
3/3
94
5.1
9/9
FS Paint
Film: Latex
Semi-Gloss
and SPE
3
432
88
3/3
1546
73
3/3
381*
185
3/3
95
1.5
9/9
6
194
31
3/3
1745
84
3/3
542
1.9
3/3
100
4.4
9/9
24
14
3.3
3/3
683
12
3/3
1669
24
3/3
95
1.1
9/9
48
5.1
0.88
3/3
317
69
3/3
2006
22
3/3
93
3.4
9/9
FS Paint
Film: Oil
Gloss and
SPE
3
1555
29
3/3
692
32
3/3
234
10
3/3
100
0.42
9/9
6
898
97
3/3
889
28
3/3
488*
1.2
3/3
91
5.1
9/9
24
6.5
1.6
3/3
677
43
3/3
1199
30
3/3
76
2.4
9/9
48
3.2
0.65
3/3
485
32
3/3
1317
37
3/3
72
2.2
9/9
-- = not applicable.
SD = standard deviation.
FOD = frequency of detection (number of samples above the quantitation limit/total number of samples).
For results less the quantitation limit, the quantitation limit (i.e., 2.6 |ig and 2.5 |ig residual mass for wipe and
coupon samples, respectively) was used for the calculation of summary statistics,
t The mean HD spike amount (per spike control recovery) was 2492 |ig.
* For one of the replicates, the weight was inadvertently left off the LVAP cell assembly for a portion of the dwell
time. The stainless steel washer was present the entire dwell time.
Notes: all laboratory blanks and procedural blanks were non-detect for HD (<2.6 |ig).
36
-------
I
TS
0)
i_
0)
>
o
u
*_
a>
>
o
u
a>
cc
a
x
ro
a>
2200
2000
1800
1600
1400
1200
1000
wipe
coupon
SPE
12 15 18
33 36 39 42 45 48
21 24 27 30
Dwell Time (h)
Figure 13. Mean HD mass recovered from FS latex flat paint films and underlying SPE disks (wipe, coupon [after wipe], and SPE
disk samples).
37
-------
0
3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48
Dwell Time (h)
Figure 14. Mean HD mass recovered from FS latex semi-gloss paint films and underlying
SPE disks (wipe, coupon [after wipe], and SPE disk samples).
3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48
Dwell Time (h)
Figure 15. Mean HD mass recovered from FS oil gloss paint films and underlying SPE
disks (wipe, coupon [after wipe], and SPE disk samples).
The HD is clearly transported into the SPE disks beneath the FS paint films (as seen in the last
three sets of rows in Table 14). The amounts of HD recovered from SPE disks were higher after
24- and 48-hour dwell times (mean HD recoveries ranged from 1,199 |ig to 2,006 |ig), while
mean HD recoveries after three- and six-hour dwell times ranged from 234 |ig to 877 |ig. Key
results of applying statistical test #2 of Section 2.2.1 to these data were as follows:
At the three-hour dwell time, the SPE disk associated with the FS latex flat paint film
generally had higher HD mass values compared to the other two film types, while the
SPE disk associated with FS oil gloss paint film had the lowest mass values. However,
38
-------
when accounting for differences present in variability, no differences in HD recovery
from the SPE disks were statistically significant at the 0.05 level.
At the six-hour dwell time, the FS latex flat paint film (814 |ig) differed significantly
from the FS latex semi-gloss (542 |ig) and FS oil gloss (488 |ig) paint films at the 0.05
level in mean HD mass measured on the SPE disks. While results of a two-sample t-test
also indicated that the HD recovered from the SPE disks associated with FS latex semi-
gloss and FS oil gloss paint films differed significantly at a 0.05 level, this outcome was
due primarily to low variability in the observed measurements, allowing for small
differences to be statistically significant.
For HD recoveries from the SPE disks at the 24-hour dwell time, the FS oil gloss paint
film (1,199 |ig) differed significantly from the FS latex flat paint film (1,523 |ig) and the
FS latex semi-gloss paint film (1,669 |ig) at the 0.05 level.
At the 48-hour dwell time, each of the three FS paint films had HD recoveries from the
SPE disk that differed significantly from each other at a 0.05 level. The FS latex semi-
gloss paint film had the highest mean HD mass on the SPE disks (2,006 |ig), while the FS
oil gloss paint film had the lowest (1,317 |ig).
When statistical test #1 from Section 2.2.1 was performed on the three HD mass
measurements from the underlying SPE disks at each dwell time, the mean HD mass
measurement was significantly higher than the quantitation limit in all cases at the 0.05
level.
3.3.3 Sealed Surfaces - VX
After spiking, VX generally had a flattened bead or "pancaked" appearance. Beading was
initially observed on some individual replicates of FS polyurethane and unsealed stainless steel.
During the dwell times, VX became spread out on the coupons, except on the FS polyurethane
sealant, which was noted as having raised blisters where the VX was applied. This led to creation
of holes in the FS polyurethane sealant after wiping.
Mean VX mass measurements recovered from wipes, coupons (coupon extractions following
wipe sampling), and SPE disks at each dwell time are presented in Table 15 and Figures 16 and
17. The purity of VX used was 96.2%, and the mean spike control recovery for VX for these
tests was 2095 |ig with a standard deviation of 182 |ig. Unsealed stainless steel coupons were not
extracted following wipe sampling based on observations from the painted stainless steel
surfaces where the maximum contribution of the coupon extract was 6% or less, depending on
the specific dwell time.
The total VX recoveries (based on the cumulative recovery of VX from the wipes, coupons, and
SPE disks) generally decreased over the 72-hour dwell time. On unsealed stainless steel, the
mean total VX recoveries decreased by 10% (89% after six hours and 79% after 72 hours). For
the sealed materials, the reductions in total VX recoveries were more dramatic, ranging from
21% for FS epoxy sealant and SPE (83% after six hours and 62% after 72 hours) to 43% for
stainless steel sealed with epoxy (78% after six hours and 35% after 72 hours).
39
-------
Table 15. Sealed Surfaces - VX Recoveries from Wipes, Coupons, and SPE Disks
Dwell
Wipe
Coupon (after wipe)
SPE Disl
V
Total VX Recovery
Material
Time
(hour)
Mean
(tig)
SD
(tig)
FOD
Mean
(tig)
SD
(tig)
FOD
Mean
(tig)
SD
(^g)
FOD
Mean
(%)
SD
(%)
FOD
Unsealed
Stainless Steel
6
1859
52
3/3
--
--
--
89
2.5
3/3
24
1861
136
3/3
--
--
--
89
6.5
3/3
72
1655
162
3/3
--
--
--
79
7.7
3/3
Sealed
6
1254
34
3/3
374
25
3/3
78
2.0
6/6
Stainless Steel:
24
476
110
3/3
773
147
3/3
60
2.1
6/6
Epoxy
72
185
8.1
3/3
538
9.7
3/3
35
0.6
6/6
Sealed
6
73
14
3/3
1115
35
3/3
57
1.0
6/6
Stainless Steel:
24
338
555
3/3
729
185
3/3
51
18
6/6
Polyurethane
72
5.5
0.24
3/3
459
17
3/3
22
0.8
6/6
FS Sealant:
6
1663
61
3/3
79
24
3/3
<2.5
0.0
0/3
83
3.1
6/9
Epoxy and
24
1212
141
3/3
470
65
3/3
<2.5
0.0
0/3
80
4.1
6/9
SPE
72*
305
63
3/3
986
61
3/3
<2.5
0.0
0/3
62
0.1
6/9
FS Sealant:
6
452
397
3/3
1143
234
3/3
40
65
2/3
78
6.9
8/9
Polyurethane
24
238
189
3/3
1193
143
3/3
220
102
3/3
79
4.3
9/9
and SPE
72
75
97
2/3
775
158
3/3
130
33
3/3
47
11
8/9
-- = not applicable.
SD = standard deviation.
< = all replicate results were less than the quantitation limit.
FOD = frequency of detection (number of samples above the quantitation limit / total number of samples).
For results less than the quantitation limit, the quantitation limit (i.e., 2.6 |ig and 2.5 |ig residual mass for wipe and
SPE samples, respectively) was used for the calculation of summary statistics,
t The mean VX spike amount (per spike control recovery) was 2095 |ig.
* One replicate FS epoxy sealant coupon flipped over when removing from Petri dish. The spiked surface briefly
contacted the SPE.
Notes: all laboratory blanks and procedural blanks were non-detect for VX (<2.6 |ig).
Mean wipe recoveries, after six hours, were relatively high (ranging from 1254 |ig to 1859 |ig)
from unsealed stainless steel, stainless steel sealed with epoxy, and FS epoxy sealant and SPE
compared to the mean wipe recoveries from stainless steel sealed with polyurethane and FS
polyurethane sealant and SPE, which ranged from 73 |ig to 452 |ig after 6 hours. Wipe
recoveries of VX for all materials tested were lower after 72 hours than after 6 hours and 24
hours.
VX was recovered from all material coupons extracted at each dwell time, but consistent trends
of increasing or decreasing VX concentrations were not apparent for the materials tested. For
stainless steel sealed with polyurethane and FS polyurethane sealant and SPE, VX recoveries
were lower after 72 hours than after 6 and 24 hours. For stainless steel sealed with epoxy and FS
epoxy sealant and SPE, VX recoveries after 72 hours were higher than those after 6 hours.
40
-------
two
TS
Q)
o
u
M-
o
to
l/>
(0
OJ
>
o
u
a>
cc
x
>
ro
a>
1200
900
600
300
FS Epoxy Sealant - wipe
FS Epoxy Sealant-coupon
FS Epoxy Sealant-SPE
FS Polyurethane Sealant - wipe
FS Polyurethane Sealant - coupon
FS Polyurethane Sealant - SPE
12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72
Dwell Time (h)
Figure 17. Mean VX mass recovered from FS sealants (epoxy and polyurethane) and underlying SPE disks (wipe, coupon [after
wipe], and SPE disk samples).
41
-------
For VX, measurable mass (exceeding the quantitation limit) was transported to the SPE disks
only beneath the FS polyurethane sealant. All measurements involving the epoxy sealant were
below the quantitation limit of 2.5 |ig. Thus, no statistical tests were performed to compare VX
recoveries from SPE disks between the two types of sealants. The amounts of VX recovered
from SPE disks beneath the FS polyurethane sealant increased from 6 to 24 hours (from 40 to
220 |ig), followed by a decrease at the 72 dwell time (to 130 |ig). These mean mass recovery
values for polyurethane were significantly higher than the quantitation limit (at the 0.05 level)
only after the 24- and 72-hour dwell times.
3.3.4 Sealed Surfaces - HP
After spiking, HD beaded on nearly all the coupons; on three unsealed stainless steel coupons,
the HD was noted as having a flattened bead or "pancaked" appearance. At each dwell time,
blistering occurred on both the FS polyurethane sealant and FS epoxy sealant. For unsealed
stainless steel and stainless steel sealed with epoxy and polyurethane, the HD appeared as a shiny
or dry spot after 48 hours (the longest dwell time).
Mean HD mass measurements recovered from wipes, coupons (coupon extractions following
wipe sampling), and SPE disks at each dwell time are presented in Table 16 and Figures 18 and
19. The purity of HD used was 96.4% and the mean spike control recovery for HD for these tests
was 2702 |ig with a standard deviation of 72 |ig. Unsealed stainless steel coupons were not
extracted following wipe sampling based on observations from the unpainted stainless steel
surfaces where the maximum contribution of the coupon extract was 1% or less of the total (wipe
+ extraction) amount, depending on the specific dwell time.
The total HD recoveries (based on the cumulative recovery of HD from the wipes, coupons, and
SPE disks) generally decreased over the 48 hours of testing, except for testing associated with
the FS polyurethane sealant and SPE. For unsealed stainless steel, the total (mean) HD recoveries
decreased from 68% at 6 hours to <0.1% at 48 hours. Similar results were obtained for HD
applied to stainless steel sealed with epoxy (64% mean HD recovery after 6 hours and 2.8%
recovery after 48 hours). Interestingly, higher HD recoveries were associated with the other
materials tested. For stainless steel sealed with polyurethane, total (mean) HD recoveries were
82% after 6 hours and 48% after 48 hours. For FS epoxy sealant and SPE, total (mean) HD
recoveries were 72% after 6 hours and 31% after 48 hours. The highest total HD recoveries were
associated with FS polyurethane sealant and SPE, which remained relatively high and ranged
from 86% to 99%.
Based on wipe sampling after the 6-hour dwell time, mean HD recoveries were relatively high
from all materials (>1481 |ig, except for the samples associated with the FS polyurethane sealant
and SPE, which was 782 |ig). After 48 hours, the HD recoveries associated with wipe sampling
were reduced to <7.8 |ig.
Coupon extraction had mean recoveries of HD ranging from 222 |ig to 857 |ig after 6 hours.
After 24 hours, the HD recoveries increased or remained relatively high from all coupons
(ranging from 344 |ig to 1539 |ig). After 48 hours, coupon extractions resulted in mean HD
recoveries lower than those recovered after 6 and 24 hours (ranging from 65 |ig to 352 |ig),
42
-------
except for stainless steel sealed with polyurethane, which had a mean HD recovery of 1292 |ig
after 48 hours.
Table 16. Sealed Surfaces - HD Recoveries from Wipes, Coupons, and SPE Disks ^
Dwell
Wipe
Coupon (after wipe)
SPE Disl
V
Total VX Recovery
Material
Time
(hour)
Mean
(tig)
SD
(tig)
FOD
Mean
(tig)
SD
(tig)
FOD
Mean
(tig)
SD
(^g)
FOD
Mean
(%)
SD
(%)
FOD
Unsealed
Stainless Steel
6
1826
91
3/3
--
--
--
68
3.4
3/3
24
827
245
3/3
--
--
--
31
9.1
3/3
48
<2.6
0.0
0/3
--
--
--
<0.1
0.0
0/3
Sealed
6
1506
55
3/3
225
20
3/3
64
1.9
6/6
Stainless Steel:
24
239
148
3/3
344
143
3/3
22
11
6/6
Epoxy
48
3.6
1.7
1/3
73
16
3/3
2.8
0.6
4/6
Sealed
6
1481
88
3/3
726
97
3/3
82
3.2
6/6
Stainless Steel:
24
191
198
3/3
1539
257
3/3
64
16
6/6
Polyurethane
48
7.8
1.7
3/3
1292
146
3/3
48
5.4
6/6
FS Sealant:
6
1711
128
3/3
222
66
3/3
17
26
1/3
72
1.7
7/9
Epoxy and
24*
27
37
3/3
385
93
3/3
558
279
3/3
36
6.5
9/9
SPE
48
<2.6
0.0
0/3
65
16
3/3
772
144
3/3
31
4.8
6/9
FS Sealant:
6
782
369
3/3
857
188
3/3
715
457
3/3
87
3.8
9/9
Polyurethane
24
6.5
3.8
3/3
697
194
3/3
1961
336
3/3
99
6.1
9/9
and SPE
48
2.9
0.4
1/3
352
193
3/3
1969
171
3/3
86
7.8
7/9
-- = not applicable.
SD = standard deviation.
< = all replicate results were less than the quantitation limit.
FOD = frequency of detection (number of samples above the quantitation limit / total number of samples).
For results less the quantitation limit, the quantitation limit (i.e., 2.6 |ig and 2.5 |ig residual mass for wipe and SPE
samples, respectively) was used for the calculation of summary statistics,
t The mean HD spike amount (per spike control recovery) was 2702 |ig.
* One replicate FS epoxy sealant coupon flipped over when being removed from Petri dish. The spiked surface
briefly contacted the SPE.
Notes: all laboratory blanks and procedural blanks were non-detect for HD (<2.6 |ig).
43
-------
DO
T3
ai
ai
>
0
~
two
TS
0)
i_
0)
>
o
u
-------
SPE disk extractions resulted in mean HD recoveries that tended to increase with increasing
dwell times. (For FS epoxy sealant and SPE, the mean HD recoveries from the SPE disks were
17 |ig after six hours and 772 |ig after 48 hours. For FS polyurethane sealant and SPE disk, the
mean HD recoveries from the SPE disks were 715 |ig after six hours and 1,969 |ig after 48
hours.) In addition, the FS polyurethane sealant was associated with higher HD permeability to
the SPE disks compared to the FS epoxy sealant at each dwell time. Key results of applying
statistical test #2 of Section 2.2.2 to these data were as follows:
At the six-hour dwell time, no significant difference was observed at the 0.05 level
between the HD recoveries from SPE disks beneath the two sealant types. However, two
of the three measurements associated with the FS epoxy sealant were below the
quantitation limit of 2.5 |ig (and the third measurement was 47 |ig), while the
measurements associated with the FS polyurethane sealant ranged from 420 jug to 1,214
|ig. Thus, the range of HD measurements from the SPE disk beneath the two sealants was
quite different. The large variability associated with the three FS polyurethane sealant
measurements served as a key cause for non-significance.
At the 24- and 48-hour dwell times, significant difference was observed at the 0.05 level
between the HD measurements in the SPE disk beneath the two sealant types, with FS
polyurethane sealant associated with measurements that were two to three times higher
than the measurements associated with the FS epoxy sealant.
The mean HD mass values recovered from the SPE disks were significantly higher than the
quantitation limit at the 0.05 level for both sealant types at the 24- and 48-hour dwell times, but
not at the six-hour dwell time.
45
-------
4.0 Quality Assurance/Quality Control
4.1 Control of Monitoring and Measuring Devices
The data quality indicators for test measurements are provided in Table 17. All data quality
indicator results were acceptable per the Quality Assurance Project Plan for Fate and Transport
of Chemical Warfare Agents VX and HD Across a Permeable Layer into Porous Subsurfaces,
Version 1 (March 2015), including checks of the measurement methods for time, volume, spike
controls, internal standard (IS), and paint thickness. The results shown in Table 17 are well
within the data quality indicators for all parameters, which limits the amount of error introduced
into the investigation results.
Temperature and RH were monitored during this investigation, but because the investigation
took place under ambient laboratory conditions and no attempts were made to control these
environmental parameters, data quality indicators were not developed for this investigation.
Nevertheless, a calibrated HOBO humidity and temperature data logger (UX-100-003, Onset
Computer Corporation, Bourne, MA) was used to monitor and record the environmental
conditions. The temperature during the fate and transport testing associated with painted surfaces
ranged from 18.7 ฐC to 21.8 ฐC and RH ranged from 51.5% to 69.6%. During the fate and
transport testing with sealed surfaces, the temperature ranged from 18.4 ฐC to 22.6 ฐC and the
RH ranged from 27.8% to 48.9%. Details on environmental conditions are summarized in
Appendix A.
Table 17. Data Quality Indicators and Results
Parameter
Measurement
Method
Data Quality Indicators
Results
Time, seconds
Timer/data
logger
Compared to the National Institute of
Standards and Technology (NIST)
website: http://NIST.time.gov once
before testing; agree ฑ2 seconds/hour.
0 seconds/hour
Thickness of
paint, sealant/
coating
Eddy current
gauge
Thickness will be acceptable if the
observed range of measurements was
within ฑ (0.2 mil) of the target thickness
All dry film measurements
were within ฑ0.2 mil for the
each particular paint or
sealant and coupon
combination.
Volume, |iL
Repeating
dispenser/
syringe
Repeating dispenser/syringe was
checked for accuracy and repeatability
one time before use by determining the
mass of water delivered. The syringe
was considered acceptable if the range
of observed masses for five droplets
was ฑ10% of expected.
The observed masses were
<4.84% of expected for 2 |iL
syringes.
Spike control
GC/MS
The mean of the spike controls included
with each day of testing should be
within 70% to 120% of the target
application and have a coefficient of
variation of <30% between replicates.
The mean spike controls
were within 102% to 110%
of the target application with
coefficients of variation
<8.7% between replicates.
46
-------
Parameter
Measurement
Method
Data Quality Indicators
Results
IS,
naphthalene-d8
Extraction,
GC/MS
The mean of the IS included with each
day of testing should be within 70% to
120% of the expected mass.
The IS peak area response
for the primary samples was
within ฑ50% of the mid-
point calibration standard IS
response.
VX or HD on
unpainted
stainless steel
with extraction
after 60 min,
(ig/inL
Extraction,
GC/MS
The mean percent recovery for a known
quantity of each VX or HD added to
unpainted stainless steel coupons must
be at least 65% but less than 120% of
the spike control and have a coefficient
of variation of <30% between
replicates.
The mean VX and HD
recoveries from unpainted
stainless steel (after a three-
hour dwell time and sampled
via wipe and coupon
extraction) were 83% to
92% relative to the spike
controls with coefficients of
variation <1.1% between
replicates.
CWAon
laboratory blank
coupons, (ig/inL
Extraction,
GC/MS
Laboratory blanks (coupons without
applied CWA, maintained outside the
testing hood) should have <1% of the
amount of analyte compared to that
found on spike controls.
No measurable VX or HD
was detected on laboratory
or procedural blank coupons.
VX or HD on
uncoated
(unsealed)
stainless steel
wipe-sampled
after 60 inin
(ig/inL
Extraction,
GC/MS
The mean percent recovery for a known
quantity of each VX or HD added to
uncoated (unsealed) stainless steel
coupons must be at least 65% but less
than 120% of the spike control and have
a coefficient of variation of <30%
between replicates.
The mean VX and HD
recoveries from unpainted
stainless steel (after a six-
hour dwell time and wipe-
sampled) were 89% and
68%, respectively, relative to
the spike controls with
coefficients of variation
<5.0% between replicates.
VX purity
GC/FID
Purity (percent) must be >85% at the
start of the test.
Spike control recoveries
were calculated based on the
amount and purity of the
CWA spiked.
The DoD VX purities, which
were used for testing with
painted surfaces, ranged
from 69.9% to 96.1%.
The synthesized VX purity
was >96% at the start of the
testing with sealed surfaces.
HD purity
GC/FID
Purity (percent) must be >85% at the
start of the test.
The DoD HD purity was
>96% at the start of testing.
47
-------
4.2 Equipment Calibrations
Section 2.8 noted the instrumentation used to quantify VX and HD. The analytical equipment
was maintained and operated according to the quality requirements and documentation of the
research laboratory. All equipment was either calibrated with appropriate standards or replaced
with a new, manufacturer-calibrated instrument at the frequency specified in Table 18.
Table 18. Equipment Calibration Schedule
Equipment
Frequency
Calibrated pipette and repeating
dispenser/syringe
Prior to the investigation and every six months thereafter
Eddy current gauge
Calibration verified each day of testing by measurement of reference
standards
Calibrated
hygrometer/thermometer
Prior to the investigation and annually thereafter
GC/MS
Beginning of each batch of test samples (calibration curve) and a calibration
verification standard after every five samples and at the end of a batch of
samples. In general, a batch was defined as all extractions associated with
one dwell time for a specific CWA
At least, a five-point calibration for VX or HD was used with a lower calibration level of
approximately 0.1 |ag/m L and an upper range of approximately 125 |ig/mL. A mean response
(relative standard deviation <15%) and linear regression (coefficient of determination [r2]
>0.980) curve fit was applied to the calibration data. Any sample exceeding the upper calibration
limit would have been diluted to a concentration within the calibration range and reanalyzed;
however, this was not necessary as results of all samples analyzed fell within the calibration
range. The GC/MS was tuned initially and as needed following manufacturer's guidelines. A
tune check was performed before running each batch using decafluorotriphenylphosphine. A 12-
hour tune time was not employed.
Following analysis of the calibration standards at the beginning of each analytical run, a solvent
blank sample was analyzed to confirm that no VX or HD carryover was occurring. Solvent blank
sample analysis results were below the value of the lowest calibration standard.
Independently prepared continuing calibration verification (CCV) standards were analyzed prior
to sample analysis, following every five samples, and at the end of each batch of samples. Two
CCV concentrations were used, one of which was equal to the low calibration standard and the
other within the calibration range. The CCV response had to be within 35% of the nominal
concentration for the lowest level CCV used and within 20% of the nominal concentration for all
other CCVs for VX or HD to be acceptable. Samples analyzed prior to or following CCVs that
were outside acceptance limits were re-analyzed (either within the same analytical run or during
a separate run for which a new calibration curve was established; results from reanalysis were
considered valid and reportable if analysis quality control objectives, as described above for
calibration curve standards and bracketing CCVs, were met).
48
-------
Neat VX or HD (concentrations corrected for percent purity) were used to create calibration
standards encompassing the appropriate analysis range. Calibration standards were kept and
used for no longer than six months from the date of creation. CCV standards were kept and
used for no longer than one month from the date of creation. The GC was recalibrated if the r2
from the regression analysis of the standards was <0.98.
Limits were placed on the percent bias observed in the low and high standards. Percent bias is
calculated for both standards as follows:
Percent bias = x 100% (5)
where:
R = expected value from calibration curve
C = observed value from standard
The percent bias for the low standard was required to fall below 25%, and the percent bias for
the remaining standards was required to fall below 15%.
As discussed above, one CCV standard was run for every five samples. The percent bias for the
low CCV standard was required to fall below 35%, and the percent bias for the remaining CCV
standards was required to fall below 20%. Every tenth test sample extract was immediately re-
injected and analyzed following the original. The result from the reanalysis of every tenth sample
was used strictly as an additional confirmation to the analyst of adequate GC performance. The
reanalysis result was required to fall within 20% of the result of the original analysis of the
sample. The tenth sample reanalysis results were included only in the primary raw analysis data
generated by the analyst and excluded from calculations and reported data. Criteria for
evaluation of the GC performance are shown in Table 19.
Table 19. Gas Chromatography Performance Parameters and Acceptance Criteria
Parameter
Acceptance Criterion
coefficient of determination (r2)
>0.98
% bias for the lowest calibration standard
<25%
% bias for remaining calibration standards (except lowest standard)
<15%
Solvent blank sample
< lowest calibration standard
% bias for the lowest CCV
<35%
% bias for remaining CCVs (except lowest CCV)
<20%
Differences between replicate samples (tenth sample reanalysis)
<20%
4.3 Technical Systems A udit
The Quality Assurance (QA) Officer performed a technical systems audit (TSA) during the
investigation (on December 7, 2015). The purpose of the TSA was to ensure that testing was
performed in accordance with this project's quality assurance project plan (QAPP). The QA
Officer reviewed the investigation methods, compared test procedures to those specified in the
49
-------
QAPP (and the associated Amendments 1 and 2), and reviewed data acquisition and handling
procedures. The QA Officer observed the CWA spiking, wiping, and extraction procedures and
found that all procedures were performed according to standard operating procedures and the
associated QAPP. In addition, all documentation followed good laboratory practice guidelines,
and no issues with documentation were noted. All staff adhered to the strict safety requirements
while working with CWA.
4.4 Performance Evaluation A udits
The scope and outcome of the performance evaluation (PE) audit are summarized in Table 20.
Acceptable tolerances were volume (ฑ10%), time (ฑ1 second/min), chemical mass (>85%), film
thickness (ฑ0.1 mil), and IS (ฑ10%). No PE audit was performed for VX or HD as there were no
other sources for a standard. The volume of VX or HD dispensed correlated to the mass on the
coupon. The IS provided confidence that the analysis systems were providing accurate data.
Because temperature and RH were measured on each day of testing but were not controlled, no
PE audit was performed for these parameters.
Table 20. Performance Evaluation Results
Parameter Audit Procedure Expected Tolerance Results
Volume
Syringe used for
dispensing chemical agent
was checked for accuracy
and repeatability one time
before use by determining
the mass of water
delivered
ฑ10%
Syringes <4.84%
Time
Compared time to
independent clock one
time before use
ฑ1 second/min
0 second/min
Chemical mass
Used GC/MS to determine
mass of CWA delivered as
2 |iL droplet into 25 mL
of extraction solvent and
compared to target
application level one time
>85% of spike target
All mean spike controls were
102% to 110% of the
theoretical spike amount.
Film thickness
Measured the thickness of
a standard reference using
the eddy current gauge
ฑ0.1 mil
A calibrated probe accurate to
0.1 mil was used to measure
the thickness.
IS
Used GC/MS to measure
from a secondary source
and compared to the
primary source one time
ฑ10%
1.85%
50
-------
4.5 Data Quality Audit
The QA Officer audited at least 10% of the investigation data and traced the data from initial
acquisition, through reduction and statistical comparisons, to final reporting. All data analysis
calculations were checked.
51
-------
5.0 Summary
5.1 Painted Surfaces
This project investigated the fate and transport of VX and HD applied to painted surfaces and
porous material beneath FS paint films. Method development and demonstration work was also
conducted with regard to wipe sampling methods and extraction methods for the CWAs. For the
fate and transport testing, VX and HD were spiked at 2 |iL onto unpainted stainless steel, painted
stainless steel, and FS paint films placed over SPE disks as representative of a porous material.
The stainless steel and FS paint films were prepared with three types of paint (latex flat, latex
semi-gloss, and oil gloss). After holding at ambient laboratory conditions for 3 hours to 72 hours,
the materials were wipe-sampled for VX or HD, and then the entire material coupon was
extracted and analyzed for VX or HD. For tests with SPE disks beneath FS paint films, the SPE
disks and the FS paint films were extracted and analyzed separately, but only the FS paint films
were wipe-sampled.
Wipe sampling methods were evaluated with a combination of wipes (gauze or DRC), wipe
wetting solvents (hexane, acetone, and IP A), and two solvent volumes (the solvent volume
needed to saturate the wipe or half the volume needed to saturate the wipe). All of these wipe
combinations (except when IPA was used at the saturating volume for both gauze and DRC
wipes) were able to recover >75% of VX and HD (relative to spike controls) from unpainted
stainless steel. Additional testing with gauze wipes with hexane at half saturating volumes for
VX and HD on FS paint films was conducted. VX and HD recoveries from FS paint films were
lower than observed from unpainted stainless steel, but recoveries were >55% relative to spike
controls and considered acceptable to use for the fate and transport testing. Note that these
recovery values are biased low as the agent was in contact with the FS paint film for 60 min
during this method development.
Method demonstration on the extraction efficiency of VX and HD directly from the materials
was tested by sonicating the spiked materials (unpainted stainless steel, the three types of FS
paint film, and SPE disks) in 25 mL of solvent for 10 min. Hexane was used as the extraction
solvent for all materials except the SPE disks, which were extracted using acetone. For all
materials, the mean percent recoveries of VX and HD were >85%. As such, this extraction
approach was used for the fate and transport investigation.
Fate and transport results for VX on painted surfaces are summarized in Table 21. Over the 72-
hour dwell time, VX recoveries generally decreased for the wipe samples and for the total
recoveries (sum of wipe, coupon, and SPE disk). VX was also recovered from most coupon and
SPE disk extractions. With regard to the percent of VX recovered from the coupons, increasing
or decreasing trends were not apparent. Lower percent VX recoveries (<15%) were associated
with the coupons of unpainted stainless steel, latex flat painted stainless steel, and oil gloss
painted stainless steel. The other materials had at least one dwell time with VX recoveries from
coupon extraction of >22%. The VX recoveries from SPE disks generally increased with dwell
time, and VX recoveries were highest (4.8% to 19%) from the SPE disk beneath the latex flat FS
paint film.
52
-------
Table 21. Percent VX Recoveries from the Painted Surfaces Investigation
Dwell
% Recovered (relative to spike controls)
Material
Time
(hour)
Wipe
Coupon
(after wipe)
SPE Disk
Total
3
81
2.6
--
83
Unpainted Stainless
6
74
3.4
--
78
Steel
24
57
3.6
--
61
72
34
8.0
--
42
3
70
9.6
--
80
Painted Stainless Steel:
6
65
8.1
--
73
Latex Flat
24
42
15
--
57
72
19
10
--
29
3
36
29
--
64
Painted Stainless Steel:
6
24
34
--
58
Latex Semi-Gloss
24
17
29
--
46
72
19
13
--
32
3
63
9.3
--
72
Painted Stainless Steel:
6
48
11
--
60
Oil Gloss
24
26
8.2
--
34
72
18
4.4
--
23
3
61
9.9
6.5
77
FS Paint Film: Latex
6
50
15
4.8
70
Flat and SPE
24
30
26
8.1
64
72
13
12
19
44
3
19
49
1.8
70
FS Paint Film: Latex
6
38
25
5.6
68
Semi-Gloss and SPE
24
18*
40
3.6*
60
72
14
23
9.1
46
3
52
11
<0.17
63
FS Paint Film: Oil Gloss
6
29
22
<0.17
51
and SPE
24
15
18
1.7
35
72
16
4.7
1.5
23
-- = not applicable
* For one of the wipe replicates, an SPE disk was inadvertently included in the extraction; thus one wipe replicate
result included the recovery from the wipe and SPE disk. The associated SPE disk results were based only on two
replicates.
For "<" values, the results were less the quantitation limit for all replicates.
Fate and transport results for HD on painted surfaces are summarized in Table 22. Over the 48-
hour dwell time, HD recoveries rapidly decreased for the wipe samples. By 48 hours, all HD
wipe recoveries were <2%. However, total HD recoveries remained relatively high (>72%) at the
48-hour dwell time for most of the materials, except unpainted stainless steel (0.23%) and the oil
gloss painted stainless steel (25%). Relatively high percentages of HD were recovered from the
painted stainless steel and FS paint film coupons (e.g., >18% was recovered after a three-hour
dwell time for all paint-related materials, and up to 89% was recovered from latex semi-gloss
painted stainless steel after a 24-hour dwell time). HD recoveries were also high for the SPE disk
(ranging from 9.4% to 80%). HD recoveries increased from the SPE disk with increasing dwell
times. The increasing HD recoveries from the SPE disk demonstrated that HD may persist in a
53
-------
porous material below a paint layer at higher levels than measured by wipe sampling and/or
coupon extractions.
Table 22. Percent HD Recoveries from the Painted Surfaces Investigation
Material
Dwell
% Recovered (relative to spike controls)
Time
(hour)
Wipe
Coupon
(after wipe)
SPE Disk
Total
Unpainted Stainless
Steel
3
91
0.88
--
92
6
82
0.19
--
83
24
40
0.38
--
41
48
0.12
0.10
--
0.23
Painted Stainless Steel:
Latex Flat
3
40
51
--
91
6
23
64
--
87
24
3.7
76
--
80
48
1.7
76
--
78
Painted Stainless Steel:
Latex Semi-Gloss
3
34
62
--
96
6
13
84
--
97
24
3.1
89
--
92
48
2.0
85
--
88
Painted Stainless Steel:
Oil Gloss
3
68
18
--
86
6
67
25
--
92
24
8.0
39
--
47
48
0.23
25
--
25
FS Paint Film: Latex
Flat and SPE
3
30
31
35
96
6
23
40
33
96
24
0.73
33
61
95
48
0.39
25
69
94
FS Paint Film: Latex
Semi-Gloss and SPE
3
17
62
15
95
6
7.8
70
22
100
24
0.54
27
67
95
48
0.20
13
80
93
FS Paint Film: Oil Gloss
and SPE
3
62
28
9.4
100
6
36
36
20
91
24
0.26
27
48
76
48
0.13
19
53
72
-- = not applicable
Clearly VX and HD are capable of permeating paints and porous materials beneath the FS paint
films. This effect was more dramatic for HD than VX, although the spike amounts differed
between these CWAs. Based on the spike controls, the mean VX spiked amount was 1470 |ig
and the mean HD spike amount was 2492 |ig. HD appears to accumulate to a greater extent
(based on percent recoveries relative to the spike controls) in SPE disks than VX (for the
materials and dwell times investigated). Based on testing with both painted and sealed surfaces,
all VX recoveries from SPE disks were <10%, except for the test with FS latex flat paint, which
had a 19% recovery after 72 hours from the associated SPE disks. For HD, all recoveries from
SPE disks ranged from 20% to 80%, with the exception of three tests (two associated with
painted surfaces and one associated with sealed surfaces). These three instances occurred at the
shortest dwell times, and in each instance the HD recoveries were higher than the VX recoveries
from SPE disks in similar tests.
54
-------
5.2 Sealed Surfaces
The fate and transport of VX and HD was also investigated as applied to sealed surfaces and
porous material beneath FS sealants. As before, VX and HD were spiked at 2 |iL onto unsealed
stainless steel, sealed stainless steel, and FS sealant films placed over SPE disks. The stainless
steel and FS sealants were prepared with two types of sealant (epoxy and polyurethane). After
holding at ambient laboratory conditions for 6 hours to 72 hours, the materials were wipe-
sampled for VX or HD, and then the entire material coupon was extracted and analyzed for VX
or HD. For tests with SPE disks beneath FS sealants, the SPE disks and the FS sealants were
extracted and analyzed separately, but only the FS sealants were wipe-sampled.
Fate and transport results for VX on sealed surfaces are summarized in Table 23. Over the 72-
hour dwell time, VX recoveries generally decreased with time for the wipe samples and for the
total recoveries (sum of wipe, coupon, and SPE disk). VX was also recovered from coupon
extractions (generally ranging from 18% to 57% recoveries). At the 72-hour dwell time the
recoveries of VX were higher from the FS sealants (37% to 47%) than from the sealed stainless
steel (22% to 26%). This might be associated with the thicker sealant films associated with the
FS sealants than those of the sealed stainless steel. For epoxy, the mean dry sealant thickness was
1 mil higher than the thickness of the sealant on the stainless steel (4.2 mil vs. 3.2 mil,
respectively). Similarly, the FS polyurethane sealant had a mean thickness 0.6 mils greater than
the mean thickness associated with the polyurethane stainless steel (2.0 mil vs. 1.4 mil,
respectively).
Coupon extractions of stainless steel sealed with polyurethane and FS polyurethane sealant
consistently resulted in VX recoveries that represented the largest contributions to the total
amount of VX recovered. VX was not recovered from SPE disks underneath FS epoxy sealants,
and low levels (<10%) of VX were recovered from SPE disk beneath FS polyurethane sealants.
55
-------
Table 23. Percent VX Recoveries from the Sealed Surfaces Investigation
Material
Dwell
% Recovered (relative to spike controls)
Time
(hour)
Wipe
Coupon
(after wipe)
SPE Disk
Total
Unsealed Stainless Steel
6
89
89
24
89
89
72
79
79
Sealed Stainless Steel:
Epoxy
6
60
18
78
24
23
37
60
72
8.8
26
35
Sealed Stainless Steel:
Polyurethane
6
3.5
53
57
24
16
35
51
72
0.3
22
22
FS Sealant: Epoxy and
SPE
6
79
3.8
<2.5
83
24
58
22
<2.5
80
72
15
47
<2.5
62
FS Sealant: Polyurethane
and SPE
6
22
55
1.9
78
24
11
57
10
79
72
3.6
37
6.2
47
-- = not applicable
For "<" values, the results were less the quantitation limit for all replicates.
Fate and transport results for HD on sealed surfaces are summarized in Table 24. Over the 48-
hour dwell time, HD recoveries rapidly decreased with time for the wipe samples. By 48 hours,
all HD wipe recoveries were <0.3%. HD was also recovered from the direct extraction of the
coupons. The amounts of HD recovered were generally higher for the coupons associated with
polyurethane, which ranged from 13% to 57%, than from coupons associated with epoxy, which
ranged from 2.4% to 14%. Trends of increasing or decreasing HD recoveries from the coupons
were not consistently demonstrated. HD was recovered from the SPE disk. HD recoveries were
greater for the SPE disk underneath the FS polyurethane sealant (26% to 73% mean HD
recoveries) than for the SPE disk underneath the FS epoxy sealant (0.6% to 29% mean HD
recoveries). As the dwell times increased, the HD recoveries from the SPE disk increased
representing a larger proportion of the total amount of HD recovered. The increasing HD
recoveries from the SPE disk demonstrated that HD can persist at higher levels than measured by
wipe sampling and/or coupon extractions alone.
56
-------
Table 24. Percent HD Recoveries from the Sealed Surfaces Investigation
Material
Dwell
% Recovered (relative to spike controls)
Time
(hour)
Wipe
Coupon
(after wipe)
SPE Disk
Total
Unsealed Stainless Steel
6
68
68
24
31
31
48
<0.1
<0.1
Sealed Stainless Steel:
Epoxy
6
56
8.3
64
24
8.9
13
22
48
0.1
2.7
2.8
Sealed Stainless Steel:
Polyurethane
6
55
27
82
24
7.1
57
64
48
0.3
48
48
FS Sealant: Epoxy and
SPE
6
63
8.2
0.6
72
24
1.0
14
21
36
48
<0.1
2.4
29
31
FS Sealant: Polyurethane
and SPE
6
29
32
26
87
24
0.2
26
73
99
48
0.1
13
73
86
-- = not applicable
For "<" values, the results were less the quantitation limit for all replicates.
The CWAs VX and HD are both capable of permeating into some sealants and porous materials
beneath the FS sealants. In general, both VX and HD appeared to have greater abilities to
permeate and migrate through polyurethane sealant compared to epoxy sealant. VX was not
detected in the SPE disk underneath the FS epoxy sealant, but VX was detected in the SPE disk
underneath the FS polyurethane sealant. HD was detected in the SPE disk underneath both the
FS epoxy sealant and FS polyurethane sealant. As the dwell times increased, wipe sample results
reflected a lower portion of the total CWA recovered based on the summation of wipe sampling,
coupon extraction, and SPE disk extraction.
57
-------
6.0 References
EPA, 2007. A Literature Review of Wipe Sampling Methods for Chemical Warfare Agents and
Toxic Industrial Chemicals. EPA/600/R-11/079. U.S. Environmental Protection Agency (EPA),
Office of Research and Development, Washington, DC.
EPA, 2013. Stability Study for Ultra-Dilute Chemical Warfare Agent Standards. EPA 600/R-
13/044. Environmental Protection Agency (EPA), Office of Research and Development,
Cincinnati, OH.
Malloy, Thomas A., 2012. Chemical Agent Sampling from Surfaces Using Solid Phase
Extraction Disks. National Environmental Monitoring Conference. August 9, 2012.
http://nemc.us/docs/2012/presentations/Thu-PM-lnnovative-TomMallov-8-8-12.pdf.
Stone 2013. Surface decontamination for blister agents Lewisite, sulfur mustard and agent
yellow, a Lewisite and sulfur mustard mixture. H. Stone, D. See, A. Smiley, A. Ellingson, J.
Schimmoeller, and L. Oudejans. Journal of Hazardous Materials 314 (2016) 59-66.
58
-------
Appendix A
Table A-l. Environmental conditions during the method development / demonstration and
subsequent fate and transport tests of VX and HD on paints and sealants.
Activity
Report
Section
Temperature Range
(ฐC)
RH Range
(%)
Wipe Method Development and
Demonstration for Paints
3.2.1
21.6-23.6
68-77
Wipe Method Development and
Demonstration for Paints - DRC and
3.2.1
19.7-22.0
41-47
acetone only
Extraction Method Development for Paints
3.2.2
19.0-20.8
44-49
Extraction Method Development for Paints
- Propagation of VX/HD from SPE
3.2.2
16.4-21.8
36-50
Wipe Method Demonstration Results for
Sealants
3.2.3
20.5-22.6
31-35
Fate and Transport Painted Surface - VX
3.3.1
18.7-21.1
56-65
Fate and Transport Painted Surface - HD
3.3.2
19.1-21.8
52-70
Fate and Transport Sealed Surface - VX*
3.3.3
18.4-22.6
28-49
Fate and Transport Sealed Surface - HD*
3.3.3
18.4-22.6
28-49
*: Simultaneous tests for both agents
59
-------
vvEPA
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
------- |