EPA/600/R-20/047 | May 2020
www.epa.gov/homeland-security-research
United States
Environmental Protection
Agency
&EPA
Physical and Chemical Removal
Options for Porous/Permeable
Materials Contaminated with the
Persistent Chemical Warfare
Agent VX
Office of Research and Development
Homeland Security Research Program
-------
This page left intentionally blank
-------
Physical and Chemical Removal Options for Porous/Permeable
Materials Contaminated with the Persistent Chemical Warfare
Agent VX
U.S. Environmental Protection Agency
Office of Research and Development
Center for Environmental Solutions and Emergency Response
Research Triangle Park, NC 27711
iii
-------
EPA/600/R-20/May 2020
DISCLAIMER
The U.S. Environmental Protection Agency (EPA) through its Office of Research and
Development funded and managed the research described herein under Contract Number EP-C-
15-002, Task Order 0020 with Battelle. It has been subjected to the Agency's review and has
been approved for publication. Note that approval does not signify that the contents necessarily
reflect the views of the Agency. Any mention of trade names, products, or services does not
imply an endorsement by the U.S. Government or EPA. The EPA does not endorse any
commercial products, services, or enterprises. The contractor role did not include establishing
Agency policy.
Questions concerning this document, or its application should be addressed to:
Lukas Oudejans, Ph.D.
Center for Environmental Solutions and Emergency Response
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
iv
-------
EPA/600/R-20/May 2020
ACKNOWLEDGMENTS
Contributions of the following individuals and organization to this report are gratefully
acknowledged:
U.S. Environmental Protection Agency (EPA) Project Team
Lukas Oudejans (Principal Investigator, Office of Research and Development, Center for
Environmental Solutions and Emergency Response (ORD/CESER), Homeland Security
and Materials Managements Division (HSMMD)
Shannon Serre and Larry Kaelin, Office of Land and Emergency Management, Office of
Emergency Management Consequence Management Advisory Division
(OLEM/OEM/CMAD)
Paul Lemieux and Timothy Boe, ORD/CESER/HSMMD
Cathy Young, Region 1
Charlie Fitzsimmons, Region 3
US EPA Technical Reviewers of Report
Kathy Hall, ORD/CESER/HSMMD
Lance Brooks, ORD/CESER/HSMMD
US EPA Quality Assurance
Ramona Sherman
Battelle
Carissa Dodds
William Hayes
Amy Andrews
Anthony Ellingson
David See
The National Caucus and Center on Black Aging, Inc.
Joan Bursey (Technical Editing)
v
-------
EPA/600/R-20/May 2020
EXECUTIVE SUMMARY
The U.S. Environmental Protection Agency (EPA) is responsible for preparing for, responding
to, and recovering from threats to public health, welfare, or the environment caused by actual or
potential hazardous materials incidents. Following either an accidental or intentional release of
chemical warfare agents (CWAs), porous building materials and permeable coatings such as
paints or sealants are likely to become contaminated. Residual CWA can then absorb into the
materials, coatings, and into the material under such coatings. The reversal of the absorption
process may not be possible or would likely take place at a rate significantly slower than the
evaporation of CWAs from hard nonporous material surfaces. Further, absorbed CWA may
become inaccessible to surface decontaminants applied as aqueous liquids due to the inability of
these decontaminants to penetrate sufficiently into the material or coating. Thus, the best course
of remediation of absorbed contamination may ultimately involve physical removal of the
contaminated materials or coatings. The ideal physical removal process would eliminate the
residual CWA entirely while simultaneously minimizing the amount of contaminated waste
generated and maintaining the integrity of the item or structure from which the materials or
coatings were removed. Following physical removal of contaminated materials and/or coatings,
the item or structure could then be resurfaced and repainted as necessary and returned to service
while the removed materials would be managed as waste according to federal, state, and local
requirements.
The primary objective of this project was to quantitatively evaluate the efficacy of select physical
removal technologies and determine the application conditions/methods necessary to
decontaminate CWA-contaminated porous materials and permeable coatings through physical
removal of the contaminated portions of the materials. Prior to testing, literature searches were
performed to identify physical removal mechanisms that could be used to remove contaminated
portions of porous materials and/or permeable coatings while simultaneously minimizing
damage to the materials and generation of wastes that may be classified as hazardous waste.
From the literature search results, grinding and chemical stripping were selected for further
evaluation. Grinding was evaluated for efficacy in removal of contaminated portions of sealed
concrete and limestone, and chemical stripping was evaluated for efficacy in removal of
contaminated coatings from low-carbon steel and hardwood.
Bench scale studies were performed using neat O-ethyl S-(2-diisopropylaminoethyl)
methylphosphonothioate (VX) as the challenge CWA. The porous materials and permeable
coatings were contaminated with a target 10 |iL of VX (equivalent to a contaminant mass of 9.4
mg of VX). The VX was allowed to dwell on the surface of the materials for a period of 24
hours to allow for penetration into the materials. Following the 24-hour dwell period, the porous
material and permeable coating coupon surfaces were sampled via wipe sampling to quantify
residual, transferable VX. Following wipe-sampling, the physical removal technologies under
test were applied to remove the contaminated portions of the material coupons. Grinding was
used to remove portions of sealed concrete and limestone at discrete 0.25 inch (in.) -thick depth
vi
-------
EPA/600/R-20/May 2020
layers. Chemical stripper was applied to the coated steel and hardwood coupons to remove the
paint/primer layers. Ground material removed from sealed concrete and limestone and coatings
stripped from steel and hardwood were extracted with solvent, and extracts were analyzed via
liquid chromatography-tandem mass spectrometry (LC-MS/MS) to quantify VX recovered from
the removed materials. The surface of steel and wood coupons was also sampled via wipe
sampling again following stripping.
Grinding Results:
Total percent of VX recovery from sealed concrete averaged only 8.5% compared to the
associated VX spike control mean recovery. Average total percent recovery from limestone was
markedly higher, at 47%. The major portion of the VX recovered from each sealed concrete and
limestone coupon via application of the grinding technology was obtained in the topmost 0.25-in.
of the material, to which the VX challenge was applied. Recoveries from the 2nd, 3rd and 4th
grinded 0.25-in. layers decreased sharply to less than 1% of the spike control mean recovery in
all cases except that of the 2nd limestone coupon, in which recoveries from the 3rd and 4th 0.25-in.
ground layer samples remained as high as 5.8% and 5.4%, respectively. However, it cannot be
discerned from the data whether lower detections in deeper layers are due to the absence of VX
(i.e., VX did not penetrate past the topmost 0.25-in. layer), degradation of VX, or an inability to
recover VX that is present. Thus, physical removal to a greater depth than just the topmost 0.25-
in. of material may be necessary. While the data suggest that VX contamination in porous
materials can be removed via application of grinding to remove contaminated portions of the
materials, the generally low total recoveries as well as the relatively higher recoveries from
deeper layers from the 2nd limestone coupon suggest that the necessary removal depths can be
inconsistent.
Paint Stripper Results:
Generally, greater recoveries of VX were obtained from the painted steel than from painted
wood coupons. Markedly less VX was recovered from the post-stripping wipe samples taken
from the steel substrate, indicating that the majority of the VX contamination was removed by
the first (pre-stripping) wipe and by removal of the permeable coating via application of the
stripper. Assuming VX does not permeate into the steel substrate, the data suggest that
remediation of VX-contaminated painted/coated steel via a combination of solvent wipe
sampling and removal of the paint/coating via chemical stripping may be possible (though
repeated solvent wipe sampling and application of the stripper may be required, depending on
the required decontamination level). The lower total recoveries from painted wood samples as
well as the higher recoveries from post-stripping wipe samples taken from the wood coupons
suggest that VX may be permeating through the paint/coating layer and into the underlying
permeable wood substrate. Such residual VX contamination could pose contact or vapor hazards
vii
-------
EPA/600/R-20/May 2020
if the VX diffuses back to the surface of the wood or if the wood is cut, ground, or otherwise
manipulated.
Waste Considerations:
Application of both the grinding and chemical stripping technologies generated wastes that
retained measurable levels of VX. Such wastes included the porous materials removed by
grinding and the permeable coatings removed from substrates via application of the chemical
stripper. Both waste types would require collection and handling using appropriate PPE and
managed appropriately as per federal, state, or local requirements. It must be noted that these
wastes are likely regulated at the state level and it is crucial to discuss the management of these
wastes prior to the time at which they are generated.
Once acceptable levels of decontamination are reached, concrete, limestone, and similar porous
materials and nonporous substrates such as steel would likely be amenable to resurfacing or
recoating and reuse following application of the grinding or chemical stripping technologies. An
exception to this may be porous substrates from which permeable coatings are removed, such as
the hardwood substrate which could be at risk of excessive damage if repeated chemical
stripping applications or additional physical removal methods (beyond removal of the coating)
are required to achieve acceptable levels of decontamination.
Health and Safety Considerations:
Grinding of porous materials such as concrete and limestone will result in dust formation. Dust
mitigation will be required since small dust particles carrying agent contamination will likely
become redistributed in the environment (and potentially transfer to other materials). Some of the
potentially contaminated particulate matter may become an inhalation hazard.
viii
-------
EPA/600/R-20/May
TABLE OF CONTENTS
EXECUTIVE SUMMARY
LIST OF ACRONYMS AND ABBREVIATIONS
1. INTRODUCTION
1.1. Purpose
1.2. Proj ect Obj ectives
1.3. Test Facility Description
1.4. Staff and Resources
2. EXPERIMENTAL METHODS
2.1. Experimental Design
2.1.1. Methods Demonstration
2.1.2. Physical Removal Efficacy Evaluation
2.1.3. Waste Generation Assessment
2.1.4. Damage Extent Assessment
2.2. Experimental Methods and Materials
2.2.1. Porous Materials and Permeable Coatings
2.2.2. Application of CWA
2.2.3. Description and Application of Physical Removal Technologies ..
2.2.4. Coupon Surface (Wipe) Sampling for VX
2.2.5. Extraction of VX from Wipe, Coupon, and Waste Samples
2.3. Analytical Methods
2.4. Calculations
3. RESULTS
3.1. Methods Demonstration
3.1.1. Physical Removal Technology Functionality Assessment Results
3.1.2. Solvent Extraction Method Demonstration Results
3.1.3. Wipe Sampling Method Demonstration Results
3.1.4. Waste Sampling Results
3.1.5. VX Depth Penetration Assessment Results
3.2. Physical Removal Efficacy Results - Grinding
3.3. Physical Removal Efficacy Results - Chemical Stripping
3.4. Waste Generation Assessment Results
3.5. Damage Extent Assessment Results
4. QUALITY ASSURANCE/QUALITY CONTROL
4.1. Data Quality Indicators
4.2. Instrument Calibration
4.2.1. Calibration Schedules
4.2.2. LC-MS/MS Calibration
4.3. Sample Handling and Custody
-------
EPA/600/R-20/May 2020
4.4. Technical Systems Audit 72
4.5. Performance Evaluation Audits 73
4.6. Data Quality Audit 73
5. SUMMARY 75
6. REFERENCES 85
x
-------
EPA/600/R-20/May 2020
LIST OF TABLES
Table 1. Anticipated Nature of Coupon and Waste Samples 6
Table 2. Solvent Extraction Method Demonstration Test Matrix 9
Table 3. Wipe-Sampling Method Demonstration Test Matrix 11
Table 4. Waste Sampling Method Demonstration Test Matrix 13
Table 5. VX Depth Penetration Assessment Test Matrix 15
Table 6. Physical Removal Efficacy Test Matrix 16
Table 7. Porous Materials and Permeable Coatings 17
Table 8. GC/FID VX Purity Sample Analysis Method Parameters 18
Table 9. VX Purity by Test 18
Table 10. Analyte Ion Transitions 29
Table 11. LC-MS/MS Conditions for VX Analysis 29
Table 12. Temperature Increase for Core Excision using Drill Press 34
Table 13. Limestone Core Slicing Results 35
Table 14. Sealed Concrete Core Slicing Results 35
Table 15. Slice Thickness Results 35
Table 16. Limestone Grinding Results 38
Table 17. Sealed Concrete Grinding Results 39
Table 18. Chemical Stripper Application Methods and Spread Results 41
Table 19. Solvent Extraction, Spike Controls 46
Table 20. Solvent Extraction, Stainless Steel Positive Controls 46
Table 21. Solvent Extraction, Glass Bead Positive Controls 46
Table 22. Solvent Extraction, Core Layer Samples 47
Table 23. Solvent Extraction, Ground Material Samples 48
Table 24. Wipe Sampling, Spike Controls 50
Table 25. Wipe Sampling, Stainless Steel Positive Controls 50
Table 26. Wipe Sampling Results 50
Table 27. Waste Sampling, Spike Controls 52
Table 28. Waste Sampling, Glass Bead Positive Controls 53
Table 29. Waste Sampling, Cutting Dust Recovery 53
Table 30. Waste Sampling, Stripped Coating Recovery 54
Table 31. VX Depth Penetration Assessment, Spike Controls 55
Table 32. VX Depth Penetration Assessment, Limestone Recovery 56
Table 33. VX Depth Penetration Assessment, Sealed Concrete Recovery 56
Table 34. Grinding, Spike Controls 58
Table 35. Sealed Concrete, Ground Layer Masses 59
Table 36. Limestone, Ground Layer Masses 59
Table 37. Grinding, Sealed Concrete Recovery 60
Table 38. Grinding, Limestone Recovery 60
Table 39. Chemical Stripping, Spike Controls 63
Table 40. Chemical Stripping, Positive Control Recovery 63
Table 41. Chemical Stripping, Painted Steel Recovery 64
Table 42. Chemical Stripping, Painted Wood Recovery 64
Table 43. Data Quality Indicators and Results 70
xi
-------
EPA/600/R-20/May 2020
Table 44. Equipment Calibration Schedule 71
Table 45. LC-MS/MS Performance Parameters and Acceptance Criteria 72
Table 46. Performance Evaluation Audit Results 73
LIST OF FIGURES
Figure 1. Single Coupon Spiking Arrangement for Multiple Wipe Replicates 11
Figure 2. Core Sample Holder with Dust Collection Tray 21
Figure 3. Porous Material Core Sampling Approach 21
Figure 4. Core Sampling Dissection 22
Figure 5. Approach for Depth Layer Sample Collection via Grinding 23
Figure 6. Concrete Coupon on Aluminum Easel 24
Figure 7. Vapor-Phase Solvent Extraction Apparatus 26
Figure 8. Wipe Pattern 28
Figure 9. Core Excision Setup 32
Figure 10. Core Slicing Setup 33
Figure 11. Broken Core 33
Figure 12. Cores Excised from Limestone (left) and Sealed Concrete (right) 34
Figure 13. Core Slices from Limestone (left) and Sealed Concrete (right) 36
Figure 14. Cutting Dust from Limestone (left) and Sealed Concrete (right) 36
Figure 15. Grinding Setup 38
Figure 16. Depth Measurement 38
Figure 17. Limestone Ground to 0.25 in. (left) and 1 in. (right) 39
Figure 18. Sealed Concrete Ground to 0.25 in. (left) and 1 in. (right) 40
Figure 19. Ground Material 40
Figure 20. Coating Removal 41
Figure 21. Coated Steel Before (left) and After Scraping (right) 42
Figure 22. Coated Wood Before (left) and After Scraping (right) 42
Figure 23. Collected Coating Using 1 mL Application with Brushing 42
Figure 24. Collected Coating Using 1.5 mL Application without Brushing 43
Figure 25. Vial for Solvent Collection 43
Figure 26. Revised Vapor-Phase Solvent Extraction Test Setup 45
Figure 27. Solvent Extraction, Average Mass Recovery 49
Figure 28. Solvent Extraction, Average Percent Recovery 49
Figure 29. Wipe Sampling, Average Mass Recovery 51
Figure 30. Wipe Sampling, Average Percent Recovery 51
Figure 31. Waste Sampling, Average Mass Recovery 54
Figure 32. Waste Sampling, Average Percent Recovery 55
Figure 33. VX Depth Penetration Assessment, VX Mass Recovery by Component 57
Figure 34. Grinding, VX Mass Recovery by Component 61
Figure 35. Chemical Stripping, VX Mass Recovery by Component 65
Figure 36. Ground Sealed Concrete Coupon 67
Figure 37. Ground Limestone Coupon 68
Figure 38. Stripped Coating, Steel 68
Figure 39. Stripped Coating, Hardwood 69
Figure 40. Solvent Extraction, Average Percent Recovery 76
Figure 41. Wipe Sampling, Average Percent Recovery 76
Figure 42. Waste Sampling, Average Percent Recovery 77
Figure 43. VX Depth Penetration Assessment, VX Mass Recovery by Component 79
Figure 44. Grinding, VX Mass Recovery by Component 80
xii
-------
EPA/600/R-20/May 2020
Figure 45. Chemical Stripping, VX Mass Recovery by Component 81
Figure 46. Limestone and Sealed Cone. Recovery, Grinding vs Core Sampling Comparison 84
ATTACHMENT
Attachment A - Environmental Data
Xlll
-------
EPA/600/R-20/May 2020
LIST OF ACRONYMS AND ABBREVIATIONS
AMC Army Materiel Command
CASARM Chemical Agent Standard Analytical Reference Material
CCDC Combat Capabilities Development Command
CCV Continuing Calibration Verification
CESER Center for Environmental Solutions and Emergency Response (EPA)
cm Centimeter(s)
cm2 Square centimeter(s)
CoC Chain of Custody
CoV Coefficient of Variation
CWA Chemical Warfare Agent
EPA U.S. Environmental Protection Agency
FID Flame ionization detector
GC Gas Chromatography
g Gram(s)
h Hour(s)
HMRC Hazardous Materials Research Center
HPLC High Performance Liquid Chromatography
HSRP Homeland Security Research Program
in. Inch(es)
IPA Isopropyl Alcohol
IS Internal Standard
|_iL Microliter(s)
L Liter(s)
LC-MS/MS Liquid Chromatography-Tandem Mass Spectrometry
LLOQ Lower Limit of Quantitation
LRB Laboratory Record Book
min Minute(s)
mL Milliliter(s)
mm Millimeter(s)
MRM Multiple reaction monitoring
NA Not applicable
ng Nanogram(s)
Pa Pascal
PB Procedural blank
PC Positive control
PPE Personal protective equipment
psi Pounds per square inch
PTFE polytetrafluoroethylene
QAPP Quality Assurance Project Plan
r2 Coefficient of determination
RDS Research dilute solution
RDT&E Research, development, test and evaluation
RH Relative Humidity
RSD Relative Standard Deviation
SC Spike control
SS Stainless Steel
TPCS Test Parameter Control Sheet
TSA Technical Systems Audit
VR Russian VX
xiv
-------
EPA/600/R-20/May 2020
VX O-ethyl S-(2-diisopropylaminoethyl) methylphosphonothioate
xv
-------
EPA/600/R-20/May 2020
1. INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is responsible for preparing for, responding
to, and recovering from threats to public health, welfare, or the environment caused by actual or
potential hazardous materials incidents. Hazardous materials include chemical, biological, and
radiological substances, whether accidentally or intentionally released. The threat of a chemical
agent release into the environment is driving EPA's Homeland Security Research Program
(HSRP) to systematically evaluate potential decontamination technologies for chemical agents.
Following either an accidental or intentional release of chemical warfare agents (CWAs), many
building materials, including porous building materials and/or permeable coatings such as paints
or sealants are likely to become contaminated. Residual CWA can then absorb into the materials
and coatings in a similar manner as how CWAs can permeate into and through gloves or other
personnel protective equipment (PPE) [1], The reversal of absorption may not be possible, and
even if it is, would likely take place at a rate significantly slower than the evaporation of CWAs
from hard nonporous material surfaces [2-4], Further, absorbed CWA may become inaccessible
to surface decontaminants applied as aqueous liquids due to the inability of the decontaminants
to penetrate sufficiently into the material or coating [4],
Thus, the best course of remediation of absorbed contamination may involve physical removal of
the contaminated materials or coatings. The ideal physical removal process would eliminate the
residual CWA entirely while simultaneously minimizing the amount of contaminated waste
generated and maintaining the integrity of the item or structure from which the materials or
coatings were removed [5], Following physical removal of contaminated materials and/or
coatings, the item or structure could then be resurfaced as necessary and returned to service. EPA
first responders have identified this high-priority knowledge gap for the HSRP to address.
1.1. Purpose
The purpose of this project was to determine the application conditions necessary and evaluate
the efficacies of select methodologies for remediation of CWA-contaminated porous materials
and permeable coatings through physical removal of the contaminated portions of the materials,
while simultaneously minimizing the amount of hazardous wastes generated and maintaining the
integrity of the surfaces or structures to which the technologies are applied.
1.2. Project Objectives
The primary objective of this project was to quantitatively evaluate the efficacy of select
technologies to physically remove CWA-contaminated portions of select porous materials and
permeable coatings through performance of bench scale laboratory studies using neat O-ethyl S-
(2-diisopropylaminoethyl) methylphosphonothioate (VX, CAS 50782-69-9) as the challenge
CWA.
1
-------
EPA/600/R-20/May 2020
The fate and transport of VX in concrete has been studied in detail by many research groups. The
general consensus is that while there is a measured fast degradation (half-life time of 2-3 hours)
[6], this fast process may only apply to VX when present at very low concentrations. The
presence of higher quantities as used in this study will lead to a prolonged persistence (weeks to
months) of VX that permeated into the concrete [7],
Prior to the physical removal efficacy evaluations, technologies that were anticipated to be
efficacious, generate minimal or no wastes, and minimize or eliminate irreparable damage were
identified via searches of existing literature and secondary data. Physical removal technologies
identified during the search were evaluated based on four primary characteristics, including:
• Anticipated or demonstrated efficacy of the technologies in removal of hazardous
contamination through physical removal of the contaminated portions of a material.
• The types, quantities, and hazard designations of wastes generated from application of the
technologies.
• The extent of irreparable damage caused (or anticipated to be caused) to
surfaces/structures from application of the physical removal technologies (factors that
impact the possibility, cost, and level of effort associated with resurfacing/restoring
treated surfaces/structures and returning them to service).
• The application rate, ease-of-use, and personal protective equipment (PPE) and cost
requirements associated with the technologies.
From the technologies identified during the literature searches, three were selected for physical
removal efficacy evaluation during this project:
• Grinding was evaluated for efficacy in removal of VX-contaminated portions of sealed
concrete and limestone.
• Chemical stripper was evaluated for efficacy in removal of VX-contaminated permeable
coatings (paint and primer) from the surface of low-carbon steel and hardwood.
• Vapor-phase solvent extraction was selected for evaluation for efficacy in removal of VX
contamination from both porous materials (sealed concrete and limestone) and permeable
coatings (paint and primer applied to low-carbon steel and hardwood). However, technical
difficulties associated with setup and application of the technology prevented a full
evaluation within this project. Refer to Sections 2.2.3.4 and 3.1.1.4 for additional
information.
Additionally, a method independent of the selected physical removal technologies, referred to as
the "core sampling approach", was developed and used for dissection of porous materials (sealed
concrete and limestone) to quantify the extent of VX penetration/contamination as a function of
depth. This core sampling experiment was conducted prior to the grinding tests and supported
the depths to which porous materials were removed.
2
-------
EPA/600/R-20/May 2020
Refer to Sections 2.2.3.1 through 2.2.3.4 for additional information related to the physical
removal technologies, the core sampling approach, and the application strategies used for each.
Section 2.2.1 provides information related to the porous materials and permeable coatings used
as test items.
Also, during the physical removal efficacy evaluation, quantities and types of hazardous (or
potentially hazardous) wastes generated were assessed, as well as the extent of damage caused to
the materials from application of the technologies. Waste generation assessments were
quantitative where possible (waste volumes and CWA contamination levels) and otherwise
qualitative (visual observations and descriptions of waste types). Damage extent to a material
was qualitatively determined and included visual assessments and descriptions of damage
caused.
1.3. Test Facility Description
All testing was performed at Battelle's Hazardous Materials Research Center (HMRC) located in
West Jefferson, Ohio. The HMRC is certified to work with chemical surety material under a
Provisioning Agreement with oversight by the U.S. Army Materiel Command (AMC;
Provisioning Agreement Battelle-1). Wherever applicable and required, the reporting
requirements for this agreement were followed.
1.4. Staff and Resources
Quantitative physical removal efficacy testing, associated methods demonstration testing, and
qualitative assessments of waste generation and damage extent were completed using staff and
resources from Battelle's HMRC (West Jefferson, OH) in consultation with the EPA's Center for
Environmental Solutions and Emergency Response (CESER).
3
-------
EPA/600/R-20/May 2020
2. EXPERIMENTAL METHODS
2.1. Experimental Design
Project objectives were achieved through execution of physical removal efficacy tests, waste
generation assessments, and material damage extent assessments. Generally, physical removal
efficacy testing proceeded according to the following approach:
• Test articles of the porous materials and permeable coatings selected for evaluation were
contaminated with VX, and the VX was allowed to dwell on the test article surface for a
period of 24 hours. Environmental conditions during the dwell period were monitored but
not controlled.
• Following the VX dwell period, test article surfaces were sampled via surface wipe-
sampling to evaluate residual VX surface hazard following the dwell period.
• After wipe sampling, test articles were transferred into a test chamber, if required (core
sampling approach and grinding technology), for application of the physical removal
technology.
• The physical removal technology was then applied to remove the contaminated portions of
the material.
• Samples of the porous material or permeable coating removed, samples of any waste
generated, and samples taken from the physical removal technology itself (where
applicable, e.g., wipe samples of the grinding wheel of the grinder) were analyzed via
liquid chromatography-tandem mass spectrometry (LC-MS/MS) to quantify residual VX
present in/on each.
During the project, test articles consisted of coupons of each of the porous materials or
permeable coatings (coatings applied to a substrate material) selected for the evaluation. Exact
coupon sizes for each material type were determined during assessments of the functionality of
the physical removal technologies selected for testing. Coupons for all material types were sized
adequately for proper application of the physical removal technologies. Refer to Section 2.2.1 for
information on the porous materials and permeable coatings included in the evaluation.
Prior to physical removal efficacy testing, the experimental methods planned for use were
demonstrated to ensure valid data would be generated. The experimental designs for each of
these phases of testing, including technology functionality assessments, methods demonstration,
physical removal efficacy testing, waste generation assessment, and damage extent assessment
are described in the following subsections.
4
-------
EPA/600/R-20/May 2020
2.1.1. Methods Demonstration
2.1.1.1. Physical Removal Technology Functionality Assessments
Detailed descriptions of the application approaches for each of the physical removal technologies
selected for evaluation as well as for the core sampling approach and for collection of wastes
generated from application of the technologies are described in Sections 2.2.3.1 through 2.2.3.4.
Prior to methods demonstration testing using VX, the functionality of each of the physical
removal technologies selected for testing and of the core sampling approach were evaluated
without VX present.
Coupon samples produced from application of each technology to each material type included
ground material (via grinder applied to sealed concrete and limestone) and excised layer samples
("slices") from material cores (via the core sampling approach applied to sealed concrete and
limestone). Physical coupon samples were not collected/harvested from painted/coated materials
following application of chemical stripper; rather, the treated surfaces of the materials were
sampled via wipe sampling.
Waste samples produced from application of each technology to each material type included the
dust created during excision of layer samples ("slices") from material cores during application of
the core sampling approach (Section 2.2.3.1) and the permeable coatings removed following
application of chemical stripper to painted steel and wood coupons (Section 2.2.3.3). Separate
waste samples were not collected during application of the grinding technology (the entirety of
the ground material produced during application of the grinder was collected as the coupon
sample).
Application of the vapor-phase solvent extraction technology to sealed concrete and limestone
cores was planned to be followed by application of the core sampling approach to harvest
coupon samples (successive depth layer "slices") and waste samples (the dust produced during
core sample cutting). As with application of chemical stripper, no physical coupon samples
would be collected from painted materials following application of vapor-phase solvent
extraction; rather, the surface of coupons would be sampled via wipe sampling. Condensed
solvent collected during and following application of vapor-phase solvent extraction to all
material types was planned for collection and analysis as well.
The form/nature of the coupon and waste samples collected from application of each technology
to each of the materials is described/summarized in Table 1.
5
-------
EPA/600/R-20/May 2020
Table 1. Anticipated Nature of Coupon and Waste Samples
Physical Removal
Technology
Material
Form/Nature of Coupon
Samples
Form/Nature of Waste
Samples
Core sampling approach A
Sealed concrete
Excised coupon sections A
Cutting dust
Core sampling approach A
Limestone
Excised coupon sections A
Cutting dust
Grinding
Sealed concrete
Ground material
Not Collected
Grinding
Limestone
Ground material
Not Collected
Vapor-phase solvent extraction
Sealed concrete
Excised coupon sections A
Cutting dust, collected solvent
Vapor-phase solvent extraction
Limestone
Excised coupon sections A
Cutting dust, collected solvent
Chemical Stripping
Painted steel
Post-treatment wipe sample
Stripped paint layer
Chemical Stripping
Painted hardwood
Post-treatment wipe sample
Stripped paint layer
Vapor-phase solvent extraction
Painted steel
Post-treatment wipe sample
Collected solvent
Vapor-phase solvent extraction
Painted hardwood
Post-treatment wipe sample
Collected solvent
NA = Not applicable A Refer to Section 2.2.3.1
To perform the physical removal technology functionality assessments, the technologies were
applied to the porous materials and permeable coatings selected for the efficacy evaluations as
described in Sections 2.2.3.1 through 2.2.3.4, using all identified equipment, procedures, and test
samples. Functionality assessments of the core sampling approach and grinding technology took
place in the test chamber, to assess the ergonomics and limitations associated with working with
the technologies inside the chamber.
These preliminary assessments were intended to evaluate the adequacy of the planned
approaches for application of each technology, assess "ease of use" for each
technology/approach, and aid in refinement (as determined necessary) of the planned approaches
for application of the technologies. Additional objectives of the functionality assessments
included:
• Determination of the exact coupon sizes required for each porous material or permeable
coating type included during physical removal efficacy testing efforts. Section 2.2.1
provides additional details (including coupon sizes) related to the porous materials and
permeable coatings included in the evaluations.
• Determination/confirmation of the nature and dimensions of coupon samples obtained from
the material coupons via application of the selected removal methods (e.g., volume/mass of
ground solids; dimensions of core layer samples).
• Determination/confirmation of the nature and amount of wastes generated by the removal
methods (e.g., volume/mass of sealed concrete and limestone cutting dust collected during
application of the core sampling approach; size/amount and characteristics of the stripped
paint layers; volume of condensed solvent collected during application of vapor-phase
solvent extraction to porous materials and permeable coatings) and how the wastes would
be collected most efficiently.
• Investigation of temperature increase in the materials during application of the core
sampling approach.
6
-------
EPA/600/R-20/May 2020
o Application of the core sampling approach to excise layer samples from the sealed
concrete and limestone material cores could generate heat and increase the temperature
of the material.
o Following dissection of each core sample into individual coupon layer samples, the
temperature of the coupon layer samples was measured using a calibrated non-contact
infrared thermometer (9248T57, McMaster-Carr, Aurora, OH).
o Temperature of the material cores was also measured prior to application of the core
sampling approach so that the increase in temperature due to application of the
approach could be determined.
Additionally, the various coupon samples and waste samples produced for each
material/technology combination (refer to Table 1) were retained for use during subsequent
solvent extraction, wipe sampling, and waste sampling methods demonstration.
2.1.1.2. Sampling for Residual VX in Material Samples (Solvent Extraction)
As discussed in Section 2.1.1.1, the coupon samples generated during the physical removal
technology functionality assessments were retained for use during testing to demonstrate the
adequacy of the methods that were used to recover VX from the samples via solvent extraction.
As identified in Table 1, such samples included:
• Excised layer samples obtained from sealed concrete and limestone cores following
application of the core sampling approach.
• Ground sealed concrete and limestone material recovered from coupons following
application of the grinding technology.
Replicate material core layer samples were each spiked in the center with 2 microliters (|iL) of
VX according to the procedures described in Section 2.2.2.2, and the VX was allowed to dwell
on the material layer sample surface for either 30 minutes (min) or 24 hours. Although a
challenge volume of 10 |iL per coupon was used during physical removal efficacy testing and the
VX depth penetration assessment, a challenge volume of 2 |iL per replicate material core layer
sample was used during solvent extraction method demonstration testing. At the time of the
method demonstration, it was anticipated that 2 |iL would be more representative of the amount
of VX that would penetrate into the porous materials during the 24-hour dwell period and thus
require extraction/recovery from the core layer samples (especially for deeper core layer
samples). Coupon layer samples were placed on top of polytetrafluoroethylene (PTFE) disks
during the VX dwell period. At the end of the 30-minute or 24-hour VX dwell period, the
coupon layer samples were extracted individually with solvent according to procedures described
in Section 2.2.5. The PTFE disks underneath each material core layer sample were extracted
separately to determine if the VX applied to the samples migrated through the (target) 0.25-inch
(in.) -thick core layer samples during the VX dwell period.
7
-------
EPA/600/R-20/May 2020
In addition to the sealed concrete and limestone core layer samples, stainless steel coupons (same
surface dimensions as the coupon layer samples, i.e., round, 1.375-in. diameter coupons, but with
a thickness of 24 gauge) were spiked with VX. The VX was allowed to dwell for the same
duration as on the core layer samples, and the steel coupons were extracted with solvent
alongside the core layer samples to act as a control material.
Ground sealed concrete and limestone samples were contained in glass jars (125 milliliters [mL]
volume). Each sample of ground material was spiked with 2 |iL of VX according to procedures
described in Section 2.2.2.2, and the VX was allowed to dwell within the ground material for
either 30 minutes or 24 hours. A challenge volume of 2 |iL (as opposed to 10 |iL for physical
removal efficacy testing) was used for each ground material sample as it was anticipated that 2
|iL would be more representative of the amount of VX that would penetrate into the porous
materials during the 24-hour dwell period. Following the VX dwell period, the ground sealed
concrete or limestone material in each jar was extracted with solvent according to procedures
described in Section 2.2.5 (solvent was added directly to the jars containing the ground material
samples).
In addition to the ground sealed concrete and limestone samples, 3-millimeter (mm) glass beads
(10-310-1, Fisher Scientific, Pittsburgh, PA; approximately 61.4 grams [g] of beads per sample)
were spiked with VX. The VX was allowed to dwell for the same duration as that on the ground
material samples, and the beads were then extracted with solvent alongside the ground material
test samples to act as a control material.
Isopropyl alcohol (IPA; A464-4, Fisher Scientific, Pittsburgh, PA) was evaluated as the core
layer and ground material extraction solvent. IPA was selected because of its use to recover
CWA via extraction from materials during previous studies [8] and because of its amenability for
use with LC-MS/MS analysis.
Table 2 provides the matrix for solvent extraction method demonstration testing that was
performed for this project.
8
-------
EPA/600/R-20/May 2020
Table 2. Solvent Extraction Method Demonstration Test Matrix
Material
Sample Form A
Sample Type
VX
Challenge
(HL)
Dwell
Period
Replicates
Sealed concrete
Coupon layer sample B
Test sample
2
30 min
3
Sealed concrete
Coupon layer sample B
Test sample
2
24 hours
3
Sealed concrete
Coupon layer sample B
Procedural blank
NA
24 hours
1
Limestone
Coupon layer sample B
Test sample
2
30 min
3
Limestone
Coupon layer sample B
Test sample
2
24 hours
3
Limestone
Coupon layer sample B
Procedural blank
NA
24 hours
1
Stainless steel
Disk c
Positive control
2
30 min
3
Stainless steel
Disk c
Positive control
2
24 hours
3
Stainless steel
Disk c
Procedural blank
NA
24 hours
1
Sealed concrete
Ground material
Test sample
2
30 min
3
Sealed concrete
Ground material
Test sample
2
24 hours
3
Sealed concrete
Ground material
Procedural blank
NA
24 hours
1
Limestone
Ground material
Test sample
2
30 min
3
Limestone
Ground material
Test sample
2
24 hours
3
Limestone
Ground material
Procedural blank
NA
24 hours
1
Glass beads
Loose beads
Positive control
2
30 min
3
Glass beads
Loose beads
Positive control
2
24 hours
3
Glass beads
Loose beads
Procedural blank
NA
24 hours
1
A Coupon layer samples and ground material samples retained following functionality assessments (refer to
Section 2.1.1.1). Ground material masses collected are provided in Section 3.1.1.2.
B Dimensions: 1.375 in. diameter, 0.25 in. thickness
c Dimensions: 1.375 in. diameter, 24-gauge thickness
In addition to the test samples (core layer samples and ground material samples), a single
procedural blank per material type/form was included (as indicated in Table 2). Procedural
blanks for solvent extraction method demonstration testing consisted of core layer samples and
ground material of the same porous material type as the associated test samples (sealed concrete
and limestone) that were not spiked with VX but that were extracted with solvent and analyzed
alongside the test samples. Additionally, three spike controls (SCs) were prepared to confirm the
VX challenge application amount. SCs consisted of a spike of an equal volume of VX (2 |iL)
directly into extraction solvent. As discussed in Section 2.2.2.2, when spiked, VX was applied to
the inside surface of the glass SC jar (rather than submerging the pipette tip into the solvent).
The coupon sample solvent extraction method would be deemed acceptable for use during
subsequent physical removal efficacy evaluations if the mean recoveries from stainless steel
(core layer sample controls) and glass beads (ground material sample controls) were within the
range of 70% to 120% of the mean of the SC results with a coefficient of variation (CoV)
between replicates of less than 30%.
Concrete has been shown to be capable of active degradation of penetrated/absorbed CWAs
[3,9], Data generated during previous studies [10] have demonstrated such difficulty in
recovering VX spiked onto concrete surfaces using similar solvent extraction techniques.
Further, while VX is commonly considered a persistent CWA (approximate vapor pressure of
0.09 pascals [Pa]), evaporation during the dwell period could still occur to some extent
9
-------
EPA/600/R-20/May 2020
(particularly during the 24-hour dwell period). While the propensity of limestone to degrade
penetrated/absorbed VX in a manner similar to concrete is unknown, it was assumed for this
testing that the evaporation rate of VX from limestone and concrete would be similar. Stainless
steel and glass provided inert surfaces, and previous studies [11,12] have demonstrated the
extractability of VX from the materials. Recoveries from stainless steel and glass provided an
indication of the amount of VX lost to evaporation during the dwell period. As it was assumed
that VX would not be absorbed or degraded by stainless steel or glass, comparison of the
recoveries from sealed concrete and limestone (core layer samples and ground material) were
compared to recoveries from stainless steel and glass to determine the amount of VX lost to
entrapment within or degradation by the concrete and limestone matrices.
Results of solvent extraction method demonstration testing are provided in Section 3.1.2.
2.1.1.3. Sampling for Residual VX on Material Surfaces (Wipe Sampling)
The coupon surface (wipe) sampling methods developed for CWAs during work on previous
EPA efforts [12] were evaluated for use in recovering residual surface VX contamination from
the materials selected for testing during this project. Wipe sampling was used to assess residual
VX surface hazard on the coupons following the 24-hour VX dwell period during physical
removal efficacy testing (prior to application of the physical removal technologies) and was also
applied to painted steel and painted hardwood following application of the chemical stripping
technology to assess efficacy of the removal method (refer to Tables 1 and 6; wipe-sampling was
also planned for use to quantify residual VX contamination following application of the vapor-
phase solvent extraction technology).
Wipe sampling method demonstration testing focused on recovery of VX from the porous
materials and permeable coatings selected for the physical removal efficacy evaluations.
Stainless steel was included as a control material. The procedure and materials used for wipe
sampling, including the specific wipe type that was used, are described in Section 2.2.4. Coupons
of each porous material or permeable coating type were contaminated with 10 |iL of VX, and the
VX was allowed to dwell on the surface of coupons for 24 hours. The 10-|iL VX challenge
volume and 24-hour dwell period were used during physical removal efficacy testing. A
challenge volume and VX dwell period of 10 |iL and 24 hours were used for the wipe sampling
method demonstration testing to ensure a representative amount of VX was remaining/available
on the surface of coupons at the time of wipe-sampling. This is in contrast to the 2 |iL VX
challenge volume per sample discussed for solvent extraction method demonstration testing.
Following the dwell period, coupons were wipe-sampled using the procedure described in
Section 2.2.4. Each wipe was extracted individually in solvent according to procedures described
in Section 2.2.5.
IPA (A464-4, Fisher Scientific, Pittsburgh, PA) was evaluated as the wipe wetting and wipe
extraction solvent. As during solvent extraction method demonstration testing, IPA was selected
10
-------
EPA/600/R-20/May 2020
because of its use to recover CWA via extraction from materials during previous studies [8] and
because of its amenability for use with LC-MS/MS analysis.
Table 3 provides the matrix for wipe-sampling methods demonstration testing that was
performed for this work.
Table 3. Wipe-Sampling Method Demonstration Test Matrix
Material
Sample Type
VX Challenge
(jiL)
Replicates A
Sealed concrete
Test coupon
10
3
Procedural blank
NA
1
Limestone
Test coupon
10
3
Procedural blank
NA
1
Painted steel
Test coupon
10
3
Procedural blank
NA
1
Painted hardwood
Test coupon
10
3
Procedural blank
NA
1
Stainless steel (control)
Test coupon
10
3
Procedural blank
NA
1
Multiple wipe replicates may be obtained from a single coupon, depending on coupon size.
Multiple sections of a single coupon were spiked with VX to provide the required number of
wipe sample replicates (each replicate area was spiked with 10 pL of VX). Figure 1 depicts the
coupon spiking arrangement for obtaining multiple wipe replicates from a single coupon.
CWA spiked
Figure 1. Single Coupon Spiking Arrangement for Multiple Wipe Replicates
In addition to the test coupons, a single procedural blank per material type was included (as
indicated in Table 3). Procedural blanks for wipe-sampling method demonstration testing
consisted of coupons of the same porous material or permeable coating and dimensions as the
associated test coupons that were not spiked with VX but that were wipe-sampled and analyzed
alongside the test coupons. Additionally, three SCs were prepared to confirm the VX challenge
11
-------
EPA/600/R-20/May 2020
application amount. SCs consisted of a spike of an equal volume of VX (10 |iL) directly into
extraction solvent. As discussed in Section 2.2.2.2, when spiked, VX was applied to the inside
surface of the glass SC jar (rather than submerging the pipette tip into the solvent).
The coupon wipe-sampling method would be deemed acceptable for use during subsequent
physical removal efficacy evaluations if the mean wipe-sampling recoveries from stainless steel
were within the range of 70% to 120% of the mean of the SC results with a CoV between
replicates of less than 30%.
Data generated during previous studies [11] have demonstrated difficulty in recovering VX
spiked onto concrete surfaces using solvent extraction techniques. Additionally, previous work
[12] has demonstrated that VX spiked onto a paint layer applied to a substrate will absorb into
the paint layer. For these reasons, difficulty in recovering VX from the surface of concrete,
limestone, painted steel, and painted hardwood coupons via wipe sampling as described above
was anticipated. Thus, only recovery from stainless steel as described above was considered
when determining the effectiveness of the wipe sampling method for use during this project.
Results of wipe-sampling method demonstration testing are provided in Section 3.1.3.
2.1.1.4. Sampling for Residual VX in Generated Waste
As discussed in Section 2.1.1.1, the waste samples generated during the physical removal
technology functionality assessments were retained for use during testing to demonstrate the
methods that were used to recover VX from the wastes via solvent extraction. As identified in
Table 1, such samples included:
• Cutting dust generated from application of the core sampling approach to excise layer
samples from sealed concrete and limestone cores.
• The permeable coatings (paint and primer) removed from painted low-carbon steel and
painted hardwood though application of the chemical stripper.
A method similar to that used to evaluate solvent extraction of VX from ground sealed concrete
and limestone material was used to evaluate extraction of VX from the cutting dust generated
during application of the core sampling method. Concrete and limestone cutting dust samples in
glass jars (60 mL volume) were spiked with 2 |iL of VX according to procedures described in
Section 2.2.2.2, and the VX was allowed to dwell for either 30 minutes or for 24 hours.
Following the dwell period, the cutting dust in each jar was extracted with solvent according to
procedures described in Section 2.2.5 (solvent was added directly to the jars containing the dust
samples).
As during the ground material solvent extraction method demonstration testing, 3 mm glass
beads were used as a control/reference material (approximately 2.1 g of beads per sample). As
discussed in Section 2.1.1.2, recovery from glass was intended to indicate the amount of VX lost
to evaporation during the dwell period, and comparison of the recoveries obtained from sealed
concrete and limestone cutting dust to the recovery obtained from the glass beads would be
12
-------
EPA/600/R-20/May 2020
indicative of the amount of VX lost to entrapment within or degradation by the concrete and
limestone matrices.
Also, stripped permeable coatings (paint/primer) collected from application of chemical stripper
were retained for use during testing to demonstrate that any VX present in the coatings could be
accurately quantified via LC-MS/MS analysis. Stripped coatings collected following application
of chemical stripper to the painted steel and painted hardwood materials were spiked with 2 |iL
of VX (spiked directly onto the stripped coatings sample). VX was allowed to dwell on the
stripped coatings samples for either 30 minutes or for 24 hours. Following the dwell period, the
spiked coatings samples were extracted in solvent. Extracts were then analyzed via LC-MS/MS.
Analysis results were evaluated to ensure that the spiked VX could be adequately recovered from
the coatings and that no interferences were present in the sample matrices from either the
paint/primer or residual chemical stripper that would negatively affect analyses.
Table 4 provides the matrix for waste sampling method demonstration testing. As during coupon
sample solvent extraction and wipe sampling method demonstration testing, IPA was evaluated
as the extraction solvent for recovery of VX from the cutting dust and stripped coatings wastes
generated during physical removal efficacy testing.
Table 4. Waste Sampling Method Demonstration Test Matrix
VX VX
Material
Sample Form A
Sample Type
Challenge
GiL)
Dwell
Period
Replicates
Sealed concrete
Cutting dust
Test sample
2
30 min
3
Sealed concrete
Cutting dust
Test sample
2
24 hours
3
Sealed concrete
Cutting dust
Procedural blank
NA
24 hours
1
Limestone
Cutting dust
Test sample
2
30 min
3
Limestone
Cutting dust
Test sample
2
24 hours
3
Limestone
Cutting dust
Procedural blank
NA
24 hours
1
Glass beads
Loose beads (approximately 2.1 g)
Test sample (control material)
2
30 min
3
Glass beads
Loose beads (approximately 2.1 g)
Test sample (control material)
2
24 hours
3
Glass beads
Loose beads (approximately 2.1 g)
Procedural blank
NA
24 hours
1
Painted steel
Stripped paint
Test sample
2
30 min
3
Painted steel
Stripped paint
Test sample
2
24 hours
3
Painted steel
Stripped paint
Procedural blank
NA
24 hours
1
Painted hardwood
Stripped paint
Test sample
2
30 min
3
Painted hardwood
Stripped paint
Test sample
2
24 hours
3
Painted hardwood
Stripped paint
Procedural blank
NA
24 hours
1
A Cutting dust samples and stripped paint samples retained following functionality assessments (refer to Section 2.1.1.1). Cutting
dust masses collected are provided in Section 3.1.1.1.
In addition to the test samples and procedural blanks identified in Table 4, three SCs were
prepared to confirm the VX challenge application amount. SCs consisted of a spike of an equal
volume of VX (2 |iL) directly into extraction solvent. As discussed in Section 2.2.2.2, when
spiked, VX was applied to the inside surface of the glass SC jar (rather than submerging the
pipette tip into the solvent).
Results of the waste sampling method demonstration testing are provided in Section 3.1.4.
13
-------
EPA/600/R-20/May 2020
As indicated in Table 1, application of the vapor-phase solvent extraction technology to the
porous materials or permeable coatings was anticipated to produce condensed extraction solvent
as a waste product. Retention of the solvent collected during the functionality assessment of
vapor-phase solvent extraction on sealed concrete, limestone, painted steel, and painted
hardwood was planned. Since vapor-phase solvent extraction was ultimately excluded from the
physical removal efficacy test matrix (based on difficulties experienced with setup and
application of the technology during functionality assessments), this portion of waste sampling
method demonstration testing was not conducted.
2.1.1.5. VXDepth Penetration Assessment
The depths to which VX would penetrate each of the porous materials selected for evaluation
during this project (sealed concrete and limestone) were evaluated. To determine these depths,
the core sampling approach (an approach independent of the selected physical removal
technologies) was developed and used for dissection of the porous materials to quantify the
extent of VX penetration as a function of depth during the VX depth penetration assessment.
Section 2.2.3.1 provides a description of the equipment and procedures that were used for the
core sampling approach. The VX depth penetration assessment using the core sampling approach
took place after functionality of the approach was assessed, and solvent extraction and wipe-
sampling method demonstration testing had been completed, so that the VX depth penetration
assessment results could be evaluated in light of the determined solvent extraction and wipe-
sampling recovery efficiencies from the sealed concrete and limestone porous materials.
VX (10 |iL) was applied to sealed concrete and limestone core samples according to procedures
described in Section 2.2.2.2. VX was applied in the center of the top surface of the cores.
Following application, the VX was allowed to dwell on the surface of the cores according to
procedures described in Section 2.2.2.3 and penetrate the material cores over the course of 24
hours.
Following the dwell period, the top (spiked) surfaces of the porous material cores were wipe-
sampled according to the procedures demonstrated during wipe sampling methods development
testing. Wipe samples were obtained from the entire area of the top surface of each core that was
initially contaminated with VX.
Following wipe sampling, the core sampling approach was applied. The contaminated core
samples were dissected into discrete layer samples. The individual layer samples were extracted
separately in solvent, and each layer extract was analyzed via LC-MS/MS for VX. Analysis
results were used to quantify the mass of VX that penetrated the material as a function of depth
(in increments of approximately 0.25 in., based on VX recovery from each layer sample).
Table 5 provides the test matrix for the VX depth penetration assessment. Each core test sample
replicate and procedural blank for each material type was sampled via wipe sampling following
the VX dwell period (top, spiked surface), then dissected into five (5) discrete layer samples. For
each discrete layer sample excised from the core (except for the lst/topmost layer), the associated
14
-------
EPA/600/R-20/May 2020
cutting dust was collected and extracted with solvent and cutting dust extracts were analyzed via
LC-MS/MS alongside the layer sample extracts.
Table 5. VX Depth Penetration Assessment Test Matrix
Material
Sample Type
Sample Form
VX Challenge
(jiL)
Dwell
Period
Replicates
Sealed concrete
Test sample
Material core A
10
24 hours
3
Sealed concrete
Procedural blank
Material core A
NA
24 hours
1
Limestone
Test sample
Material core A
10
24 hours
3
Limestone
Procedural blank
Material core A
NA
24 hours
1
A Dimensions: 1.5 in. diameter, 2.0 in. thickness
In addition to the test samples and procedural blanks identified in Table 5, three SCs were
prepared to confirm the VX challenge application amount. SCs consisted of a spike of an equal
volume of VX (10 |iL) directly into extraction solvent. When spiked, VX was applied to the
inside surface of the glass SC jar (rather than submerging the pipette tip into the solvent).
Results of the VX depth penetration assessment are provided in Section 3.1.5.
2.1.2. Physical Removal Efficacy Evaluation
Coupons of the porous materials or permeable coatings (coatings applied to the low-carbon steel
and hardwood substrates; coupon dimensions provided in Table 7) were contaminated with VX.
VX was applied as a single 10-|iL droplet. Following application, VX was allowed to dwell on
the coupon surface for 24 hours to penetrate the porous material or permeable coating.
Observations of the spreading/soaking/etc. nature of the VX after application were recorded on
the test parameter control sheet (TPCS; refer to Section 4.4).
Following the dwell period, wipe samples were collected from each coupon surface prior to
application of the physical removal methods to quantify residual VX surface hazard on the
material (transferable hazard that had not absorbed/permeated/penetrated into the material
during the dwell period). Wipe samples were obtained using the procedures described in Section
2.2.4 that had been successfully demonstrated as described in Section 2.1.1.3.
Following wipe sampling, the selected physical removal technologies were applied as described
in Sections 2.2.3.2 and 2.2.3.3 to obtain samples from the coupons (the form/nature of samples
obtained depended on the material and physical removal technology applied). Generally, coupon
samples (either ground material samples or post-technology application wipe samples) were
collected from each coupon in the area below the location of VX contamination. Coupon
samples were extracted in solvent and analyzed via LC-MS/MS to quantify VX in the samples.
Wastes generated during removal (i.e., stripped coating layers) were collected, sampled and
analyzed via LC-MS/MS to quantify VX in the waste. Wipe samples were collected from the
parts of the removal technologies/equipment (as applicable, e.g., grinding wheel and deflector
shield of the grinding approach) that contact the contaminated areas of the coupons, and the
wipes were extracted in solvent and analyzed for VX.
15
-------
EPA/600/R-20/May 2020
Four (4) tests were executed to complete the physical removal efficacy test matrix. Each
individual test involved evaluation of a single physical removal method and coupon material
type, to avoid cross contamination of samples by any migrating chemicals or dusts and debris
created during application of the physical removal technologies that could not be completely
captured (for waste analysis). Table 6 summarizes and provides information for the four tests
that were performed, including indication of the types of samples that were collected from
replicate test coupons during each test.
Table 6. Physical Removal Efficacy Test Matrix
Pre-
0.25 in.
Post-
Material
Type
Removal
Method
Sample Type
VX
Challenge
Application
Wipe
Sample
Depth
Layer
Samples
Application
Wipe
Sample
Waste
Removal
Technology
Replicates
Sealed
Grinding
Test sample
¦/
¦/
¦/
~
3
Concrete
Procedural blank
¦/
¦/
¦/
1
Limestone
Grinding
Test sample
¦/
¦/
¦/
¦/
3
Procedural blank
¦/
¦/
¦/
1
Painted
Steel
Chemical
Test sample
¦/
¦/
¦/
¦/
3
Stripping
Procedural blank
¦/
¦/
¦/
1
None
Positive Control
¦/
¦/
¦/
1
Painted
Hardwood
Chemical
Test sample
¦/
¦/
¦/
¦/
3
Stripping
Procedural blank
¦/
¦/
¦/
1
None
Positive Control
¦/
¦/
¦/
1
Positive controls included in tests of the chemical stripping technology consisted of painted steel
and hardwood material coupons that were contaminated with VX and sampled following the
dwell period alongside the test coupons (via wipe sampling), but to which the chemical stripper
was not applied.
Procedural blanks consisted of coupons that were tested alongside the test coupons, including the
dwell, physical removal technology application, and sampling and analysis steps, but that were
not contaminated with VX.
Three SC samples were prepared as well during each test by spiking VX (same volume as that
applied to test coupons) directly into solvent to confirm the contamination amount applied to
coupons.
Results from the four physical removal efficacy tests that were performed are provided in
Sections 3.2 and 3.3.
2.1.3. Waste Generation Assessment
In addition to the quantitative measurements discussed in Sections 2.1.1.4, 2.2.3.1, and 2.2.3.3
(recovery of VX from core sample cutting dusts and stripped paint), the wastes generated during
physical removal efficacy testing were characterized by nature (chunks, dust, stripped materials,
etc.) and volume/weight.
2.1.4. Damage Extent Assessment
During the physical removal technology functionality assessments and following application of
the technologies during physical removal efficacy testing, the appearance of coupons was
16
-------
EPA/600/R-20/May 2020
assessed and representative photographs of the damage portions of the materials (portions to
which the technologies were applied) were taken.
2.2. Experimental Methods and Materials
Experimental methods and materials used to conduct the testing described in Sections 2.1.1
through 2.1.4 are described in the subsections below.
2.2.1. Porous Materials and Permeable Coatings
Porous material and permeable coating information is provided in Table 7. Information on the
stainless steel and glass beads used as control samples during methods demonstration tests is
provided in Table 7 also.
Coupons of the porous materials and permeable coatings used during the grinding and chemical
stripper physical removal efficacy tests, respectively, were prepared and used at the dimensions
provided in Table 7. Core samples of the two porous materials used during the VX depth
penetration test were excised from coupons prior to testing. For both porous materials, core
samples were 1.5-in. -diameter cylinders, with height dependent upon the original material
coupon thickness. Limestone was not sealed prior to use. Concrete coupons (5.75 in. length by
5.75 in. width by 2 in. thick) were fabricated and allowed to cure for a period of five (5) days
prior to sealing.
Table 7. Porous Materials and Permeable Coatings
Material
Description
Supplier
Coupon Dimensions
Preparation
Sealed
concrete
Water repellent (Siloxane PD,
PROSOCO, Lawrence, KS) sealed
concrete (5 parts sand; 2 parts cement
(Buzzi Unicem LISA, Greencastle Plant,
Greencastle, IN); custom preparation,
sealed on all surfaces
Snowden Concrete
Products, LLC,
Cincinnati, OH
5.75 in. length
5.75 in. width
2 in.thick
Cleaned with dry air to
remove loose dust
Limestone
Limestone pavers/cobbles
Roby Monuments
London, OH
7.5 in. length
7.5 in. width
2.25 in. thick
Scrubbed with a water-wetted
brush to remove grime/debris;
dried in an oven at approx.
30.5°C for at least 24 hours
Painted steel
Low-Carbon Steel
McMaster-Carr
Aurora, OH
(6544K13)
7.5 in. length
7.5 in. width
22-gauge thickness
(plus coating layer
thickness)
1. Cut coupon to size
2. Applied coat of primer
3. Allowed to dry
4. Applied coat of paint
5. Allowed to dry
6. Cleaned using dry air to
remove debris
LATEX White Interior/Exterior Multi-
Surface Primer, Sealer, and Stain Blocker
Home Depot
(100096395)
High Performance Protective Enamel
Gloss White Oil-Based Interior/Exterior
Metal Paint
Home Depot
(202067206)
Painted
Hardwood
Red Oak Hardwood
Lowe's
(1054)
5.5 in. length
5.5 in. width
0.75 in. thickness
(plus coating layer
thickness)
1. Cut coupon to size
2. Applied coat of primer
3. Allowed to dry
4. Applied coat of paint
5. Allowed to dry
6. Cleaned using dry air to
remove debris
LATEX White Interior/Exterior Multi-
Surface Primer, Sealer, and Stain Blocker
Home Depot
(100096395)
High Performance Protective Enamel
Gloss White Oil-Based Interior/Exterior
Metal Paint
Home Depot
(202067206)
Stainless steel
(control)
Type 304 stainless steel disks
Adept Products, Inc.
West Jefferson. OH
1.375 in. dia. disk
24-gauge thickness
None
Glass beads
(control)
Borosilicate glass beads
Fisher Scientific,
Pittsburgh. PA
(10-310-1)
3-mm dia. beads
None
17
-------
EPA/600/R-20/May 2020
2.2.2. Application of CWA
2.2.2.1. CWA
All quantities of VX used during this project were synthesized at Battelle's HMRC under
Chemical Weapons Convention program guidelines, with accountability through the U.S. AMC.
All VX used originated from the same synthesis lot. All VX was stored in the HMRC CWA
vault in accordance with HMRC security and CWA storage policies until needed for testing.
Once prior to use during testing (prior to both methods demonstration and physical removal
efficacy testing), purity of the VX was measured. Target purity for VX was > 90%. A VX purity
sample was prepared (900 |ag/m L concentration for VX) from the supply of neat VX available
for use on the project. The purity sample was analyzed by gas chromatograph (GC)/flame
ionization detector (FID) to determine the relative abundance of VX as determined by peak area
and reported as percent purity. Impurities and composition can influence VX degradation rates.
Solvent blanks were used to correct for possible solvent contaminants. GC/FID method
parameters for purity sample analysis are provided in Table 8.
Table 8. GC/FID VX Purity Sample Analysis Method Parameters
Parameter
Description A
Instrument
Hewlett Packard Model HP 6890 Gas Chromatograph equipped with FID
and Model 7683 Automatic Sampler
Data System
Chromeleon 7 (Thermo Electron Corporation)
Column
Rtx-5 30 m x 0.25 |im x 0.25 mm or equivalent
Carrier Gas Flow Rate
1.5 mL/min
Injection Volume
0.10 nL
Column Temp
40 °C initial temperature, hold 2 min, 20 °C/min to 280 °C. hold 5 min
Injection Temperature
Cool on column (track oven temperature)
FID Temp
250 °C
Table 9 provides purity information for the VX used during each test.
Table 9. VX Purity by Test
Purity
Tests Used
92.7%
Solvent extraction, wipe-sampling, and waste sampling methods demonstration tests
93.1%
VX depth penetration test, all physical removal efficacy tests
2.2.2.2. Coupon Spiking
Coupons, core samples, and core layer samples were inspected visually prior to contamination
with neat VX to ensure any samples with surface anomalies were not used. Neat VX was applied
to each designated sample as a single 2-|iL droplet using a l-|iL- to 10-|iL-range Gilson™
MICROMAN™ E positive displacement pipette (FD10001G, Fisher Scientific, Pittsburgh, PA,
or similar), or a single 10-|iL droplet using a 3-|iL- to 25-|iL-range Gilson™ MICROMAN™ E
positive displacement pipette (FD10002G, Fisher Scientific, Pittsburgh, PA, or similar). VX
18
-------
EPA/600/R-20/May 2020
droplets were generally placed in the center of samples (or in each coupon "quadrant" during the
wipe sampling method demonstration test).
When spiking ground concrete and limestone coupon samples during coupon sample solvent
extraction method demonstration testing, VX was applied to the ground material samples by
submerging the tip of the positive displacement pipette into the ground material, ejecting VX
into the material while the tip was submerged, and then using the pipette tip to "stir" the spiked
VX into the material. This approach was intended to ensure adequate mixing and dispersal of the
VX throughout the ground material. However, we observed that the dispensed liquid VX caused
"clumping" of ground material and the "clumped" ground material would stick to the pipette tip,
possibly preventing adequate dispersal and mixing of VX throughout the material. Refer to the
solvent extraction methods demonstration results in Section 3.1.2 for further discussion.
Following spiking each sample, the pipette tip used to spike/stir each sample was discarded (i.e.,
a new, unused tip was used to spike/stir each ground material sample).
A similar approach was initially used for spiking and mixing VX into the sealed concrete and
limestone cutting dust during the waste sampling methods demonstration test as that used for the
ground materials (VX was applied to the cutting dust by submerging the tip of the positive
displacement pipette into the dust, ejecting VX while the tip was submerged, and then using the
pipette tip to "stir" the spiked VX into the dust). The same "clumping" behavior was observed
after the 1st cutting dust sample was spiked (i.e., liquid VX caused clumping of the cutting dust,
and portions of the "clumped" dust sample stuck to the pipette tip; 1st sealed concrete dust
sample; refer to the EPATO20-MD3 TPCS provided in Attachment B), so the spiked/"clumped"
sample was discarded and a new sample was prepared and spiked using an alternative method.
For the alternative cutting dust spiking method, the jar containing each dust sample was tilted to
collect the dust to one side of the jar, exposing the bottom of the glass jar. VX was then applied
to the bottom of the glass jar, and the jar was then closed and swirled by hand to mix the cutting
dust and liquid VX droplet. The remaining sealed concrete and limestone cutting dust samples
were then spiked using this alternative approach. Following spiking each sample, the pipette tip
was discarded (i.e., a new, unused tip was used to spike each dust sample).
SC samples were prepared by delivering the same quantity of VX (2 or 10 |iL) directly into 20
mL of extraction solvent in a 60 mL glass extraction jar (same used for sample extraction), rather
than onto a sample surface. When spiked, VX was applied to the inside surface of the glass SC
jar (rather than submerging the pipette tip into the solvent). When extracted, SCs were processed
in a manner similar to wipe, coupon, or waste sample extracts (that is, SCs were sonicated and
aliquoted as described for test sample extracts in Section 2.2.5).
2.2.2.3. CWA Dwell Period
Following application of VX, the contaminated samples were allowed to remain undisturbed for
a 30-minute or 24-hour CWA dwell period (depending on the test). During the dwell period, the
samples were subjected to the ambient atmosphere within the test hood. Samples were left
19
-------
EPA/600/R-20/May 2020
uncovered during the dwell period. Temperature and relative humidity (RH) of the sample
environment within the hood were monitored and recorded but not controlled. RH was not
expected to have an impact on evaporation or penetration of VX into the materials. Temperature
and RH conditions within the hood were measured and recorded using a HOBO UX100
Datalogger (UX100-003, Onset® Computer Corporation, Bourne, MA) on each day of testing.
Environmental data from each test are available as Attachment A.
2.2.3. Description and Application of Physical Removal Technologies
2.2.3.1. Core Sampling Approach
A method independent of the selected physical removal technologies was developed and used for
dissection of porous materials (sealed concrete and limestone) to quantify the extent of VX
penetration as a function of depth during the VX depth penetration assessment performed during
methods demonstration.
A cordless drill/driver (1001592743, Home Depot, Columbus, OH, or similar) equipped with a
1.625 in.-diameter carbide hole saw (301697684, Home Depot, Columbus, OH, or similar) was
used to excise 1.5 in.-diameter full-depth cylindrical core samples from coupons of each porous
material. During the VX depth penetration assessment, the core samples were contaminated in
the center of the top surface with VX, and the VX was allowed to dwell on the core surface for
24 hours (cores were oriented upright), after which the top (contaminated) surface of the core
was wipe-sampled. Following wipe-sampling, the contaminated core was inserted into a holder
with the bottom/uncontaminated surface of the core sample facing toward the front of the holder
(Figure 2). A cordless reciprocating saw (1002338813, Home Depot, Columbus, OH, or similar)
equipped with a diamond-tipped saw blade (1000683506, Home Depot, Columbus, OH, or
similar) was then used to cut the core sample into discrete layer samples at (target) 0.25 in. depth
increments. Layer samples were cut from the core beginning from the bottom/uncontaminated
side of the core and progressing toward the top/contaminated side. The layer samples from the
core were then extracted separately with solvent and layer sample extracts were analyzed via LC-
MS/MS for VX. Cutting dust generated during use of the reciprocating saw (to dissect the core
into individual slices) was collected in a tray placed underneath the core and retained for solvent
extraction and analysis via LC-MS/MS for VX. No air monitoring was conducted to collect
(contaminated) fine particulate matter which may have spread across a larger area.
20
-------
EPA/600/R-20/May 2020
Figure 2. Core Sample Holder with Dust Collection Tray
(uncontaininated surface of core is visible)
Figure 3 illustrates the overall core sampling approach used to obtain discrete depth layer
samples of each porous material core to evaluate VX penetration into the material.
CWA applied and allowed to dwell
Figure 3. Porous Material Core Sampling Approach
Figure 4 illustrates the approach used to dissect each individual material core to harvest the
discrete layer samples.
21
-------
EPA/600/R-20/May 2020
Top surface of core
(VX challenge and wipe sampling)
Layer 1
"Slicing"
progression
Layer 2 cutting dust
Layer 2
Layer 3 cutting dust
Layer 3
Layer 4 cutting dust
Layer 4
Layer 5 cutting dust
Layer 5
Figure 4. Core Sampling Dissection
Once the core had been cut into discrete layer samples, a wipe sample was obtained from the
reciprocating saw blade. The wipe covered the full surface of the blade. The blade was then
replaced prior to cutting the next core sample (i.e., each blade was used to cut only a single core
into layer samples).
2.2.3.2. Grinding
Grinding is a commonly used approach for surface layer removal in which coarse-grained
abrasives in the form of grinding wheels or surfacing disks are applied to a material. The rotating
wheel abrades the material, grinding it and removing surface layers. During this project, an angle
grinder (DeWalt 11 amp corded, 4.5 in. small angle grinder, 1001672186, Home Depot,
Columbus, OH, or similar) equipped with a fine-grit diamond grinding wheel (203061023, Home
Depot, Columbus, OH, or similar) was used.
Following application of VX, the 24-hour dwell period, and the initial coupon wipe sample, the
grinder was applied to the coupon surface to remove material to a target depth of 0.25 in.
Removal to this depth was achieved using a stepwise approach, wherein the grinder was applied
to remove material across a wide area of the coupon surface to a depth visually less than 0.25 in.
After the first "pass", the achieved removal depth was measured to gauge the additional material
that needed to be removed via reapplication of the grinder (a second "pass") to reach a final 0.25
in. depth. Once the final depth of approx. 0.25 in. had been achieved across the removed area of
the coupon surface, the aggregate ground material removed was collected and retained for
solvent extraction and analysis via LC-MS/MS to quantify VX in the removed material. Prior to
22
-------
EPA/600/R-20/May 2020
solvent extraction, the mass of ground material removed in the depth layer sample was
determined gravimetrically.
Following removal of the first 0.25 in.-depth layer sample, the grinding wheel was sampled via
wipe sampling, and the wipe was extracted with solvent and analyzed via LC-MS/MS to assess
VX contamination transferred to the grinding wheel. The wipe sample obtained from the wheel
covered the full surface of the wheel that contacted the coupon during grinding.
The above approach was then repeated to remove an additional (second) 0.25 in. depth layer
sample (thus to a cumulative target depth of 0.5 in. from the original coupon surface). The area
of the coupon over which the grinder was applied to collect the second (and subsequent) 0.25 in.-
depth layer sample was slightly less than the area of the first (or previous) 0.25 in.-depth layer
sample. Following collection of each successive depth layer sample after the second, the
grinding wheel was wiped using a solvent-soaked wipe, but the wipe was not extracted with
solvent and analyzed (i.e., the wipe was used only to clean the wheel between collection of the
second and third, and subsequent, depth layer samples).
The general approach for collecting 0.25 in.-depth layer samples from the porous material
coupons is illustrated in Figure 5.
1st "pass" removes
material to a depth
less than 0.25".
2nd "pass" (following
measurement of 1st
"pass" depth) removes
additional material.
3rd (as necessary)
"pass" to achieve
0.25" removal depth.
Additional 0.25" depth layer samples
are removed. The area of each across
the coupon surface is decreased
slightly with each subsequent layer,
so that depth layers are discernable.
Figure 5. Approach for Depth Layer Sample Collection via Grinding
Coupon setup for application of the grinder involved orientation of the coupon upright on a side
within the test chamber using an aluminum easel. Figure 6 depicts a sealed concrete coupon on
the aluminum easel.
23
-------
EPA/600/R-20/May 2020
Figure 6. Concrete Coupon on Aluminum Easel
Figure 6 also shows that a deflector shield was positioned to the right of the coupon. When the
grinder was applied, ground material removed from the surface of the coupon was thrown by the
grinding wheel against the deflector shield and directed downward into an aluminum collection
pan. Once a depth layer sample of 0.25 in. had been removed, the ground material collected in
the aluminum collection pan was extracted with solvent according to procedures described in
Section 2.2.5, and the extract was analyzed via LC-MS/MS for VX. Similar to the grinding
wheel, the inside surface of the deflector shield that was contacted by removed ground material
was sampled via wipe sampling following removal of the first depth layer sample, and the wipe
was extracted with solvent and analyzed via LC-MS/MS. The deflector shield was also wiped
between collection of the second and third, and subsequent, depth layer samples, but these wipes
were not extracted and analyzed for VX (i.e., wipe sampling was performed only to clean the
inner surface of the deflector shield between replicates to reduce/eliminate the potential for
cross-contamination).
2.2.3.3. Chemical Stripping
Klean-Strip® KS-3 Premium finish/paint stripper (100144685, Home Depot, Columbus, OH, or
similar) is a dichloromethane-based stripper. It is a thickened semi-paste that can be applied via
brushing and is intended to cling to vertical surfaces without running or dripping.
24
-------
EPA/600/R-20/May 2020
Following application of VX, the 24-hour dwell period, and the initial coupon wipe sample, 1
mL of the chemical stripper was applied directly to the permeable coating (primer/paint) on
coupons using an Eppendorf Repeater M4 Pipette (14-287-150, Fisher Scientific, Pittsburgh,
PA). The stripper was then manually spread across an area with diameter of approx. 1.375 in.
using a V2 in.-width brush (7237T84, McMaster-Carr, Aurora, OH), centered over the area of the
coupon that was contaminated with VX (or in the center of the coupon in the case of procedural
blanks). Following application, the stripper was allowed to dwell on the coating for 30 minutes
(manufacturer-recommended contact time). After 30 minutes, the coating was visually assessed.
If the coating did not appear to be detached from the steel or hardwood material (as evidenced by
a "bubbled" or "swollen" appearance of the paint/primer), the stripper was allowed to dwell and
react with the coating for an additional 15 minutes (total 45-minute stripper dwell period). The
total chemical stripper residence time on the coating surface was recorded on the TPCS.
Once the coating layer appeared to be visibly detached from the steel or hardwood material, the
coating was stripped from the material using a 2 in.-width plastic joint knife (3546A421,
McMaster-Carr, Aurora, OH). Generally, the coating was removed as a single, solid piece (a
"bubbled" or "swollen" section of degraded paint/primer) that was physically scraped away from
the material. The stripped coating (potentially containing VX) was collected, extracted with
solvent, and analyzed for VX via LC-MS/MS.
2.2.3.4. Vapor-Phase Solvent Extraction
Vapor-phase solvent extraction is a physical removal technology that involves vaporization of an
organic solvent with a low boiling point so that the vaporized solvent will circulate within a
building (or similar enclosed space). As they are circulated, the vapors permeate into porous
building materials, where they condense, solubilize contaminants, and then diffuse outward.
Condensed solvent containing the removed contaminants is then decontaminated using a separate
approach.
As discussed in Section 1.2, vapor-phase solvent extraction was selected following the literature
search for evaluation for efficacy in removal of VX contamination from both porous materials
and permeable coatings. As vapor-phase solvent extraction is primarily designed for application
in buildings and similar large, enclosed spaces, a modified, laboratory-scaled approach was
conceptualized and intended for use during this project. As discussed in Section 2.1.1.1, the
conceptualized apparatus and approach for application of vapor-phase solvent extraction were
evaluated during the technology functionality assessments conducted during the methods
demonstration phase of the project. The apparatus and test stand setup that were evaluated for
application of the vapor-phase solvent extraction physical removal technology are depicted in
Figure 7.
25
-------
EPA/600/R-20/May 2020
Temp.
Solvent controller
Clean, dried house air (regulated to 20 pounds per square inch [psi]) was supplied to a mass flow
controller that set the flow rate of the airstream to approximately 300 mL/min. The air stream
coupled with an infusion (via a syringe pump) of solvent (acetone, A929-4, Fisher Scientific,
Pittsburgh, PA) at a tee fitting. Heat tape (VV-03106-32, Cole Parmer, Vernon Hills, IL) was
used to heat the tee fitting and air supply tubing to vaporize the acetone (heated to above the
boiling point of the solvent). The vaporized acetone would then be carried by the airstream
though an inline static mixer (3530K43, McMaster-Carr, Aurora, OH).
It was planned that following application of VX, the 24-hour dwell period, and the initial coupon
wipe sample, the test or control coupon would be set into a test stand that would orient the
coupon in an upright position. The solvent vapor stream from the static mixer would then be
directed onto the contaminated area of the coupon (or the center of the coupon in the case of
procedural blanks). The solvent vapor stream would be allowed to impact the coupon surface for
a period of 2.5 hours. The syringe pump would be set to infuse solvent into the airstream at a rate
of 150 |iL/min, Thus, the total amount of solvent vaporized would be approx. 22.5 mL.
As the solvent vapor stream impacted the coupon surface, it was expected that solvent would
either: (1) migrate into the porous materials, (2) be deflected off the coupon and dissipate into
the air, or (3) condense on the coupon surface, run down, and drip from the bottom edge of the
coupon. Solvent vapor that deflected off the coupon surface and into the air would not be
collected. Solvent that condensed on the coupon surface and ran downward would be collected
for analysis. The design of the test stand included channels that would direct condensed solvent
into a graduated glass collection vessel below the coupon. The collection vessel would be kept
on dry ice to keep the condensed, collected solvent cold to minimize evaporation during the 2.5-
hour vapor-phase solvent extraction application. It was suspected that solvent that penetrated
(spiked) porous materials would either carry penetrated CWA farther into the coupon or would
allow for migration of penetrated CWA out of the coupon. It was assumed that CWA that
26
-------
EPA/600/R-20/May 2020
migrated back out of the coupon through the penetrated solvent would be carried by condensed
solvent running down the surface of the coupon into the collection vessel below. At the end of
2.5-hour vapor-phase solvent extraction application, the volume of condensed solvent collected
in the vessel would be recorded. The solvent would then be analyzed via LC-MS/MS for VX.
Following solvent vapor treatment, the core sampling approach would be used to dissect and
sample the treated porous materials (sealed concrete and limestone). Permeable coatings (painted
steel and painted hardwood) would be wipe-sampled a second time, following the first wipe
collected prior to solvent vapor treatment.
During the technology functionality assessment of the vapor-phase solvent extraction apparatus,
operating conditions/parameters to produce a viable solvent vapor stream could not be
determined using the apparatus as conceptualized (setup as depicted in Figure 7). Initial revisions
of the apparatus setup were unsuccessful in generating a useable solvent vapor stream. Budget
and schedule limitations of the project then precluded further development of the technology, so
vapor-phase solvent extraction was not carried forward into further methods demonstration test
phases or into physical removal efficacy testing. Refer to Section 3.1.1.4 for additional
information.
2.2.4. Coupon Surface (Wipe) Sampling for VX
The wipe sampling method used during this project was consistent with a method developed and
used during a previous EPA study [12] to sample transferable/residual CWA on the surface of
materials. Leveraging the wipe-sampling methods development testing conducted during the
previous work, the wipe sampling method used during this project included the following details:
• Wipes were lint-free 2 in. x 2 in. (5 centimeters [cm] x 5 cm) four-ply rayon/polyester
blend (gauze) sponges (22-037-921, Fisher Scientific, Pittsburgh, PA).
• The same solvent selected for coupon sample and waste sample extraction (IPA) was used
as the wipe wetting and wipe extraction solvent.
• Wipes were wetted with 1.5 mL of solvent. This volume of solvent added to this particular
wipe was determined to be an amount that is approximately half-saturating for the wipe, as
determined gravimetrically by weighing three wipes before and after soaking the wipes in
solvent (half of the amount of solvent remaining on the wipe 30 seconds after immersion in
solvent and hanging vertically to allow excess solvent to drip off) [12],
• Each coupon was wiped using four horizontal and four vertical strokes (using the same
wipe) over an area of no larger than approximately 100 square centimeters (cm2; adjusted
as necessary based on wipe sampling a coupon or a core sample) centered on the area of
the coupon/core that was contaminated with VX. Adequate overlap between strokes
occurred so that the entire 100 cm2 (or as adjusted) area was wiped. The leading edge of the
wipe was maintained between passes. The wipe was folded, as necessary, for
manageability during wipe sampling. Figure 8 depicts the wipe pattern that was used.
27
-------
EPA/600/R-20/May 2020
Pass 1 Pass 2
Stroke 1 Stroke 2 Stroke 3 Stroke 4
Figure 8. Wipe Pattern
Following collection, wipes were extracted in solvent in the same manner as coupon samples and
waste samples, using the same solvent as that used to wet the wipes IPA. Wipe extracts were
then analyzed for VX by LC-MS/MS.
2.2.5. Extraction of VX from Wipe, Coupon, and Waste Samples
All core layer samples, wipe samples, and waste samples (cutting dust and stripped paint) were
extracted by placing each into a separate 60-mL glass jar (05-719-257, Fisher Scientific,
Pittsburgh, PA, or similar) containing extraction solvent (IPA). Ground material samples were
placed into a 125-mL glass jar (05-719-57, Fisher Scientific, Pittsburgh, PA, or similar). SCs,
wipe samples, core layer samples, core sampling cutting dust, and stripped coating samples were
extracted with 20 mL of solvent. Ground material samples were extracted with 80 mL of solvent.
PTFE disks included during the solvent extraction method demonstration test were extracted
with 10 mL of solvent. These jars and these volumes of solvent were sufficient to submerge all
sample types fully in the extraction solvent.
Following addition of coupon samples, wipe samples, or waste samples to the extraction solvent
within each jar, the jars were swirled by hand for approximately 5-10 seconds and placed into a
sonicator (15-336-103, Fisher Scientific, Pittsburgh, PA, or similar). Extraction jars were
sonicated at 40 - 60 kilohertz for 10 min. Within 30 min of completing this process, aliquots of
approximately 0.5 mL from each extraction jar were transferred to individual analysis sample
vials (21140 [vial], 24670 [cap], Fisher Scientific [Restek Corp.], Hanover Park, IL 60133) and
sealed. Samples that were not analyzed the same day were stored at -20 ± 10 °C.
2.3. Analytical Methods
Coupon, wipe, and waste sample extracts were analyzed using LC-MS/MS to quantify the
amount of residual VX present. An AB SCIEX 5500 triple quadrupole MS (SCIEX,
Framingham, MA) coupled to a Shimadzu 20 XR series LC (Shimadzu, Columbia, MD) was
used for sample analysis. VX was quantitated in sample extracts using a reversed-phase high
performance liquid chromatography (HPLC) method and multiple reaction monitoring (MRM).
MRM provides high specificity and sensitivity and is typically used in quantitative applications.
28
-------
EPA/600/R-20/May 2020
The MRM transition with the best signal-to-noise ratio is usually selected for quantitation. VR
(Russian VX; nominal concentration 0.5 nanograms [ng]/mL) was used as the internal standard
(IS) for quantitation of VX, and was added to calibration standards, controls, and test samples
just prior to LC-MS/MS analysis (specifically, a solution of VR in water was used as the diluent
for dilution of samples prior to analysis). The VR used during this project was a Chemical Agent
Standard Analytical Reference Material (CASARM) obtained from the U.S. Army Combat
Capabilities Development Command (CCDC) Chemical Biological Center. Table 10 provides
the ion transitions that were used for detection and quantitation of VX.
Table 10. Anafyte Ion Transitions
Analyte
Precursor Ion
Product Ion Quantifier
VX
268
128
VR
268
100
The lower limit of quantitation for VX (i.e., the lowest concentration standard used in the
calibration curve) was 0.010 ng/mL, which was above, but near, the instrument detection limit of
the LC-MS/MS.
The concentration of VX in the samples was interpolated using the VX area/IS area ratio and the
regression equation generated from the calibration standards. Samples that were quantitated
below the lowest calibration standard concentration (< 0.010 ng/mL), or displayed area counts
below the area counts of the lowest concentration on the calibration curve, were reported as less
than the lower limit of quantitation (LLOQ; e.g., < 0.010 ng/mL). The less-than-the-LLOQ value
was corrected to account for any sample dilution factor (minimum of 10-fold). Samples that were
quantitated above the highest calibration standard (2.0 ng/mL) were diluted (i.e., a new analytical
sample was prepared from the original extract using a higher dilution factor) and reanalyzed.
Refer to Section 4.2.2 for LC-MS/MS calibration details. All data were reported to two
significant figures. LC-MS/MS parameters that were used for analyses are provided in Table 11.
Table 11. LC-MS/MS Conditions for VX Analysis
Parameter Description
Ionization Mode and Polarity
Electrospray Ionization, Positive Mode
HPLC Column
Restek Allure PFPP, 2.1 x 50 mm, 5 (un, (part no. 9169552)
Column Temperature
Ambient
Mobile Phase
A: 2 mM Formic Acid/2 mM Ammonium Formate in Water
B: 2 mM Formic Acid/2 mM Ammonium Formate in Methanol
Mobile Phase Gradient
Time (minutes)
%B
0.0
20
1.0
20
2.0
100
4.0
100
4.1
20
4.5
20
Flow Rate
Ramp from 0.5 to 0.7 mL/min from 1.0 to 2.0 minutes
Ramp from 0.7 to 0.5 mL/min from 4.0 to 4.1 minutes
29
-------
EPA/600/R-20/May 2020
Parameter
Description
0.5 mL/min for all remaining method segments
Injection Volume
2 - 50 jiL
Run Time
4.5 minutes
Samples (in IP A) were diluted a minimum of 10-fold prior to LC-MS/MS analysis. Higher
dilution factors were used based on the concentration of VX in the samples. Sample dilutions
were performed using calibrated positive displacement pipettes and were documented on the
sample chain of custody (CoC) forms (refer to Section 4.3), the laboratory record book (LRB),
and the raw analysis data files.
2.4. Calculations
Diluted test, control, and blank wipe sample, coupon sample (core layers or ground material),
and waste sample (cutting dust or stripped coating) extract concentrations were provided in units
of ng of VX per mL of extract by the LC-MS/MS Analyst® software (ver. 1.7, SCIEX,
Framingham, MA) through comparison of VX and IS peak areas to the calibration curve. After
correction of extract concentrations for dilutions performed prior to analyses, mass recovered
from the samples via extraction was determined according to Equation 1:
M = C0nCExtXV0lExt
1000 1 '
where:
MassRec = CWA mass recovered from the sample (|ig)
ConcExt = Sample extract concentration provided by the LC-MS/MS software (ng/mL)
Vol Ext = Volume of extraction solvent (mL)
Total mass recovered from test, control, or blank coupons or cores was the sum of the masses
recovered from wipe samples taken from the coupons or cores, from extraction of coupon
samples in solvent (core layers or ground material layers), and from extraction of waste samples
(cutting dust or stripped coating). Multiple wipe and/or coupon samples contributed to the total
mass recovered from coupons or cores (depending on the test; e.g., wipe samples were taken
from the coupon, grinder, and deflector shield, and multiple ground material layer samples were
collected during application of the grinding technology to a single coupon). Total mass was
calculated according to Equation 2:
MassTot = MassRec (wipe) + MCLSSftec (coupon)
+ MCLSSftec (waste) (2)
where:
MassTot = Total CWA mass recovered from the coupon or core (|ig)
MassRec (wipe) = CWA mass recovered from the wipe(s) (|ig)
30
-------
EPA/600/R-20/May 2020
MassRec (coupon) = CWA mass recovered from the coupon sample(s) (|ig)
MassRec (waste) = CWA mass recovered from the waste sample(s) (|ig)
Percent recovery was calculated for each individual test coupon or core according to Equation 3:
Recovery = ( MasSTot ) x 100% (3)
\M asss CAvg )
where:
MassscAvg = Average CWA mass recovered from the SCs (|ig)
MassTot = Total CWA mass recovered from the coupon (|ig)
31
-------
EPA/600/R-20/May 2020
3. RESULTS
3.1. Methods Demonstration
3.1.1. Physical Removal Technology Functionality Assessment Results
3.1.1.1. Core Sampling Approach
The method planned initially for application of the core sampling approach involved
contamination of sealed concrete and limestone coupons with VX (10-iiL droplet spiked at the
center of the coupon) prior to application of the approach. After the 24-hour dwell period, core
samples would then be excised from the coupons underneath the area of the coupon to which VX
was applied. Following excision of the core, the core would be sliced into discrete layer samples
to assess depth of VX penetration into the material. This initially planned approach was
evaluated during the technology functionality assessment of the core sampling approach
(performed prior to testing without VX present)
Method
Limestone (8 in. length by 8 in. width by 2.25 in. thick) or sealed concrete (12 in. length by 12
in. width by 2 in. thick) coupons were placed vertically in a holder inside the test chamber with a
weigh dish placed under the coupon to collect cutting dust (refer to Figure 9). A drill
(1001592743, Home Depot, Columbus, OH) with a 1.5 in.-diameter carbide hole saw bit
(301697689, Home Depot, Columbus, OH) was used to drill a 1.375 in.-diameter core sample
out of the coupon. An infrared thermometer (9248T57, McMaster-Carr, Aurora, OH) was used to
measure the surface temperature of each coupon before and after core excision. The core sample
was then inserted into a nylon block/holder for slicing, with a weigh dish placed underneath to
collect cutting dust (refer to Figure 10). A reciprocating saw (1002338813, Home Depot,
Columbus, OH) with a 6 in. diamond grit saw blade (1000683506, Flome Depot, Columbus, OH)
was used to slice the core into 0.25 in. thick slices. The infrared thermometer was again used to
measure the temperature of the core before and after sli cing. After slicing, the cutting dust
collected in the dish was poured into a sample jar and the mass of dust was determined
gravimetrically. A ruler was used to measure the thickness of each slice.
Figure 9. Core Excision Setup
32
-------
EPA/600/R-20/May 2020
Figure 10. Core Slicing Setup
Results
A core sample was excised from sealed concrete in 22 minutes. Temperature of the sealed
concrete coupon prior to excision of the core was 72.1°F as measured on the top surface of the
coupon. Temperature of the core following excision from the coupon was 93 °F at the top of the
core and 110°F at the bottom of the core.
Difficulty was experienced with starting the coring process as the drill/hole saw had a tendency
to "travel" on the coupon surface until some depth was achieved (the pilot bit typically used with
a hole saw could not be used since during normal testing the area drilled by the pilot bit would be
the VX-contaminated area of the coupon). The coupon was removed from the chamber and a
circular groove/score was pre-drilled into the surface. After approximately 6 minutes of drilling,
the operator became fatigued and the angle of the drill shifted causing a (approximately) 0.25 in.-
thick piece of the core to break off (still i nside the coupon), as depicted in Figure 11.
At that point, a 2 operator continued the coring process. Multiple times during drilling, the
operator lost control of the drill due to ergonomics and positioning and also to the hole saw
becoming seized within the coupon, leading to concerns regarding the safety of appli cati on of the
method within the test chamber on coupons that would be contaminated with VX during physical
removal efficacy testing. By the time the operator finished drilling through the material, the hole
Figure 11. Broken Core
33
-------
EPA/600/R-20/May 2020
saw bit had again become seized within the coupon. The coupon had to be removed from the
chamber to remove the hole saw bit. The excised core was also stuck inside the hole saw bit and
operators had to use a screwdriver to remove the core sample from the hole saw. An attempt was
made to pre-drill/score another coupon but was unsuccessful as the hole saw bit had become dull
from use during the first core sampling.
Given the complications and safety concerns described above, it was decided that the core
sampling approach would not be applied as initially planned during the VX depth penetration
assessment and physical removal efficacy testing (wherein the coupon is spiked with VX prior to
coring, and the hole saw is used to excise a core from the contaminated coupon). Alternatively,
pre-excised core samples would be used; i.e., core samples were obtained from uncontaminated
coupons prior to testing, and the cores would be spiked and treated using the technologies (rather
than spiking coupons and obtaining core samples from the contaminated portion of the coupons
after the dwell period). Because pre-excised cores would be used, the nominal core sample
diameter (for the VX depth penetration assessment) was increased to 1.5 in. (excised using a
1.625 in.-diameter carbide hole saw bit, 301697684, Home Depot, Columbus, OH).
Additional 1.375 in.-diameter core samples were prepared outside the lab for preparation of layer
samples for solvent extraction method demonstration testing. The temperature of the core was
measured during drilling (once with the used, dull hole saw bit for concrete and limestone and
once with a new bit for concrete). An increase in temperature was observed and results are
provi ded in Table 12. Photographs of the core samples excised from limestone and sealed
concrete are provided in Figure 12. Such temperature increase was also observed in other studies
and resulted in thermal degradation of VX [13].
Table 12. Temperature Increase for Core Excision using Drill Press
Start
Temp at % in.
Temp at Vz in.
Temp at 1 in.
Temp at 2
Material
Temp
Depth
Depth
Depth
in. Depth
(°F)
(°F)
(°F)
(°F)
(°F)
Limestone (old bit)
61.7
Not measured
Not measured
151.6
145.3
Concrete (old bit)
67.8
141.0
145.5
196.1
179.1
Concrete (new bit)
68.2
91.1
109.1
174.9
163.7
Figure 12. Cores Excised from Limestone (left) and Sealed Concrete (right)
34
-------
EPA/600/R-20/May 2020
The time required to slice the core samples, temperature changes, and collected cutting dust
masses are provided in Table 13 for limestone and Table 14 for sealed concrete. The thickness of
the resulting slices for limestone and sealed concrete are provided in Table 15. Both materials
were amenable to slicing using the method described in Section 2.2.3.1. Representative
photographs of core sample slices are provided in Figure 13 and associated cutting dust samples
are provided in Figure 14.
Table 13. Limestone Core Slicing Results
Core
Slice
Slicing Time (min)
Start Temp (°F)
End Temp (°F)
Cutting Dust Mass (g)
1
10
72.1
93.5
2.85
2
3
93.5
97.3
2.73
1
3
3
97.3
97.8
2.66
4
2
97.8
107.9
2.18
5
3
107.9
121.8
2.66
6
2
105.1
110.3
2.92
1
4
81.9
98.0
2.41
2
2
2
98.0
104.6
2.07
3
1
104.6
97.8
2.43
Average
2.55
Table 14. Sealed Concrete Core Slicing Results
Core
Slice
Slicing Time (min)
Start Temp (°F)
End Temp (°F)
Cutting Dust Mass (g)
1
5
81.6
114
1
2
2
89
124
8.69 A
3
3
94
120
1
2
78.1
80.8
1.39
2
2
3
80.8
83.8
1.85
3
3
83.8
93.8
1.66
4
2
93.8
105.3
1.90
1
1
82.1
101.3
1.09
3
2
3
76.1
90.3
1.25
3
2
90.3
106.9
1.70
4
3
106.9
117.5
1.66
Average
1.56
A Cutting dust from all three slices was combined (data not used for average calculation).
Table 15. Slice Thickness Results
Rep
Concrete Thickness
Limestone Thickness
(inches)
(inches)
1
0.25
0.25
2
0.25
0.25
3
0.25
0.25
4
0.25
0.25
5
0.25
0.25
6
0.375
0.19
7
0.25
0.31
8
0.25
0.31
9
0.25
0.25
10
0.25
0.25
11
0.50 A
-
12
0.375 A
-
Average
0.29
0.26
A Remaining end of cores
35
-------
EPA/600/R-20/May 2020
Figure 13. Core Slices from Limestone (left) and Sealed Concrete (right)
Figure 14. Cutting Dust from Limestone (left) and Sealed Concrete (right)
3.1.1.2. Grinding
The method planned initially for application of the grinding technology involved use of a
vacuum dust shroud (302674944, Home Depot, Columbus, OH) attached to the grinder to collect
ground material as it was removed from the coupon. This initially planned approach was
evaluated during the technology functionality assessment of the grinding technology (performed
prior to testing without VX present).
Method
Difficulty was experienced with application of the grinding technology using the method initially
planned that incorporated a vacuum dust shroud attached to the grinder to collect the ground
material that was produced. Difficulties experienced included:
36
-------
EPA/600/R-20/May 2020
• Operators were unable to orient the grinder against the coupon as necessary so that the
grinding wheel contacted the material properly, due to the position and shape of the
vacuum shroud attachment.
• When attempts were made to orient the grinder to obtain successively deeper Vi in. depth-
layer samples, the vacuum shroud attachment could not be interfaced properly with the
surface of the coupon and vacuum was lost, leading to loss of the ground material sample.
• When a means for collecting the ground material sample using the vacuum shroud was
added to the setup (i.e., a capture vessel installed in-line between the shroud and vacuum
source), vacuum suction strength was lost and was inadequate to collect the ground
material sample.
• The high speed of the grinder exacerbated loss of ground material sample, given the above
discussed inability to adequately contain and capture ground material via vacuum.
Due to the issues summarized above, the method for application of the grinding technology was
revised. Limestone (8 in. length by 8 in. width by 2.25 in. thick) or sealed concrete (12 in. length
by 12 in. width by 2 in. thick) coupons were placed inside the test chamber. The coupons were
leaned against the back of the chamber and, if necessary, to prevent them from slipping forward,
placed on a PTFE block. A sheet of aluminum foil with folded sides was placed under the
coupon to collect the removed/ground material. An aluminum shield was placed inside the foil to
the right of the coupon to deflect and collect the ground material (refer to Figure 15).
An angle grinder (1001672186, Home Depot, Columbus, OH) fitted with a diamond cup
grinding wheel (203061023, Home Depot, Columbus, OH) was connected to an adjustable
autotransformer (W5MT3, Variac®, General Radio Company) to slow the grinding speed. Using
the grinder, a section of the coupon was ground down at 0.25 in. depth increments. The depth of
each grind was measured using a ruler (refer to Figure 16). After grinding, the ground material
deflected by the shield and collected in the foil was poured into a sample jar and the mass
collected was determined gravimetrically.
An infrared thermometer (9248T57, McMaster-Carr Part, Aurora, OH) was used to measure the
surface temperature of each coupon before and after grinding.
A wet/dry vacuum (648846002842, Home Depot, Columbus, OH) was connected to the top of
the chamber to draw and capture the fine grinding dust created that was not captured/redirected
by the shield to the right of the coupon (this fine grinding dust was thus not included with the
ground material sample collected for each grinding pass).
37
-------
EPA/600/R-20/May 2020
Figure 15. Grinding Setup
Figure 16. Depth Measurement
Results
The depth reached after each grinding pass, temperature difference between before and after
grinding, and the mass of ground material collected for each depth layer are provided in Table 16
for limestone and in Table 17 for sealed concrete.
Table 16. Limestone Grinding Results
Coupon
Layer
Target
Depth
Temp Diff
(°F)
Grinding
Time (min)
1st Grind
Pass Depth
(in.)
2nd Grind
Pass Depth
(in.)
3rd Grind
Pass Depth
(in.)
Ground Material
Mass (g)
1
m
-0.3
8
1/8
1/8 - 1/4
%
36.34
1
2
¦
0.4
9
3/8-1/2
1
N/A
60.03
3
%
1.2
5
y4
N/A
N/A
59.80
4
i
0.9
11
i
N/A
N/A
57.72
1
ii
0.7
6
1/8
]/4
N/A
55.60
2
2
i
-1.1
12
3/8
%
N/A
67.07
3
%
1.7
8
5/8
3/4
N/A
59.51
Average
56.58
38
-------
EPA/600/R-20/May 2020
Table 17. Sealed Concrete Grinding Results
Coupon
Layer
Target
Depth
(in.)
Temp Diff
(°F)
Grinding
Time (min)
1st Grind
Pass Depth
(in.)
2nd Grind
Pass Depth
(in.)
3rd Grind
Pass Depth
(in.)
Ground Material
Mass (g)
1
1.1
8
3/16
S4
N/A
68.66
1
2
Jfi
0.3
10
3/8
%
N/A
66.68
3
It
3.2
8
5/8
%
N/A
68.56
4
i
0.8
6
1
N/A
N/A
62.77
1
m
2.4
6
N/A
N/A
56.70
2
2
is
2.5
7
%
N/A
N/A
68.05
3
%
2.7
7
5/8
14
N/A
71.72
Average
66.16
Both materials were amenable to grinding using the method described above. As the ground
material was ejected from the grinding wheel toward the right side of the setup (due to the nature
of operation of the grinder), the material was deflected/directed downward by the shield and
collected in the foil underneath the coupon. As mentioned, fine dust generated during grinding
was not captured as part of the sample collected for each grinding pass. After the first few
grinding applications, a sturdier aluminum baking dish (8.13 in. by 12.25 in., 551537495,
Walmart, Columbus, OH) that was better able to capture the ground material was used instead of
the folded aluminum sheet. Additionally, an aluminum easel was fabricated and used to hold the
coupons in place during grinding, which provided increased stability and better orientation of the
coupon (Figure 6). The size of the limestone coupons was much preferred over the larger sealed
concrete coupons, which were very difficult to maneuver inside the test chamber, so the size of
the concrete coupons was reduced to 5.75 in. by 5.75 in. for physical removal efficacy testing.
The reduced size allowed for proper application of the grinding technology while allowing for
simpler and safer manipulation of the sealed concrete coupons within the chamber.
Figure 17. Limestone Ground to 0.25 in. (left) and 1 in. (right)
39
-------
EPA/600/R-20/May 2020
Figure 18. Sealed Concrete Ground to 0.25 in. (left) and 1 in. (right)
Figure 19. Ground Material
3.1.1.3. Chemical Stripping
Method
Chemical stripper (Klean Strip GKS3 KS-3 Premium Paint Stripper) was pipetted onto the
coated surface of each material using an Eppendorf Repeater M4 Pipette (14-287-150, Fi sher
Scientific, Pittsburgh, PA). The chemical stripper was then either manually spread across an
approximately 1.375 in. area using a 14 in. brush (7237T84, McMaster-Carr, Aurora, OH) or
allowed to spread out on its own. The chemical stripper was allowed to dwell on the surface of
each material for a total 45 minutes (initial 30-minute dwell, then an additional 15-minute dwell
determined necessary based on visual inspection of the stripping progress, per manufacturer
directions). A 2 in. disposable putty knife (3546A421, McMaster-Carr, Aurora, OH) was then
used to scrape the coating from the surface of each material and the coating removed was
collected into a sample jar.
40
-------
EPA/600/R-20/May 2020
Results
The spread diameters for each of the different application methods are summarized in Table 18.
Table 18. Chemical Stripper Application Methods and Spread Results
Material
Paint Stripper Volume
Stripper Brushed
(Y/N)
Spread Diameter
(in.)
1 mL
Y
1.375
Painted Steel
1.5 mL
N
1.25
2 mL
N
1.375
Painted Wood
1 mL
Y
1.375
2 mL
N
1.5
For all replicates, the paint started to "bubble up" almost immediately after the chemical stripper
was applied to the materials. The treated paint was easily removed from all replicates after the
dwell time (Figure 20). The brushed replicates had a slightly cleaner surface (i.e., more complete
removal of paint, based on subjective visual assessment) after scraping (Figure 21 and Figure
22). Generally, a greater volume of residual paint stripper was observed on unbrushed replicates
after the dwell period (Figure 23 and Figure 24).
The method of pipetting 1 mL of chemical stripper onto coupons accompanied by brushing
produced the best results, defined as most complete removal of paint (again, subjective visual
assessment) using the least volume of chemical stripper. This method was then used to produce
additional replicate samples for the remainder of the method demonstration testing (solvent
extraction of VX from the stripped paint [waste sampling]).
Figure 20. Coating Removal
41
-------
EPA/600/R-20/May 2020
f
» ml. ^Iill lir hi til. -il |
1 ml laatiruilirdf
III hIm u,%llril|<
I rn(, ilirmiivilf
[
IJS ml |HhrH>ii»Klti
I ml lbnnhnJ>
Figure 21. Coated Steel Before (left) and After Scraping (right)
Figure 22. Coated Wood Before (left) and After Scraping (right)
ifirA i
— Vw lArj
jM
v
Figure 23. Collected Coating Using 1 niL Application with Brushing
42
-------
EPA/600/R-20/May 2020
Figure 24. Collected Coating Using 1.5 mL Application without Brushing
3.1.1.4. Vapor-Phase Solvent Extraction
Operating conditions/parameters to produce a viable solvent vapor stream could not be
determined using the vapor-phase solvent extraction test apparatus as conceptualized (setup as
depicted in Figure 7) or using revised apparatus setups described below. Budget and schedule
limitations of the project then precluded further development of the technology, so vapor-phase
solvent extraction was not carried forward into further methods demonstration test phases or into
physical removal efficacy testing.
Method
A limestone (8 in. length by 8 in. width by 2.25 in. thick) coupon was placed vertically in a
holder in the test chamber (same coupon holder used for the initially planned application method
for the core sampling approach; see Figure 9). A 20-mL scintillation vial with a funnel inserted
was placed underneath the holder and diy ice was packed around the vial (Figure 25).
Figure 25. Vial for Solvent Collection
43
-------
EPA/600/R-20/May 2020
The vapor-phase solvent extraction technology was set up as described in Section 2.2.3.4 and
depicted in Figure 7. The end of the inline static tube mixer (from which the solvent vapor
stream would be ejected) was attached to a bulkhead fitting installed in the side of the test
chamber, and the limestone coupon in the holder was oriented in front of the tube mixer opening
(outlet).
The heat tape affixed to the tee fitting and tube mixer was turned on and the temperature of the
system was increased so that the temperature of the tee fitting measured approximately 250°F
and the temperature at the end of the tube mixer was approximately 150°F. Conditioned (clean,
dried) house air was supplied through the system at approximately 300 mL/min. Acetone was
then infused into the airstream at the heated tee fitting using a syringe pump at a rate of 375
|iL/min (a 60-minute infusion duration would thus result in a total infused solvent volume of
22.5 mL).
Results
No solvent vapor was visible from the outlet of the tube mixer. Furthermore, after several
minutes, no condensed solvent was observed to collect on the surface of the coupon. Brown
Kraft paper and M8 Chemical Detector paper were also held against the outlet end of the tube
mixer for periods of up to approximately 60 minutes, but no visible collection of condensed
solvent or darkening (wetting) of the Kraft or M8 paper were observed. The acetone infusion rate
was increased incrementally from 375 |iL/min to up to 4.5 mL/min while the supply airflow rate
through the system was also increased incrementally from 300 mL/min to up to 1.9 L/min. No
combination of airflow rate and acetone infusion rate was identified that resulted in a visible
solvent vapor stream from the end of the tube mixer or in formation of condensed solvent
droplets on the surface of the coupon, Kraft paper, or M8 paper (though lower airflow rates and
higher acetone infusion rates did result in liquid acetone dripping from the end of the tube
mixer).
Alternative setups were assembled and evaluated as well. These included:
• Elimination of the static tube mixer. The heated tee was attached directly to the bulkhead
fitting at the side of the test chamber. The intent of this revised setup was to minimize the
distance between the heated tee and the point at which the solvent vapor stream would be
ejected and prevent loss of vapor to condensation within cooler portions of the system.
• Revision of the system design. Rather than generating solvent vapor via infusion of liquid
acetone into a heated tee fitting, a 1-liter (L) glass beaker filled with approximately 500 mL
of acetone was submerged in a second glass beaker containing water. The water was heated
to approximately 100°F (using heat tape wrapped around the outside of the second/outer
beaker) to act as a heated jacket. Conditioned (cleaned, dried) house air was then bubbled
into the heated acetone (various flowrates from 2 to 5 L/min were evaluated). The solvent
vapor generated in the headspace was then directed into the test chamber via the bulkhead
and the vapor stream was directed at the limestone coupon. Figure 26 provides a schematic
of the revised vapor-phase solvent extraction setup that was evaluated.
44
-------
EPA/600/R-20/May 2020
Solvent
vapor stream
Conditioned
air in
Acetone
Heated
water jacket
Figure 26. Revised Vapor-Phase Solvent Extraction Test Setup
Neither of these alternative setups produced visible solvent vapor streams or resulted in
collection of condensed solvent droplets or formation of wetted areas on the test coupon, Kraft
paper, or M8 paper.
3.1.2. Solvent Extraction Method Demonstration Results
The methods used for recovery of VX from coupon samples (core layer samples and ground
material samples) were demonstrated experimentally prior to the start of physical removal
efficacy testing to verify the adequacy of the methods. A single test was conducted using core
layer and ground material samples of each porous material type (sealed concrete and limestone;
produced during the technology functionality assessments of the grinding and core sampling
approaches) to evaluate recovery of VX via solvent extraction for all anticipated coupon sample
types after VX dwell periods of both 30 minutes and 24 hours. Testing was conducted as
described in Section 2.1.1.2.
Stainless steel coupons and glass bead samples were included in the test matrix to serve as
nonporous, inert positive control samples for core layer samples and ground material samples,
respectively. Positive controls were spiked with VX and then subsequently extracted after VX
dwell periods of 30 minutes and 24 hours.
Spike control recovery results are provided in Table 19. VX recovery results from the stainless
steel and glass bead positive control samples are provided in Tables 20 and 21. A single
procedural blank of each material type was included during the test as well. VX was detected in
all procedural blank extracts except for the ground limestone material procedural blank, but all
detections were within specification (Table 43).
45
-------
EPA/600/R-20/May 2020
Table 19. Solvent Extraction, Spike Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs target)
Avg
(Mg)
Avg % vs
Target
St Dev
%RSD
Spike Control 1
90500
1810
97%
Spike Control 2
91000
1820
97%
1809
97%
11
0.62%
Spike Control 3
89880
1798
96%
Table 20. Solvent Extraction, Stainless Steel Positive Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs SC avg)
Avg
(Mg)
Avg % vs
SC Avg
St Dev
%RSD
SS PC 1 (30 mill)
83470
1669
92%
SS PC 2 (30 mill)
107900
2158
119%
1851
102%
268
14%
SS PC 3 (30 mill)
86240
1725
95%
SS PC 1 (24 h)
96780
1936
107%
SS PC 2 (24 h)
98260
1965
109%
1806
100%
250
14%
SS PC 3 (24 h)
75910
1518
84%
SS = Stainless Steel; PC = Positive Control; SC = Spike Control
Table 21. Solvent Extraction, Glass Bead Positive Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs SC avg)
Avg
(Mg)
Avg % vs
SC Avg
St Dev
%RSD
Glass Beads PC 1 (30 mill)
30000
2400
133%
Glass Beads PC 2 (30 mill)
32230
2578
143%
2636
146%
269
10%
Glass Beads PC 3 (30 mill)
36610
2929
162%
Glass Beads PC 1 (24 h)
29800
2384
132%
Glass Beads PC 2 (24 h)
27020
2162
119%
2277
126%
111
4.9%
Glass Beads PC 3(24 h)
28580
2286
126%
PC = Positive Control; SC = Spike Control
As discussed in Section 2.1.1.2, the coupon sample solvent extraction method would be deemed
acceptable for use during subsequent physical removal efficacy evaluations if the mean
recoveries from stainless steel positive controls and glass bead positive controls were within the
range of 70% to 120% of the mean of the SC results with a relative standard deviation (RSD)
between replicates of less than 30%. Recoveries from stainless steel at both VX dwell periods
met this requirement.
Average recovery of VX from the glass beads was higher than the criterion defined in Section
2.1.1.2 of 70% to 120% of the spike control average, with average recoveries of 146% at 30
minutes and 126% at 24 hours. RSD in both cases was within specification (10% at 30 minutes,
4.9% at 24 hours). No immediately attributable cause for the high recoveries from the glass
beads was available, but we speculated that the beads may have been exposed to additional VX
(past the 2 |iL spiked into the beads) due to the manner in which the beads were spiked. The tip
of the positive displacement pipettor used to spike the glass bead positive controls was
submerged into a vial containing VX to draw up the 2 |iL to be spiked. The end of the pipettor
tip was touched to an absorbent wipe to ensure excess VX drips/droplets hanging to the tip were
removed, but the remaining exterior of the tip was not wiped off. The tip was then inserted into
the beads and the VX expelled from the tip. The tip was then used to mix the spiked VX into the
46
-------
EPA/600/R-20/May 2020
beads. We suspect that VX still present on the outside surface of the pipettor tip (from drawing
VX into the tip) was transferred to the glass beads, leading to higher recoveries.
VX recoveries from sealed concrete and limestone core layer samples via solvent extraction after
VX dwell periods of 30 minutes and 24 hours are provided in Table 22. As described in Section
2.1.1.2, PTFE disks were placed below the core layer samples during the VX dwell periods to
assess for VX breakthrough (i.e., VX migrating through the 0.25 in.-thick core layer samples)
during the dwell periods. PTFE disks were extracted in solvent and analyzed via LC-MS/MS
alongside the core layer samples. Recoveries from PTFE disks are provided in Table 22 as well.
Table 22. Solvent Extraction, Core Layer Samples
Sample Description
Extract Cone.
(ng/mL)
Mass
(M^g)
% Recovery
(vs Pos avg)
Avg
(Mg)
Avg % vs
Pos Avg
St Dev
%RSD
Concrete Core Layer (30 min) 1
22110
442
24%
Concrete Core Layer (30 min) 2
19410
388
21%
411
22%
28
6.8%
Concrete Core Layer (30 min) 3
20150
403
22%
Concrete Core Layer (24 h) 1
9982
200
11%
Concrete Core Layer (24 h) 2
7839
157
8.7%
182
10%
22
12%
Concrete Core Layer (24 h) 3
9439
189
10%
Concrete Core Layer (30 min) 1 - PTFE
0.36
0.004
0.0002%
Concrete Core Layer (30 min) 2 - PTFE
0.51
0.005
0.0003%
0.003
0.0002%
0.002
62%
Concrete Core Layer (30 min) 3 - PTFE
0.11
0.001
0.0001%
Concrete Core Layer (24 h) 1 - PTFE
1.9
0.02
0.001%
Concrete Core Layer (24 h) 2 - PTFE
2.2
0.02
0.001%
0.02
0.001%
0.004
23%
Concrete Core Layer (24 h) 3 - PTFE
1.4
0.01
0.001%
Limestone Core Layer (30 min) 1
18010
360
19%
Limestone Core Layer (30 min) 2
13920
278
15%
362
20%
84
23%
Limestone Core Layer (30 min) 3
22320
446
24%
Limestone Core Layer (24 h) 1
7497
150
8.3%
Limestone Core Layer (24 h) 2
5063
101
5.6%
133
7.4%
28
21%
Limestone Core Layer (24 h) 3
7398
148
8.2%
Limestone Core Layer (30 min) 1 -
PTFE
0.67
0.01
0.0003%
Limestone Core Layer (30 min) 2 -
PTFE
0.44
0.004
0.0002%
0.01
0.0003%
0.003
44%
Limestone Core Layer (30 min) 3 -
PTFE
1.1
0.01
0.0004%
Limestone Core Layer (24 h) 1 - PTFE
0.96
0.01
0.0004%
Limestone Core Layer (24 h) 2 - PTFE
0.68
0.01
0.0003%
0.01
0.0005%
0.01
48%
Limestone Core Layer (24 h) 3 - PTFE
1.7
0.02
0.001%
Average recoveries of VX from the sealed concrete and limestone core layer samples at 30
minutes were 22% and 20%, respectively. Average recoveries at 24 hours were lower at 10%
(sealed concrete) and 7.4% (limestone). Such recoveries from these materials appeared to be
generally consistent with previously acquired data [10] for solvent extraction recovery of CWAs
from concrete and with studies that have shown that concrete is capable of active degradation of
penetrated/absorbed CWAs [9],
VX recoveries from sealed concrete and limestone ground material samples via solvent
extraction after VX dwell periods of 30 minutes and 24 hours are provided in Table 23.
47
-------
EPA/600/R-20/May 2020
Table 23. Solvent Extraction, Ground Material Samples
Sample Description
Extract Cone.
(ng/mL)
Mass
(M^g)
% Recovery
(vs Pos avg)
Avg
(MS)
Avg % vs
Pos Avg
St Dev
%RSD
Concrete Ground Material (30 min) 1
23740
1899
72%
Concrete Ground Material (30 min) 2
15950
1276
48%
1070
41%
949
89%
Concrete Ground Material (30 min) 3
449
36
1.4%
Concrete Ground Material (24 h) 1
17850
1428
63%
Concrete Ground Material (24 h) 2
235
19
0.82%
560
25%
759
136%
Concrete Ground Material (24 h) 3
2921
234
10%
Limestone Ground Material (30 min) 1
22350
1788
68%
Limestone Ground Material (30 min) 2
1815
145
5.5%
1361
52%
1068
78%
Limestone Ground Material (30 min) 3
26860
2149
82%
Limestone Ground Material (24 h) 1
27780
2222
98%
Limestone Ground Material (24 h) 2
27230
2178
96%
2157
95%
78
3.6%
Limestone Ground Material (24 h) 3
25890
2071
91%
Recovery of VX from ground sealed concrete ranged from 0.82% to 72%. Average recovery at
30 minutes was 41% and at 24 hours was 25%, and RSDs were high at 89% (30 minutes) and
136% (24 hours). The test was conducted as described in Section 2.1.1.2. Samples were spiked
as described in Section 2.2.2.2 in that VX was spiked into the ground material, and the pipettor
tip was then used to mix the spiked VX into the ground material. Use of this approach caused the
ground material to "clump" and "stick" to the pipettor tip. Attempts were made to remove the
stuck material from the tip into the extraction jar but based on the data it appears that this
approach was incomplete and inconsistent. Thus, recoveries from ground sealed concrete are
likely impacted by the amount of "clumped" or "stuck" material that was able to be removed
from the pipette tip. Recoveries from ground limestone after a VX dwell period of 30 minutes
also appeared to have been impacted by this issue of "clumping" VX/ground material. Oddly
though, average recovery of VX from ground limestone at 24 hours was 95% with only 3.6%
RSD. No explanation could be determined for why the "clumping" and "sticking" issue impacted
recoveries from ground sealed concrete (both dwell times) and ground limestone after 30
minutes but did not affect recovery from ground limestone after a 24-hour dwell period.
Figure 27 summarizes average VX mass recoveries via solvent extraction after dwell periods of
30 minutes and 24 hours from the core layer and ground material coupon samples of sealed
concrete and limestone. Figure 28 summarizes average percent recoveries. Given the known
difficulty associated with recovery of CWA from concrete [9,10] (and likely similar materials),
the suspected cause of the higher-than-specification recoveries from the glass bead positive
controls, and the acceptable recoveries from the stainless steel positive controls, the solvent
extraction method was used as evaluated during method demonstration testing during subsequent
physical removal efficacy tests (i.e., the method was not revised or altered from that described in
Section 2.2.5 based on the results of method demonstration testing).
48
-------
EPA/600/R-20/May 2020
Solvent Extraction Method Demonstration
Average VX Mass Recovery
(error bars equal ± 1 standard deviation)
Sample
¦
¦
Figure 2 7. Solvent Extraction, Average Mass Recovery
Solvent Extraction Method Demonstration
Average Percent Recovery
(SCs vs target; PCs vs SCs; test samples vs PCs)
Sample
Figure 28. Solvent Extraction, Average Percent Recovery
3.1.3. Wipe Sampling Method Demonstration Results
A single test to assess the adequacy of the wipe sampling method was performed. VX (10 |iL)
was applied to coupons of sealed concrete, limestone, painted steel, and painted wood and
allowed to dwell on the surface of coupons for 24 hours. Stainless steel coupons were spiked
49
-------
EPA/600/R-20/May 2020
with VX as well to serve as positive controls. Following the dwell period, wipe samples were
collected from the coupons.
Spike control recovery results are provided in Table 24.
Table 24. Wipe Sampling, Spike Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs target)
Avg
(Mg)
Avg % vs
Target
St Dev
%RSD
Spike Control 1
517500
10350
111%
Spike Control 2
484000
9680
104%
10000
107%
336
3.4%
Spike Control 3
498500
9970
107%
VX recoveries from the stainless-steel positive control samples are provided in Table 25.
Table 25. Wipe Sampling, Stainless Steel Positive Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs SC avg)
Avg
(Mg)
Avg % vs
SC Avg
St Dev
%RSD
Stainless Steel Wipe 1
363700
7783
78%
Stainless Steel Wipe 2
359400
7691
77%
7867
79%
230
2.9%
Stainless Steel Wipe 3
379800
8128
81%
SC = Spike Control
VX recoveries via wipe-sampling from the porous materials and permeable coatings are provided
in Table 26. Average VX mass recoveries are summarized in Figure 29. Average percent
recoveries are summarized in Figure 30.
Table 26. Wipe Sampling Results
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs Pos avg)
Avg
(Mg)
Avg % vs
Pos Avg
St Dev
%RSD
Concrete Wipe 1
2509
54
0.68%
Concrete Wipe 2
2993
64
0.81%
60
0.77%
5.9
9.7%
Concrete Wipe 3
2977
64
0.81%
Limestone Wipe 1
1781
38
0.48%
Limestone Wipe 2
1446
31
0.39%
32
0.41%
5.4
17%
Limestone Wipe 3
1282
27
0.35%
Painted Steel Wipe 1
184000
3938
50%
Painted Steel Wipe 2
199500
4269
54%
4586
58%
852
19%
Painted Steel Wipe 3
259400
5551
71%
Painted Wood Wipe 1
172500
3692
47%
Painted Wood Wipe 2
149000
3189
41%
3293
42%
357
11%
Painted Wood Wipe 3
140200
3000
38%
50
-------
EPA/600/R-20/May 2020
Wipe Sampling Method Demonstration
Average VX Mass Recovery
(error bars equal ± 1 standard deviation)
ll i.
in in in in in in
— 100%
90% 79%
* 80%
§ 70% 5Wo
60% H H _
50% ¦ 42%
IS ¦ ¦
10% 0.77% 0.41%
0%
in in m in m in
— 0) 0) 0) 0) 0)
2 9-9-9-9-9-
!§§§§§
O QJ 5^ QJ ^
ij £ £! o £ §
~ ^ " tt "> g
m c d) -q
1/1 o e aj ~a
Q) LJ C 4-J d)
K Sample ^ 2
Figure 30. Wipe Sampling, Average Percent Recovery
51
11000
10000
_ 9000
j£ 8000
| 7000
g 6000
8 5000
<1J
®; 4000
tf)
J 3000
2000
1000
0
-------
EPA/600/R-20/May 2020
The coupon wipe-sampling method would be deemed acceptable for use during subsequent
physical removal efficacy evaluations if the mean wipe-sampling recovery from the stainless-
steel positive controls was within the range of 70% to 120% of the mean of the SC results with
an RSD between replicates of less than 30%. Average percent recovery from the stainless-steel
positive controls via wipe-sampling was 79% with an RSD of 2.9%, so the recovery criterion
was met.
Data generated during previous studies [10] have demonstrated difficulty in recovering VX
spiked onto concrete surfaces using solvent extraction techniques. Additionally, previous work
[11] has demonstrated that VX spiked onto a paint layer applied to a substrate will absorb into
the paint layer. The data generated during wipe-sampling method demonstration testing
conducted during this project were consistent with these findings, as average percent recoveries
measured only 0.77% from sealed concrete, 0.41% from limestone, 58% from painted steel, and
42% from painted wood (compared to the stainless-steel positive controls). Despite these low
recoveries, the wipe sampling method was deemed adequate for use during subsequent physical
removal efficacy testing given that the criterion for recovery from stainless steel was met, as
discussed above.
3.1.4. Waste Sampling Results
The methods that would be used for recovery of VX from wastes generated during application of
the physical removal technologies were experimentally evaluated prior to physical removal
efficacy testing. Samples of the wastes produced during the technology functionality assessments
(sealed concrete and limestone cutting dusts produced from application of the core sampling
approach, and coatings stripped from steel and hardwood from application of the chemical
stripper) were spiked with VX and allowed to dwell for 30 minutes or 24 hours. Following the
specified dwell period, waste samples were extracted with solvent. As during ground material
solvent extraction method demonstration testing, glass bead samples were used as positive
controls at both dwell periods.
Spike control recovery results are provided in Table 27.
Table 27. Waste Sampling, Spike Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(M^g)
% Recovery
(vs target)
Avg
(Mg)
Avg % vs
Target
St Dev
%RSD
Spike Control 1
97320
1946
104%
Spike Control 2
88770
1775
95%
1856
99%
86
4.6%
Spike Control 3
92300
1846
99%
VX recoveries from the glass bead positive control samples at both dwell periods are provided in
Table 28. Average percent VX recovery via solvent extraction from the glass bead positive
controls (vs the spike control mean) after a dwell period of 30 minutes was 108% with an RSD
of 10%. Average percent recovery after a dwell period of 24 hours was lower. However, at only
51% with an RSD of 18%. The amount of glass beads used for the positive control samples was
based on the average mass of the cutting dust samples collected during the technology
52
-------
EPA/600/R-20/May 2020
functionality assessments and used during the waste sampling method demonstration test.
Approximately 2.1 g of glass beads were used for each positive control sample. This
number/amount/volume of beads did not cover the bottom of the 60 mL extraction jar fully (i.e.,
the beads formed a single "layer" across the bottom of the jar). Although VX is considered a
persistent hazard (vapor pressure of 0.09 Pa), it is suspected that the beads provided a greater
evaporative surface area (i.e., thinner "coat" of VX on/across the beads, as opposed to a single
droplet), and a greater amount of VX was lost via evaporation during the 24-hour dwell period.
Table 28. Waste Sampling, Glass Bead Positive Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(M^g)
% Recovery
(vs SC avg)
Avg
(Mg)
Avg % vs
SC Avg
St Dev
%RSD
Glass Beads PC 1 (30 min)
112000
2240
121%
Glass Beads PC 2 (30 min)
93940
1879
101%
2002
108%
206
10%
Glass Beads PC 3 (30 min)
94390
1888
102%
Glass Beads PC 1 (24 h)
57490
1150
62%
Glass Beads PC 2 (24 h)
40920
818
44%
956
51%
173
18%
Glass Beads PC 3(24 h)
44950
899
48%
PC = Positive Control; SC = Spike Control
VX mass recoveries and percent recoveries from sealed concrete and limestone cutting dust
samples are provided in Table 29. Percent recoveries were determined via comparison to the
glass bead positive control mean for each dwell period. Given the lower recoveries obtained
from glass beads at the 24-hour dwell period, percent recovery from sealed concrete and
limestone cutting dust samples at 24 hours were also determined via comparison to the spike
control mean recovery.
Table 29. Waste Sampling, Cutting Dust Recovery
Sample Description
Extract
Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs Pos avg)
Avg
(Mg)
Avg % vs
Pos Avg
St Dev
%RSD
% Recovery
(vs SC avg) A
Avg % vs
SC Avg
Concrete Cutting Dust (30 min) 1
95080
1902
95%
-
Concrete Cutting Dust (30 min) 2
106300
2126
106%
1982
99%
125
6.3%
-
-
Concrete Cutting Dust (30 min) 3
95880
1918
96%
-
Concrete Cutting Dust (24 h) 1
79200
1584
166%
85%
Concrete Cutting Dust (24 h) 2
90870
1817
190%
1755
184%
150
8.5%
98%
95%
Concrete Cutting Dust (24 h) 3
93130
1863
195%
100%
Limestone Cutting Dust (30 min) 1
97960
1959
98%
-
Limestone Cutting Dust (30 min) 2
100400
2008
100%
1961
98%
47
2.4%
-
-
Limestone Cutting Dust (30 min) 3
95740
1915
96%
-
Limestone Cutting Dust (24 h) 1
104300
2086
218%
112%
Limestone Cutting Dust (24 h) 2
103600
2072
217%
1980
207%
172
8.7%
112%
107%
Limestone Cutting Dust (24 h) 3
89050
1781
186%
96%
SC = Spike Control A Compared also to SC avg because of lower recovery from glass beads at 24 hours
VX mass recoveries and percent recoveries from permeable coatings stripped from steel and
hardwood are provided in Table 30. Average percent recoveries from coatings stripped from
steel and hardwood after a 30-minute VX dwell period (vs the spike control mean recovery) were
95% and 91%, respectively. Average percent recoveries from coatings stripped from steel and
hardwood after a 24-hour dwell period were lower at 65% and 82%, respectively. The lower
recoveries after the longer dwell period are consistent with findings from previous studies
53
-------
EPA/600/R-20/May 2020
demonstrating that VX spiked onto a paint layer applied to a substrate will absorb into the paint
layer [12], if the absorption decreases "extractability" of VX from the paint in some manner.
Table 30. Waste Sampling, Stripped Coating Recovery
Sample Description
Extract Cone.
(ng/mL)
Mass
(M^g)
% Recovery
(vs SC avg)
Avg
(MS)
Avg % vs
SC Avg
St Dev
%RSD
Stripped Paint Steel (30 min) 1
97850
1957
105%
Stripped Paint Steel (30 min) 2
90220
1804
97%
1767
95%
212
12%
Stripped Paint Steel (30 min) 3
76920
1538
83%
Stripped Paint Steel (24 h) 1
36620
732
39%
Stripped Paint Steel (24 h) 2
83330
1667
90%
1198
65%
467
39%
Stripped Paint Steel (24 h) 3
59760
1195
64%
Stripped Paint Wood (30 min) 1
85610
1712
92%
Stripped Paint Wood (30 min) 2
87370
1747
94%
1686
91%
78
4.6%
Stripped Paint Wood (30 min) 3
79940
1599
86%
Stripped Paint Wood (24 h) 1
86660
1733
93%
Stripped Paint Wood (24 h) 2
66160
1323
71%
1518
82%
206
14%
Stripped Paint Wood (24 h) 3
74930
1499
81%
SC = Spike Control
Average VX mass recoveries from cutting dust and stripped coating samples are summarized in
Figure 31. Average percent recoveries are summarized in Figure 32.
Waste Sampling Method Demonstration
Average VX Mass Recovery
(error bars equal ± 1 standard deviation)
2500
Sample
Figure 31. Waste Sampling, Average Mass Recovery
54
-------
EPA/600/R-20/May 2020
Waste Sampling Method Demonstration
Average Percent Recovery
(SCs vs target; PCs vs SCs; cutting dust vs PCs; stripped paint vs SCs)
225% 207%
a 200%
O 175%
ec 150%
125%
100%
75%
50%
g 25%
a>
a.
ai
m
ro
<
0%
]
.84°/
108%
33% "
99/0
9o/o
yb%
91%
Q")0/
65%
¦
_
_
_
—
_
—
—
—
¦
_
_
¦
Sample
Figure 32. Waste Sampling, Average Percent Recovery
3.1.5. VX Depth Penetration Assessment Results
The core sampling approach was used to assess the depth to which VX penetrates the porous
materials selected for physical removal efficacy evaluations. Core samples were excised from
sealed concrete and limestone coupons, the cores were spiked with 10 |iL of VX (center of top
surface of each core sample), and the VX was allowed to penetrate into the cores for a period of
24 hours. Following the 24-hour dwell period, the core sampling approach was applied to
separate the core samples into 0.25 in.-thick layers that were individually extracted in solvent.
VX depth penetration into the cores was then assessed based on recovery of VX from the
individual core layer samples.
Spike control recovery results are provided in Table 31.
Table 31. VX Depth Penetration Assessment, Spike Controls
Sample
Description
Extract Cone.
(ng/mL)
Mass
(M^g)
% Recovery
(vs target)
Avg
(Mg)
Avg % vs
Target
St Dev
%RSD
Spike Control 1
435400
8708
86%
8445
84%
493
5.8%
Spike Control 2
393800
7876
78%
Spike Control 3
437500
8750
87%
VX mass recovery and percent recovery results for each limestone core sample included during
the VX depth penetration assessment are provided in Table 32. Mass recovery and percent
recovery results for sealed concrete core samples are provided in Table 33. Mass recovery results
from all core samples are also summarized in Figure 33.
55
-------
EPA/600/R-20/May 2020
Table 32. VX Depth Penetration Assessment, Limestone Recovery
Sample
Description
Limestone Core 1
Extract Cone. Mass % Recovery
(ng/mL) (jig) (vs SC avg)
Limestone Core 2
Extract Cone. Mass % Recoveiy
(ng/mL) (jig) (vs SC avg)
Limestone Core 3
Extract Cone. Mass % Recoveiy
(ng/mL) (jig) (vs SC avg)
Avg Mass
(Mg)
Avg Mass
% Recoveiy
(vs SC Avg)
St Dev
%RSD
Wipe
1806
39
0.46%
1744
37
0.44%
853
18
0.22° o
31
0.37%
11
36%
Layer 1
43150
863
10%
42710
854
10%
51110
1022
12%
913
11%
95
10%
Cutting Dust 2
0.55
0.01
0.0001%
18
0.36
0.004%
1.1
0.02
0.0003%
0.13
0.002%
0.20
152%
Layer 2
1.1
0.02
0.0003%
4.1
0.08
0.001%
7.3
0.15
0.002%
0.08
0.001%
0.06
75%
Cutting Dust 3
<0.10
<0.002
<0.00002%
<0.10
<0.002
<0.00002° b
0.11
0.002
0.00003%
0.002
0.00002%
0.0001
6.9%
Layer 3
0.68
0.01
0.0002%
0.84
0.02
0.0002%
3.9
0.08
0.001%
0.04
0.0004%
0.04
99.9%
Cutting Dust 4
<0.10
<0.002
<0.00002%
0.10
0.002
0.00002%
0.12
0.002
0.00003%
0.002
0.00003%
0.0003
12%
Layer 4
0.28
0.01
0.0001%
0.25
0.01
0.0001%
2.1
0.04
0.0005%
0.02
0.0002%
0.02
121%
Cutting Dust 5
<0.10
<0.002
<0.00002%
<0.10
<0.002
<0.00002° b
<0.10
<0.002
<0.00002%
0.002
0.00002%
0.00
0.00%
Layer 5
<0.10
<0.002
<0.00002%
<0.10
<0.002
<0.00002%
0.30
0.01
0.0001%
0.003
0.00004%
0.002
70%
Blade Wipe
0.41
0.01
0.0001%
0.38
0.01
0.0001%
0.54
0.01
0.0001%
0.01
0.0001%
0.002
19%
Total Mass
NA
902
11%
NA
892
11%
NA
1041
12%
945
11%
83
8.8%
Table 33. VX Depth Penetration Assessment, Sealed Concrete Recovery
Sample
Sealed Concrete Core 1
Sealed Concrete Core 2
Sealed Concrete Core 3
Extract Cone.
% Recovery Extract Cone.
% Recovery Extract Cone.
% Recovery
Avg Mass
Avg Mass
% Recovery
(ng/mL)
(Mg)
(vs SC avg)
(ng/mL)
(Mg)
(vs SC avg)
(ng/mL)
(Mg)
(vs SC avg)
(vs SC Avg)
Wipe
2704
58
0.69° o
2845
61
0.72%
1979
42
0.50%
54
0.64%
9.9
19%
Layer 1
67960
1359
16%
56630
1133
13%
47790
956
11%
1149
14%
202
18%
Cutting Dust 2
No sample A
1.1
0.02
0.0003%
0.42
0.01
0.0001%
0.02
0.0002%
0.01
65%
Layer 2
No sample A
0.59
0.01
0.0001%
0.73
0.01
0.0002%
0.01
0.0002%
0.002
15%
Cutting Dust 3
0.52
0.01
0.0001%
0.54
0.01
0.0001%
0.23
0.005
0.0001%
0.01
0.0001%
0.004
41%
Layer 3
0.63
0.01
0.0001%
1.0
0.02
0.0002%
0.51
0.01
0.0001%
0.01
0.0002%
0.005
37%
Cutting Dust 4
0.16
0.003
0.00004%
0.39
0.01
0.0001%
0.17
0.003
0.00004%
0.005
0.0001%
0.003
55%
Layer 4
0.47
0.01
0.0001%
0.55
0.01
0.0001%
0.17
0.003
0.00004%
0.01
0.0001%
0.004
50%
Cutting Dust 5
<0.10
<0.002
<0.00002%
0.28
0.01
0.0001%
0.34
0.01
0.0001%
0.005
0.0001%
0.003
52° o
Layer 5
0.22
0.004
0.0001%
0.28
0.01
0.0001%
<0.10
<0.002
<0.00002%
0.004
0.00005%
0.002
46%
Blade Wipe
0.34
0.01
0.0001%
0.22
0.005
0.0001%
0.11
0.002
0.00003%
0.005
0.0001%
0.002
51%
Total Mass
NA
1417
16.8%
NA
1194
14%
NA
998
12%
1203
14%
210
17%
L Only four (4) layers were cut and collected. No "Layer 2" or "Cutting Dust 2" samples.
56
-------
EPA/600/R-20/May 2020
VX Depth Penetration
VX Mass Recovery by Core Sample Component
Wipe, 39 ug
Layer 1, 863 ug
Dust 2, 0.01 ug
Layer2, 0.02 ug
Dust 3,0.002 ug
Layer 3, 0.01 ug
Dust 4,0.002 ug
Layer4, 0.01 ug
Dust 5,0.002 ug
Layer5, 0.002 ug
Blade, 0.01 ug
Sample
Wipe, 37 ug
Layer 1, 854 ug
Dust 2,0.36 ug
Layer 2, 0.08 ug
Dust 3, 0.002 ug
Layer 3, 0.02 ug
Dust 4, 0.002 ug
Layer4, 0.01 ug
Dust 5, 0.002 ug
Layer5, 0.002 ug
Blade, 0.01 ug
Wipe, 18 ug
Layer 1,1022 ug
Dust 2,0.02 ug
Layer 2, 0.15 ug
Dust 3,0.002 ug
Layer 3, 0.08 ug
Dust 4, 0.002 ug
Layer4, 0.04 ug
Dust 5, 0.002 ug
fl
Layer5, 0.006 ug
Blade, 0.01 ug
Wipe, 61 ug
Layer 1, 1133 ug
Dust 2,0.02 ug
Layer2, 0.01 ug
Dust 3,0.011 ug
Layer 3, 0.02 ug
Dust 4,0.008 ug
Layer4, 0.01 ug
Dust 5, 0.006 ug
Layer 5, 0.006 ug
Blade, 0.005 ug
Wipe, 42 ug
Layer 1, 956 ug
Dust 2,0.01 ug
Layer 2, 0.01 ug
Dust 3,0.005 ug
Layer 3, 0.01 ug
Dust 4, 0.003 ug
Layer4, 0.003 ug
Dust 5, 0.007 ug
Layer5, 0.002 ug
Blade, 0.002 ug
Figure 33. VX Depth Penetration Assessment, VX Mass Recovery by Component
¦ Wipe
¦ Layer 1
Dust 2
Layer 2
¦ Dust 3
¦ Layer 3
¦ Dust 4
¦ Layer 4
¦ Dust 5
¦ Layer 5
¦ Blade
57
-------
EPA/600/R-20/May 2020
Generally, total recoveries from core samples (sum of the masses recovered from each core
layer, the cutting dusts collected during "slicing" of each layer from the core, and wipe samples
obtained from the top of the core following the 24-hour dwell period and from the blade
following application of the core sampling approach) were low compared to the associated spike
control mean recoveries, measuring only 11% average total recovery from limestone cores and
14% average total recovery from sealed concrete cores.
The majority of VX recovered from each core sample was obtained from solvent extraction of
the 1st layer sample (the "topmost" layer of the core that was initially contaminated with VX).
The next highest recovery from each core sample was obtained via the wipe sample taken from
the top surface of the core following the 24-hour dwell period (prior to application of the core
sampling approach). VX mass recoveries from the core layer and cutting dust samples collected
below the lst/topmost layer (layer and cutting dust samples 2 through 5) drop off significantly,
indicating that either VX does not penetrate into the materials past the topmost approximately
0.25 in. depth (via gravity, over the course of 24 hours), or VX becomes increasingly
unrecoverable or degrades as it penetrates farther than approximately 0.25 in. into the materials.
The low recoveries are also consistent with [10] and potentially attributable to previously
implied degradation of VX on concrete due to the presence of basic catalytic sites [6,7,9], Such
degradation may even be enhanced at elevated material temperatures as observed during the
cutting of the concrete slices.
3.2. Physical Removal Efficacy Results - Grinding
Sealed concrete and limestone coupons were contaminated with 10 |iL of VX (single 10-|iL
droplet applied in the center of the top surface of the coupon), and the VX was allowed to dwell
on the surface of the coupon for 24 hours. Following the 24-hour dwell period, the surface of the
coupon was sampled via wipe sampling, then the grinding technology was applied to collect
successive 0.25 in. depth layer samples from the coupon.
Spike control recovery results from the grinding tests are provided in Table 34.
Table 34. Grinding, Spike Controls
Sample Description
Extract Conc.
(ng/mL)
Mass
(M^g)
% Recovery
(vs target)
Avg
(Mg)
Avg % vs
Target
St Dev
%RSD
SCI (sealed conc. test)
470400
9408
93%
SC2 (sealed conc. test)
452800
9056
90%
9163
91%
213
2.3%
SC3 (sealed conc. test)
451200
9024
89%
SCI (limestone test)
434700
8694
86%
SC2 (limestone test)
423700
8474
84%
8694
86%
220
2.5%
SC3 (limestone test)
445700
8914
88%
The mass of each ground layer sample that was collected though application of the grinding
technology was determined gravimetrically. Sealed concrete and limestone ground layer masses
are provided in Tables 35 and 36.
58
-------
EPA/600/R-20/May 2020
Table 35. Sealed Concrete, Ground Layer Masses
Coupon 1
Coupon 2
Coupon 3
PB
Sample Description
Mass
Mass
Mass
Mass
(g)
(g)
(g)
(g)
Sealed Concrete Ground layer 1
33.3
52.4
57.3
40.5
Sealed Concrete Ground layer 2
74.4
51.2
56.4
46.8
Sealed Concrete Ground layer 3
81.3
84.6
73.5
46.5
Sealed Concrete Ground layer 4
72.8
45.5
43.8
26.8
Avg
65.5
58.4
57.8
40.2
St Dev
21.7
17.7
12.2
9.4
%RSD
33%
30%
21%
23%
Total Mass
261.8
233.7
231.0
160.6
PB = Procedural blank
Table 36. Limestone, Ground Layer Masses
Coupon 1
Coupon 2
Coupon 3
PB
Sample Description
Mass
Mass
Mass
Mass
(g)
(g)
(g)
(g)
Limestone Ground layer 1
25.6
64.3
47.1
72.8
Limestone Ground layer 2
50.0
None A
43.6
92.6
Limestone Ground layer 3
38.3
51.7
52.7
87.0
Limestone Ground layer 4
42.0
55.6
43.6
72.5
Avg
39.0
57.2
46.8
81.2
St Dev
10.2
6.5
4.3
10.2
%RSD
26%
11%
9%
13%
Total Mass
155.9
171.6
187.0
324.9
PB = Procedural blank
A Inadvertently ground deeper than target 1/4 in. for 1st layer; no sample for 2nd layer, since
already ground to 1/2 in.
VX mass recovery and percent recovery results for the wipe and ground layer samples collected
from each sealed concrete coupon included during grinding technology physical removal
efficacy testing are provided in Table 37. Mass recovery and percent recovery results for the
wipe and ground layer samples collected from limestone coupons are provided in Table 38. Wipe
sample and ground layer sample mass recovery results from all coupons are also summarized in
Figure 34.
59
-------
EPA/600/R-20/May 2020
Table 3 7. Grinding, Sealed Concrete Recovery
Sealed Concrete Coupon 1
Sealed Concrete Coupon 2
Sealed Concrete Coupon 3
Avg Mass
Sample
Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs SC avg)
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs SC avg)
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs SC avg)
Avg Mass
(Mg)
% Recovery
(vs SC Avg)
St Dev
%RSD
Coupon Wipe
1732
37
0.40%
1609
34
0.38%
1001
21
0.23%
31
0.34%
8.4
27%
Grinder Wipe
8.8
0.19
0.002%
7.6
0.16
0.002%
3.1
0.07
0.001%
0.14
0.002%
0.06
47%
Shield Wipe
21
0.46
0.01%
167
3.6
0.04%
69
1.5
0.02%
1.8
0.02%
1.6
86%
Ground layer 1
5500
440
4.8%
17940
1435
16%
4351
348
3.8%
741
8.1%
603
81%
Ground layer 2
43
3.4
0.04%
66
5.3
0.06%
36
2.9
0.03%
3.9
0.04%
1.3
33%
Ground layer 3
24
1.9
0.02%
25
2.0
0.02%
24
1.9
0.02%
2.0
0.02%
0.07
3.5%
Ground layer 4
5.8
0.47
0.01%
17
1.4
0.01%
14
1.2
0.01%
1.0
0.01%
0.47
47%
Total Mass
NA
484
5.3%
NA
1482
16%
NA
377
4.1%
781
8.5%
610
78%
Table 38. Grinding, Limestone Recovery
Limestone Coupon 1
Limestone Coupon 2
Limestone Coupon 3
Avg Mass
Sample
Description
Extract Cone.
Mass
% Recovery
Extract Cone.
Mass
% Recoveiy
Extract Cone.
Mass
% Recoveiy
Avg Mass
(Mg)
% Recoveiy
St Dev
%RSD
(ng/mL)
(Mg)
(vs SC avg)
(ng/mL)
(Mg)
(vs SC avg)
(ng/mL)
(Mg)
(vs SC avg)
(vs SC Avg)
Coupon Wipe
1466
31
0.36%
930
20
0.23%
571
12
0.14%
21
0.24%
9.6
46%
Grinder Wipe
228
4.9
0.06%
184
3.9
0.05%
215
4.6
0.05%
4.5
0.05%
0.48
11%
Shield Wipe
150
3.2
0.04%
58
1.2
0.01%
373
8.0
0.09%
4.1
0.05%
3.5
84%
Ground layer 1
47730
3818
44%
35810
2865
33%
55080
4406
51%
3697
43%
778
21%
Ground layer 2
323
26
0.30%
None A
176
14
0.16%
20
0.23%
8.3
42%
Ground layer 3
139
11
0.13%
6339
507
5.8%
12
0.98
0.01%
173
2.0%
289
167%
Ground layer 4
28
2.2
0.03%
5823
466
5.4%
5.5
0.44
0.01%
156
1.8%
268
172%
Total Mass
NA
3897
45%
NA
3863
44%
NA
4447
51%
4069
47%
328
8.1%
A Inadvertently ground deeper than target 1/4 in. for 1st layer; no sample for 2ntl layer, since already ground to 1/2 in.
60
-------
EPA/600/R-20/May 2020
Grinding
VX Mass Recovery by Sample Component
¦ Panel Wipe, 31
¦ Grinder Wipe, 4.9
• Shield Wipe, 3.2
Panel
Wipe, 20
Grinder
Wipe, 3.9
Shield
Wipe, 1.2
Panel Wipe, 12
Grinder Wipe, 4.6
¦ Shield Wipe, 8.0
Panel Wipe,
37
Panel Wipe,
34
Panel Wipe,
21
Grinder Wipe,
0.19
Shield Wipe,
0.46
Layer 1,
440
Layer 2, 3.4
Layer 3, 1.9
Layer 4, 0.47
Grinder Wipe,
0.16
Shield Wipe,
3.6
Layer 1, 1435
Layer 2, 5.3
Layer 3, 2.0
• Layer 4, 1.4
Grinder Wipe,
0.07
Shield Wipe,
1.5
Layer 1, 348
Layer 2, 2.9
- Layer 3, 1.9
Layer 4, 1.2
Layer 1, 3818
Layer 1, 2865
Layer 1, 4406
Layer 2, 26
Layer 3, 11
Layer 4, 2.2
Layer 3, 507
Layer 4, 466
Layer 4, 0.44
I Panel Wipe
I Grinder Wipe
I Shield Wipe
I Layer 1
I Layer 2
Layer 3
Layer 4
-------
EPA/600/R-20/May 2020
The majority of VX recovered from each coupon (both sealed concrete and limestone) was
obtained in the 1st ground layer sample. Total percent recovery (sum of the masses recovered via
wipe samples taken from the top surface of the coupon following the 24-hour dwell period, the
wheel of the grinder, and from the inside surface of the deflector shield, and from solvent
extraction of all four ground layer samples) averaged only 8.5% (vs the associated spike control
mean) from sealed concrete. Average total percent recovery from limestone was markedly
higher, at 47% of the associated spike control mean recovery. The higher recovery from ground
limestone is consistent with the results of the ground material solvent extraction method
demonstration testing, during which an average 25% recovery of VX from sealed concrete was
obtained following a 24-hour dwell period, compared to an average 95% recovery of VX from
ground limestone after a 24-hour dwell period.
Total mass recovery from the 2nd sealed concrete coupon was markedly higher than the total
mass recovered from the 1st and 3rd sealed concrete coupons. No readily attributable cause was
noted or could be determined for the difference in total recovery from the 2nd coupon versus from
the 1st and 3rd coupons.
The next highest recovery from each coupon (both sealed concrete and limestone) was obtained
from the wipe sample taken from the top surface of the coupon following the 24-hour dwell
period (prior to application of the grinding technology), except in the case of the 2nd limestone
coupon. Markedly higher masses of VX were recovered from extraction of the 3rd and 4th ground
layer samples taken from the 2nd limestone coupon than from the 3rd and 4th ground layer
samples taken from other coupons, suggesting that VX had penetrated more deeply into the 2nd
limestone coupon than into other coupons (or alternatively, that VX that had penetrated deeply
into the 2nd coupon was more amenable to recovery via grinding and solvent extraction than from
similarly deep layers from other coupons). No observations or anomalies were noted during
testing that would explain the higher recoveries from the 3rd and 4th ground layer samples
collected from the 2nd limestone coupon. It was discussed that since limestone is a porous
material, it may be possible that the porosity of the material throughout the full coupon is
inhomogeneous, and areas may exist within the material wherein there are relatively higher and
relatively lower abundances of pores within the material matrix. It was discussed that VX may
have been applied to the 2nd limestone coupon over an area of higher porosity compared to the 1st
and 3rd limestone coupons.
3.3. Physical Removal Efficacy Results - Chemical Stripping
Painted steel and painted hardwood coupons were contaminated with 10 |iL of VX (single 10-|iL
droplet applied in the center of the top surface of the coupon, equivalent to 9.4 mg of VX) and
the VX dwelled on the surface of the coupon for 24 hours. Following the 24-hour dwell period,
the surface of the coupon was sampled via wipe sampling, then chemical stripper was applied.
Following the chemical stripper contact period (total 45 minutes), the paint/coating stripped from
the surface of the coupons was scraped and collected, the stripped coating samples were
extracted with solvent, and repeat wipe samples were taken from the surface of the coupons.
62
-------
EPA/600/R-20/May 2020
Spike control recovery results from the chemical stripping tests are provided in Table 39.
Table 39. Chemical Stripping, Spike Controls
Sample Description
Extract Cone.
(ng/mL)
Mass
(MS)
% Recovery
(vs target)
Avg
(MS)
Avg % vs
Target
St Dev
%RSD
SCI (painted steel test)
423200
8464
84%
SC2 (painted steel test)
493000
9860
98%
8911
88%
823
9.2%
SC3 (painted steel test)
420400
8408
83%
SC1 (painted wood test)
451000
9020
89%
SC2 (painted wood test)
396000
7920
79%
8562
85%
573
6.7%
SC3 (painted wood test)
437300
8746
87%
Positive controls included in tests of the chemical stripping technology consisted of painted steel
and hardwood material coupons that were contaminated with VX and sampled following the
dwell period alongside the test coupons (via wipe sampling), but to which the chemical stripper
was not applied (two wipe samples were collected from the surface of positive controls (pre and
post-wipes, consistent with and alongside the test coupons), but without application of chemical
stripper between the wipe samples, in contrast to the test coupons). Painted steel and painted
hardwood positive control recovery results are provided in Table 40.
Table 40. Chemical Stripping, Positive Control Recovery
Sample Description
Extract Cone.
(ng/mL)
Mass
(MS)
% Recovery
(vs SC avg)
Painted Steel PC Pre-Wipe
332900
7124
80%
Painted Steel PC Post-Wipe
66570
1425
16%
Painted Steel PC Total Mass
NA
8549
96%
Painted Wood PC Pre-Wipe
106100
2271
27%
Painted Wood PC Post-Wipe
69350
1484
17%
Painted Wood PC Total Mass
NA
3755
44%
VX mass recovery and percent recovery results for the pre-stripping and post-stripping wipe
samples and stripped coating solvent extraction samples collected from each painted steel
coupon included during chemical stripping technology physical removal efficacy testing are
provided in Table 41. Mass recovery and percent recovery results for the pre-stripping and post-
stripping wipe samples and stripped coating solvent extraction samples collected from painted
hardwood coupons are provided in Table 42. Wipe sample and stripped coating extraction
sample mass recovery results from all coupons are also summarized in Figure 35. Generally,
greater recoveries of VX were obtained from the painted steel coupons, which is suspected to be
due mostly to migration of VX though the coating layer and into the underlying wood substrate
(consistent with previous findings [12]).
63
-------
EPA/600/R-20/May 2020
Table 41. Chemical Stripping, Painted Steel Recovery
Painted Steel Coupon 1
Painted Steel Coupon 2
Painted Steel Coupon 3
Avg Mass
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs PC)
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recoveiy
(vs PC)
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recoveiy
(vs PC)
Avg Mass
(Mg)
% Recovery
(vs PC)
St Dev
%RSD
Pre-Wipe
226900
4856
68%
249400
5337
75%
194500
4162
58%
4785
67%
591
12%
Post-Wipe
5309
114
8.0%
5607
120
8.4%
6814
146
10%
126
8.9%
17
13%
Stripped Paint
103400
2068
NA
89860
1797
NA
149500
2990
NA
2285
NA
625
27%
Total Mass
NA
7037
82%
NA
7254
85%
NA
7298
85%
7197
84%
140
1.9%
Table 42. Chemical Stripping, Painted Wood Recovery
Painted Wood Coupon 1
Painted Wood Coupon 2
Painted Wood Coupon 3
Avg Mass
Sample Description
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recovery
(vs PC)
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recoveiy
(vs PC)
Extract Cone.
(ng/mL)
Mass
(Mg)
% Recoveiy
(vs PC)
Avg Mass
(Mg)
% Recoveiy
(vs PC)
St Dev
%RSD
Pre-Wipe
158200
3385
149%
74730
1599
70%
61500
1316
58%
2100
93%
1122
53%
Post-Wipe
12030
257
17%
17430
373
25%
14890
319
21%
316
21%
58
18%
Stripped Paint
88490
1770
NA
127200
2544
NA
118500
2370
NA
2228
NA
406
18%
Total Mass
NA
5413
144%
NA
4516
120%
NA
4005
107%
4645
124%
713
15%
64
-------
EPA/600/R-20/May 2020
9000
Chemical Stripping
VX Mass Recovery by Sample Component
Post-Wipe, 1425
8000
7000
-TS 6000
¦a
ai
ai
>
o
5000
4000
3000
2000
1000
Post-Wipe, 114
Pre-
Wipe,
7124
n \m- i ->n Post-Wipe, 146
Post-Wipe, 120 " '
Stripped
Jf
J?
Post-Wipe, 257
Post-Wipe, 1484
Pre-Wipe,
4162
-
Pre-Wipe,
1
l~ 2271
J?
Post-Wipe, 373
Post-Wipe, 319
jr
J?
Post-Wipe
I Stripped Paint
I Pre-Wipe
Sample
Figure 35. Chemical Stripping, VX Mass Recovery by Component
65
-------
EPA/600/R-20/May 2020
3.4. Waste Generation Assessment Results
Grinding
As the physical removal efficacy test results provided in Section 3.2 show, VX was recovered in
the sealed concrete and limestone ground material removed from coupons, thus potentially
requiring treatment (e.g., via incineration) prior to disposal. Application of the grinding
technology to an area of approximately 100 cm2 to a total depth of approximately 1 in. produced
average total ground material masses of 221.8 g from sealed concrete and 209.9 g from
limestone (Tables 35 and 36). While the large majority of VX recovered from coupons by
application of the grinding technology was recovered in the topmost 0.25 in.-thick layer
collected, it cannot be discerned from the data produced during this testing whether lower
detections in deeper layers are due to the absence of VX (i.e., VX did not penetrate past the
topmost 0.25 in. layer), degradation of VX, or an inability to recover VX that is present (given
the low recovery efficiency of the solvent extraction method used to recover VX from ground
concrete and from ground limestone after a 30-minute dwell period). Thus, physical removal to a
greater depth than just the topmost 0.25 in. of material would likely be considered necessary, and
application of a grinding approach similar to the method used here to surface areas much larger
than the approximately 100 cm2 used during this testing would generate a proportionally larger
amount of ground material waste, all of which might require treatment prior to disposal.
Application of the grinding technology using the procedure developed for this testing created
both coarse ground material that was collected at each discrete 0.25 in. depth layer (using the
deflector shield and aluminum collection pan underneath the coupon) as well as fine dust that
was not captured within/redirected by the deflector shield. A wet/dry vacuum was attached to the
top of the test chamber in which grinding operations took place to attempt to draw in and collect
the fine dust, but this approach was only partially successful, leaving a portion of the fine dust
uncollected/uncontained. Attempts were made to use a vacuum shroud to collect the ground
material, but difficulties were experienced that precluded its use. In a field application of a
grinding approach to remove larger areas of sealed concrete or limestone (and likely other porous
material types) contaminated with CWAs or other hazardous compounds, similar fine, airborne
dusts would likely also be generated. Many grinding and cutting technologies incorporate
attachments for application of water (e.g., mist, spray, or similar) to wet the ground materials
produced and reduce airborne dusts, but adequate PPE, including respiratory protection, would
be necessary. Management of these wastes are typically regulated at the state level, and the
appropriate regulatory authorities should be contacted to discuss waste management practices
including waste acceptance criteria for any treatment or disposal facility accepting these
materials.
Chemical Stripper
Klean-Strip® KS-3 Premium finish/paint stripper is a thickened semi-paste that can be applied
via brushing and is intended to cling to vertical surfaces without running or dripping. During the
66
-------
EPA/600/R-20/May 2020
functionality assessment of the Klean-Strip® KS-3 Premium stripper, the stripper was absorbed
into the coating following application and "lifted" the coating from the substrate, leaving behind
minimal liquid residue. The stripped coating layer was then generally easily removed from the
substrate surface via scraping with a 2 in. plastic joint knife (although not completely removed in
the case of the hardwood substrate, as can be seen in Figure 22). Repeated applications of the
stripper would likely be required in some cases to achieve adequately thorough removal of
coatings, and complete removal may not be possible for some substrates (e.g., very porous
surfaces). As can be seen from the physical removal efficacy test results provided in Section 3 .3,
VX was recovered in the stripped coating samples, revealing that contamination is retained in
coatings stripped using the technology and collection and disposal of stripped coatings would
require use of appropriate PPE and post-removal decontamination or waste treatment
methodologies.
3.5. Damage Extent Assessment Results
Grinding
Application of the grinding technology to remove sealed concrete and limestone to a depth of
approximately 1 in. across an area of approximately 100 cm2 left superficial void spaces in
sealed concrete and limestone coupons that were, generally, smooth across the surface and not of
an excessively odd or inaccessible shape (subjective, visual assessment). Refer to Figures 17, 18,
and Figures 36 and 37 below. Based on the findings and observations from this testing, surfaces
to which similar grinding technologies have been applied would be amenable to resurfacing.
Figure 36. Ground Sealed Concrete Coupon
67
-------
EPA/600/R-20/May 2020
Figure 3 7. Ground Limestone Coupon
Chemical Stripper
As can be seen in Figures 21, 22 and Figures 38 and 39 below, a single application of the
chemical stripper technology to paint/primer on low-carbon steel and hardwood appeared to
achieve removal of a large portion of the coating layers without excessive damage to the
underlying substrates.
Figure 38. Stripped Coating, Steel
68
-------
EPA/600/R-20/May 2020
Figure 39. Stripped Coating, Hardwood
Most likely for low-carbon steel, or other similar substrates (e.g., stainless steel), repeated
applications of the stripper could be performed to achieve increasingly higher levels of physical
removal efficacy still without excessive damage to the material, allowing for the surface to be
restored/repainted once acceptable levels of decontamination are achieved. Conversely, repeated
applications of the chemical stripper technology to hardwood (or similar substrates that are not as
inherently resistant to damage as steel substrates) may cause softening, hardening, discoloration,
or other damage. Furthermore, acceptable levels of decontamination by physical removal via
chemical stripping may not be achievable for porous substrates, such as hardwood. In these
cases, portions of the substrate itself may require removal, obviously leading to much greater
levels of damage to the surface which might preclude resurfacing and reuse.
69
-------
EPA/600/R-20/May 2020
4. QUALITY ASSURANCE/QUALITY CONTROL
Quality objectives and performance criteria described in the sections below provide the
requirements used for determining the adequacy of data generated during this project. Methods
were considered acceptable and valid data were assumed if the data quality objectives for the test
measurements were met, and the Technical Systems Audit (TSA), Performance Evaluation (PE),
and data quality audits demonstrated acceptable results, as described in Sections 4.5, 4.6, and
4.7. Accuracy was ensured by the calibration of the instruments. The PE audits further confirmed
that valid data were generated.
4.1. Data Quality Indicators
Data quality indicators and results are provided in Table 43. In general, the data quality indicator
results were acceptable per the Quality Assurance Project Plan (QAPP) including checks of the
measurement methods for temperature, RH, time, volume, and VX recovery from blank samples
and spike controls. Attainment of these data quality indicator results limited the amount of error
introduced into the evaluation results.
Table 43. Data Quality Indicators and Results
Parameter
Measurement
Method
Data Quality Indicators
Results
Temperature
(°C)
HOBO
UX100
Datalogger
Compare against calibrated
thermometer once before testing;
agree ±1°C through 60 minutes.
The HOBO UX100 datalogger used during the project
remained within 0.04°C of the calibrated reference
through one hour.
Relative
Humidity
(%) '
HOBO
UX100
Datalogger
Compare against calibrated
hygrometer once before testing; agree
±10% through 60 minutes.
The HOBO UX100 datalogger used during the project
remained within 5.26% of the calibrated reference
through one hour.
Time
(sec)
Timer/data
logger
Compare to time provided at
NIST.time.gov once before testing;
agree ±2 second/hour.
No difference was observed between the timer and
NIST.time.gov after one hour.
Volume
(PL)
Calibrated
pipette
(CWA
delivery)
Checked for accuracy and
repeatability one time before use by
determining the mass of water
delivered. Acceptable if the range of
observed masses for five droplets is
±10% of expected.
Two pipettes used for VX application were checked:
• 1 to 10 (iL range Gilson pipette, error ranged from
0.17% to 4.83 % of theoretical
• 3 to 25 (iL range Gilson pipette, error ranged from
2.83% to 4.83 % of theoretical
VX in
Procedural
Blank Sample
Extracts
(lig/mL)
Extraction,
LC-MS/MS
Procedural blanks (coupons without
applied agent that are processed
alongside test coupons) should have
less than 1% of the average SC
amount.
No VX outside the stated criteria was measured in any
procedural blank sample extracts throughout testing.
Refer to Section 3 for complete test results.
VX in SC
Extracts
(lig/mL)
LC-MS/MS
The mean of the SCs included with
each test should be within 80% to
120% of the target application and
have a CoV of <30% between
replicates.
Spike control means throughout testing were within
specification.
Refer to Section 3 for complete test results.
4.2. Instrument Calibration
70
-------
EPA/600/R-20/May 2020
4.2.1. Calibration Schedules
Instrumentation needed for the investigation was maintained and operated according to the
quality and safety requirements and documentation of Battelle's HMRC. Except for the LC-
MS/MS, all instruments utilized during the project were calibrated as stipulated by the
manufacturer or, at a minimum, annually. The LC-MS/MS was calibrated as described in Section
4.2.2. Table 44 provides calibration schedules for instruments that were used during the
evaluation.
Table 44. Equipment Calibration Schedule
Equipment
Frequency
Calibrated pipette and repeating
dispenser/syringe
Prior to the investigation and annually thereafter.
Calibrated UX100 HOBO
Hygrometer/Thermometer
Prior to the investigation by the manufacturer.
Timer
Prior to the investigation by the manufacturer.
LC-MS/MS
Beginning of each batch of test samples (calibration curve) and a calibration
verification standard after every ten samples.
4.2.2. LC-MS/MS Calibration
Neat VX was used to create calibration standards (concentrations corrected for percent purity)
encompassing the appropriate analysis range. The expiration date for VX calibration standards
was six months. The expiration date for VX continuing calibration verifications (CCVs) was one
month. A seven (7)-point calibration for VX was used with a lower calibration level of 0.010
ng/mL and an upper limit of 2.0 ng/mL. A linear or quadratic regression (specified in the raw
data product) was used to describe the data with 1/x2 weighting. The origin was not included for
regression. The coefficient of determination (r2) from the regression analysis of the calibration
standards was required to be > 0.990. Limits were also placed on the percent bias (Equation 4)
observed in the standards.
Bias = x 100% (4)
where: Ev = expected value from calibration curve
Ov = observed value from standard
The percent bias for the low standard was required to be less than or equal to 25%, and the
percent bias for the remaining standards was required to be less than or equal to 15%. The signal-
to-noise ratio of the lowest calibration standard was required to be approximately 3:1 at
minimum. Retention time for each target compound (VX) and IS in each injection was reviewed
to confirm that it was within ±0.1 minutes of the retention time for the same components in the
mid-level calibration standard. Solvent blank and double blank samples were included during
analytical runs and were analyzed to confirm that no VX carryover was occurring, and no
significant analyte signal was originating from the IS. Solvent blank sample analysis results were
required to be below the value of the lowest calibration standard.
Independently prepared CCV standards were analyzed prior to sample analysis, following every
ten test/control samples (not including blanks), and at the end of each set of samples. Two CCV
71
-------
EPA/600/R-20/May 2020
concentrations were used, one of which was equal to the low calibration standard (0.010 ng/mL)
and the other within the calibration range (1.0 ng/mL). CCV response was required to be within
25% of the nominal concentration for the 0.010 ng/mL CCV and within 15% of the nominal
concentration for the 1.0 ng/mL CCV.
Calibration standards and CCVs were matched to the samples undergoing analysis as closely as
possible. For example, IPA samples prepared for analysis by a 10-fold dilution in water were
quantitated by standards and CCVs prepared in 10% IPA.
The area of VR IS in the test samples was compared to that in the nearest passing calibration
standard or passing CCV. VR area in the test samples was required to fall within 50% to 200%
of the area of the IS in the calibration standard or CCV to which it was compared (criteria per
EPA Method 8000D [14]). It was assumed that any test sample matrix would affect analysis of
VX and VR IS in a similar manner. Given that assumption, IS response variability within the
range of 50% to 200% of that of the nearest passing calibration standard or CCV was considered
acceptable and IS was assumed to be properly compensating for similar effects on VX response
due to sample matrices.
Table 45 summarizes LC-MS/MS analysis performance parameters and acceptance criteria.
Table 45. LC-MS/MS Performance Parameters and Acceptance Criteria
Parameter
Criterion
Coefficient of determination (r2)
>0.990
% Bias for the lowest calibration standard
<25%
% Bias for remaining calibration standards (except lowest standard)
<15%
Solvent blank samples
< lowest calibration standard
% Bias for the lowest CCV
<25%
% Bias for remaining CCVs (except lowest CCV)
<15%
Signal-to-noise ratio for the lowest calibration standard
Minimum of 3:1
Retention time for target compound and IS
±0.1 min. as same compounds in mid-level
calibration standard
VR IS area in samples
50% to 200% area of nearest passing calibration
standard or passing CCV
4.3. Sample Handling and Custody
At all times during the project, protocols required by the U.S. Army were followed in the
movement and use of VX and Research, Development, Test, and Evaluation (RDT&E) Dilute
Solutions (RDS) within the HMRC. CoC forms were used to ensure that test samples generated
during the project were traceable throughout all phases of testing.
4.4. Technical Systems Audit
A Quality Assurance (QA) Officer performed a TSA during the VX depth penetration method
demonstration test. The purpose of the TSA was to ensure that testing was performed in
accordance with the QAPP. The QA Officer reviewed the investigation methods, compared test
procedures to those specified in the QAPP (and the associated amendments), and reviewed data
acquisition and handling procedures. The QA Officer did not identify any findings that required
corrective action.
72
-------
EPA/600/R-20/May 2020
4.5. Performance Evaluation Audits
PE audits, summarized in Table 46, addressed those reference measurements that factored into
the data used in quantitative analysis during the project, including volume and time
measurements and LC-MS/MS calibration and performance. The volume of VX dispensed
correlated directly with the mass of VX in the wipe, coupon layer, and waste sample extracts.
Volume of solvent used to extract samples directly impacted measured concentrations. The
measured time that VX was allowed to remain in contact with the coupons directly influenced
depth of VX penetration and extent of VX spread. Calibration of the LC-MS/MS and IS recovery
provided confidence that the analysis system was providing accurate data.
Temperature and RH were measured and recorded on each day of testing, but not monitored or
controlled. Therefore, no PE audit of these parameters was performed. See Attachment A for a
summary table of measured temperature and RH ranges.
Table 46. Performance Evaluation Audit Results
Parameter
Audit Procedure
Required
Tolerance
Results
Volume
(mL, |xL)
Pipettes were checked for accuracy and
repeatability one time before use by
determining the mass of water delivered.
The pipette was acceptable if the range of
observed masses for five droplets is
±10% of expected.
±10%
Two pipettes used for VX application were
checked:
• 1 to 10 (iL range Gilson pipette, error ranged
from 0.17% to 4.83 % of theoretical
• 3 to 25 (iL range Gilson pipette, error ranged
from 2.83% to 4.83 % of theoretical
Time (sec)
Compare to time provided at
NIST.time.gov once before testing; agree
±2 second/hour.
±2 sec/hour
No difference was observed between the tinier
and NIST.time.gov after one hour.
VX in SC
Extracts
(Hg/mL)
Use LC-MS/MS to determine mass of
agent delivered as a 2- or 10-(iL droplet
into extraction solvent and compare to
target application level.
>80% of spike
target; < 120% of
spike target; <
30% CoV
Spike control means throughout testing were
within specification.
LC-MS/MS
VX
Calibration
Standards (%)
Verify all standards and CCVs used to
calibrate and confirm calibration of the
All standards and CCVs were within
specification for all reported data.
LC-MS/MS system used for analysis fall
within the requirements provided in
Section 4.2.2.
Refer to Table 45
4.6. Data Quality Audit
Validation of the data included verification of the completeness of the data, compliance with the
acceptance criteria in the QAPP, recalculation checks, and tracing of the data from instrument
outputs through the final report. The data were reviewed to verify completeness and ensure the
data were valid and met the acceptance criteria of the QAPP. One hundred percent (100%) of all
data was reviewed prior to use in calculations, and data manipulation was completed before the
data quality audit.
The QA Manager, operating independently of the laboratory testing effort, audited
approximately 10% of the data generated during testing. The QA Manager traced the data from
initial acquisition through reduction and to final reporting. All data analysis calculations were
checked. Through the data quality audit, the TSA, and the review of the draft and final reports,
73
-------
EPA/600/R-20/May 2020
the Battelle QA Manager ensured that data generated during the project were valid, meeting the
requirements of the QAPP.
74
-------
EPA/600/R-20/May 2020
5. SUMMARY
The primary objective of this project was to quantitatively evaluate the efficacy of select
technologies and determine the application conditions/methods necessary to decontaminate
CWA-contaminated porous materials and permeable coatings through physical removal of the
contaminated portions of the materials.
Prior to testing, literature searches were performed to identify technologies that could be used to
physically remove contaminated portions of porous materials and/or permeable coatings while
simultaneously minimizing damage to the materials and generation of hazardous wastes. From
the literature search results, grinding and chemical stripping were selected for evaluation.
Grinding was evaluated for efficacy in removal of contaminated portions of sealed concrete and
limestone and chemical stripping was evaluated for efficacy in removal of contaminated coatings
from low-carbon steel and hardwood.
Bench scale studies were performed using neat VX as the challenge CWA. The porous materials
and permeable coatings were contaminated with VX and the VX was dwelled on the surface of
the materials for a period of 24 hours to allow the VX to penetrate the materials. Following the
24-hour dwell period, the porous material and permeable coating coupon surfaces were sampled
via wipe sampling to quantify residual, transferable VX. Following wipe-sampling, the physical
removal technologies under test were applied to remove the contaminated portions of the
material coupons. Grinding was used to remove portions of sealed concrete and limestone at
discrete 0.25 in.-thick depth layers. Chemical stripper was applied to the coated steel and
hardwood coupons to remove the paint/primer layers. Ground material removed from sealed
concrete and limestone and coatings stripped from steel and hardwood were extracted with
solvent, and extracts were analyzed via LC-MS/MS to quantify VX recovered from the removed
materials. The surface of steel and wood coupons were also sampled via wipe sampling again
following stripping.
A method independent of the selected physical removal technologies was also developed and
used for dissection of porous materials (sealed concrete and limestone) to quantify the extent of
VX penetration into the porous materials as a function of depth. The core sampling approach
involved excision of 1.5 in. diameter cylindrical core samples from coupons of sealed concrete
and limestone that were contaminated on the top surface with VX. Following a 24-hour VX
dwell period, the contaminated surfaces of the core samples were sampled via wipe sampling.
The core samples were then dissected into discrete 0.25 in.-thick layer samples that were
extracted individually with solvent, and extracts were analyzed via LC-MS/MS. Analysis results
were then used to determine the depth to which VX had penetrated the cores over the course of
the 24-hour dwell period, based on the amount of VX recovered from each core layer sample.
Prior to physical removal efficacy testing and VX depth penetration testing (using the core
sampling approach), the methods used for solvent extraction of coupon, wipe, and waste
samples, and for wipe sampling of coupon surfaces were evaluated. Results of methods
demonstration testing are summarized in Figures 40, 41, and 42.
75
-------
EPA/600/R-20/May 2020
Solvent Extraction Method Demonstration
Average Percent Recovery
(SCs vs target; PCs vs SCs; test samples vs PCs)
Sample
Figure 40. Solvent Extraction, Average Percent Recovery
Wipe Sampling Method Demonstration
Average Percent Recovery
(SCs vs target; PCs vs SCs; test samples vs PCs)
130%
*20% Tt%
£• 110%
> 100%
90% 79%
* 80%
§ 70% ¦ 58"%
60% H H _
50% ¦ 42%
40% ¦ ¦ ¦ H
I ¦ ¦ ¦ ¦
10% 0.77% 0.41%
0%
to in in in in in
¦7; (D
-------
EPA/600/R-20/May 2020
£•
ai
>
o
u
ai
cc
Wipe Sampling Method Demonstration
Average Percent Recovery
(SCs vs target; PCs vs SCs; cutting dust vs PCs; stripped paint vs SCs)
225% 207%
200%
175%
150%
125%
100%
75%
a>
a- 50%
So 25%
ra
*_
a>
>
<
0%
18.
1
4%
H
3/0 rjr
1 51% 1
1 I 1
•% —
ri
— 9%
H
1% —
ri
h
82%
hrn
Sample
Figure 42. Waste Sampling, Average Percent Recovery
VX recovery criteria in coupon samples (ground material and core layer samples), wipe samples,
and waste samples (cutting dust and stripped coating samples) were 70% to 120% of the
associated spike control mean recovery with an RSD between replicates of < 30%. Acceptable
recoveries from waste samples were achieved using the solvent extraction method demonstrated
during the project (submersion in IP A with sonication), but generally lower recoveries were
obtained when using the method to recover VX from coupon samples. Generally lower
recoveries (< the 70% acceptance criterion) were obtained also using the wipe sampling method
that was demonstrated. The lower recoveries are generally consistent, though, with previous
studies that describe similar difficulties with recovery of CWAs from similar materials [3] as
well as degradation of CWAs in material matrices [2,6,7],
Results of the VX depth penetration assessment using the core sampling approach are
summarized in Figure 43. Generally, total recoveries from core samples were low compared to
the associated spike control mean recoveries, measuring only 11% average recovery from
limestone cores and 14% average recovery from sealed concrete cores. The majority of VX
recovered from each core sample was obtained from solvent extraction of the 1st layer sample
(the "topmost" layer of the core that was initially contaminated with VX). The next highest
recovery from each core sample was obtained via the wipe sample taken from the top surface of
the core. VX mass recoveries suggest that either VX does not penetrate into the materials past
the topmost approximately 0.25 in. depth (via gravity-driven diffusion over the course of 24
hours), or VX becomes increasingly unrecoverable or degrades as it penetrates farther than
approximately 0.25 in. into the materials. The low recoveries are also consistent with [3] and
77
-------
EPA/600/R-20/May 2020
potentially attributable to previously implied degradation of VX on concrete due to the presence
of basic catalytic sites [2],
78
-------
EPA/600/R-20/May 2020
VX Depth Penetration
VX Mass Recovery by Core Sample Component
Wipe, 39 ug
Wipe,
37 ug
Wipe, 58 ug
Layer 1, 863 ug
Blade, 0.01 ug
Layer 1, 854 ug
Dust 2,0.36 ug
Dust 2, 0.01 ug
Layer 2, 0.08 ug
Layer 2, 0.02 ug
Dust 3,0.002 ug
Dust 3,0.002 ug
Layer 3, 0.02 ug
Layer 3, 0.01 ug
Dust 4,0.002 ug
Dust 4, 0.002 ug
Layer4, 0.01 ug
Layer4, 0.01 ug
Dust 5, 0.002 ug
Dust 5,0.002 ug
Layer5, 0.002 ug
Layers, 0.002 ug
Blade, 0.01 ug
Wipe, 61 ug
Wipe, 42 ug
Layer 1, 1133 ug
Dust 2, 0.02 ug
Layer 2, 0.01 ug
Dust 3, 0.011 ug
Layer 3, 0.02 ug
Dust 4, 0.008 ug
Layer4, 0.01 ug
Dust 5,0.006 ug
Layer5, 0.006 ug
Blade, 0.005 ug
Layer 1, 956 ug
Dust 2, 0.01 ug
Layer 2, 0.01 ug
Dust 3,0.005 ug
Layer 3, 0.01 ug
Dust 4, 0.003 ug
Layer4, 0.003 ug
Dust 5, 0.007 ug
Layer5, 0.002 ug
Blade, 0.002 ug
i Wipe
I Layer 1
Dust 2
Layer 2
[ Dust 3
i Layer 3
I Dust 4
I Layer 4
I Dust 5
iLayer 5
I Blade
J
\e-
Sample
J
&
&
Figure 43. VX Depth Penetration Assessment, VX Mass Recovery by Component
79
-------
EPA/600/R-20/May 2020
As during the VX depth penetration assessment, the major portion of the VX recovered from
each sealed concrete and limestone coupon via application of the grinding technology was
obtained in the 1st ground layer sample (the topmost 0.25 in. of the material, to which the VX
challenge was applied). Total percent recovery averaged only 8.5% (versus the associated spike
control mean) from sealed concrete. Average total percent recovery from limestone was
markedly higher, at 47% of the associated spike control mean recovery. The higher recovery
from ground limestone is consistent with the results of ground material solvent extraction
method demonstration testing, during which an average 25% recovery of VX from sealed
concrete was obtained following a 24-hour dwell period, compared to an average 95% recovery
of VX from ground limestone after a 24-hour dwell period. After the 1st ground layer sample,
recoveries then decreased sharply to less than 1% of the spike control mean recovery in all cases
except that of the 2nd limestone coupon, in which recoveries from the 3rd and 4th 0.25 in. ground
layer samples remained as high as 5.8% and 5.4%, respectively. Results from the assessments
that were conducted of physical removal efficacy via grinding are summarized in Figure 44.
Grinding
VX Mass Recovery by Sample Component
4500.00
4000.00
3500.00
"SP 3000.00
Panel Wipe, 31
¦ Grinder Wipe, 4.9
¦ Shield Wipe, 3.2
2500.00
TS
0)
O
a 2000.00
J 1500.00
1000.00
500.00
0.00
Panel
Wipe, 37
Grinder
Wipe, 0.19
Shield
Wipe, 0.46
Layer
1,440
• Layer 2, 3.4
• Layer 3, 1.9
Layer 4, 0.47 I
Panel
Wipe, 34
Grinder
Wipe, 0.16
Shield
Wipe, 3.6
Layer 1, 1435
Layer 2, 5.3
Layer 3, 2.0
Layer 4, 1.4
Panel
Wipe, 21
Grinder
Wipe, 0.07
Shield
Wipe, 1.5
Layer 1, 348
|- Layer 2, 2.9
Layer 3, 1.9
Layer 4, 1.2
Layer 1, 3818
Panel
Wipe, 20
Grinder
Wipe,
3.9
Shield
Wipe,
Layer 1, 2865
Layer 2, 26
¦ Layer 3, 11
Layer 4, 2.2
Layer 3, 507
Layer 4, 466
J*
J*
J*
jr
J*
jr
r/
xO
<»
xCF
*cF
~
iF
Sample
v
¦ Panel Wipe, 12
• Grinder Wipe, 4.6
Shield Wipe, 8.0
Layer 1, 4406
¦ Panel Wipe
¦ Grinder Wipe
¦ Shield Wipe
- Layer 2, 14 ® Layer 1
Layer 3, 0.98 ¦ Layer 2
Layer 4, 0.44 Laye|_ 3
¦ Layer 4
Figure 44. Grinding, VX Mass Recovery by Component
80
-------
EPA/600/R-20/May 2020
It cannot be discerned from the data whether lower detections in deeper layers are due to the
absence of VX (i.e., VX did not penetrate past the topmost 0.25 in. layer), degradation of VX, or
an inability to recover VX that is present. Thus, in a field-application of grinding to remove
contamination, physical removal to a greater depth than just the topmost 0.25 in. of material is
likely necessary. While the data suggest that VX contamination in porous materials can be
removed via application of grinding to remove contaminated portions of the materials, the
generally low total recoveries as well as the relatively higher recoveries from deeper layers from
the 2nd limestone coupon suggest that the depths necessary for removal to safe (i.e.,
nonhazardous) levels can be inconsistent.
Wipe sampling and stripped coating extraction sample mass recovery results from all painted
steel and painted wood coupons are summarized in Figure 45. Generally, greater recoveries of
VX (as compared to mean positive control recoveries) were obtained from the painted steel
coupons.
9000
8000
7000
"m 6000
£ 5000
dl
>
O
4000
U)
U)
ro
§ 3000
2000
1000
0
Post-Wipe, 1425
Chemical Stripping
VX Mass Recovery by Sample Component
Post-Wipe, 114
P re-
Wipe,
7124
Post-Wipe, 120 Post-Wipe, 146
Stripped
Stripped
i— Paint,
i 2068
j— Paint,
1797
Pre-
j— Wipe,
1 5337
Pre-
/— Wipe,
4856
J
J
Post-Wipe, 257
Post-Wipe, 373
Post-Wipe, 319
jv
J?
J?
Post-Wipe
I Stripped Paint
I Pre-Wipe
*y
~
~
Sample
«er
sy
Figure 45. Chemical Stripping, VX Mass Recovery by Component
81
-------
EPA/600/R-20/May 2020
Markedly less VX was recovered from wipe samples taken from the coupons following removal
of the paint/coating layer from the steel substrate via application of chemical stripper, indicating
that the majority of the VX contamination was removed by the first (pre-stripping) wipe and by
removal of the permeable coating via application of the stripper. The data suggest that
remediation of VX-contaminated painted/coated steel via a combination of solvent wipe
sampling (i.e., wipe sampling with solvent-soaked wipes) and removal of the paint/coating via
chemical stripping may be possible. This assumes that VX does not permeate into the steel
substrate (given that it's a non-porous, relatively inert material) A repeated solvent wipe
sampling and application of the chemical stripper may be required, depending on the required
decontamination level. The lower total recoveries from painted wood samples as well as the
higher recoveries from post-stripping wipe samples taken from the wood coupons suggest that
VX may have permeated through the paint/coating layer and into the underlying permeable wood
substrate. Such residual VX contamination could potentially pose contact or vapor hazards later
if the VX diffuses back to the surface of the wood or if the wood is cut, ground, or otherwise
manipulated.
The grinding technology and the core sampling approach were applied to both porous materials
(sealed concrete and limestone). Application of both physical rem oval/sampling methods
produced depth layer samples at 0.25 in. increments into the material samples to which they were
applied. Also, prior to application of both the grinding technology and the core sampling
approach, the surfaces of coupons/cores were sampled via wipe sampling.
Figure 46 provides VX mass recovered by the surface wipe and in each successive depth layer
sample collected from sealed concrete and limestone coupons/core samples. As Figure 46 shows,
the largest amount of VX recovered from both material types using both removal/sampling
methods was obtained in the first 0.25 in.-thick depth layer sample (either core layer sample or
grinding layer sample). Based on visual assessment of Figure 46, it appears that generally similar
total amounts of VX were recovered from sealed concrete and limestone core samples via
application of the core sampling approach and from sealed concrete coupons via application of
the grinding technology, but significantly greater VX was recovered from limestone coupons via
application of the grinding technology.
Both grinding and chemical stripping, as applied to porous materials and permeable coatings
during this project, generate wastes that retain hazardous contaminants and would require
collection and handling using appropriate PPE and decontamination to acceptable levels prior to
disposal. Further, dust mitigation will be required since small dust particles carrying agent
contamination would likely become redistributed in the environment (and potentially transfer to
other materials). Some of the potentially contaminated particulate matter may become an
inhalation hazard.
The porous materials and permeable coating substrates evaluated during this project would likely
be amenable to resurfacing and reuse following application of the grinding and chemical
stripping technologies to remove contaminated portions of the materials, except in the case of the
82
-------
EPA/600/R-20/May 2020
hardwood permeable coating substrate which could be at risk of excessive damage if repeated
chemical stripping applications or additional physical removal methods (beyond removal of the
coating) are required to achieve acceptable levels of decontamination..
83
-------
EPA/600/R-20/May 2020
Core Sampling and Grinding Comparison
Limestone and Sealed Concrete
VX Mass Recovery by Sample Component
mill
I
e. '
f<^ f°"" rS* rS* rf" rf* rf*
^ c/ c/ c/ CO-9 c/ c/
~ c/' c/'
/ ^ ^ ^
s „-y
c,e." c?- c,®
C° ° ^
\> \> \>
/" / /
^ .& J?
Material Sample, Method
Figure 46. Limestone and Sealed Cone. Recovery, Grinding v.s Core Sampling Comparison
-------
EPA/600/R-20/May 2020
6. REFERENCES
1. NIOSH Estimating the Permeation Resistance of Nonporous Barrier Polymers to Sulfur Mustard
(HD) and Sarin (GB) Chemical Warfare Agents Using Liquid Simulants.
https://www.cdc .gov/niosh/docs/2008-141/pdfs/2008-14 l.pdf?id=10.26616/NIOSHPUB2008141.
Last accessed April 27, 2020.
2. Willis M.P., Mantooth B.A., Lalain T.A. Novel Methodology for the Estimation of Chemical Warfare
Agent Mass Transport Dynamics, Part I: Evaporation. The Journal of Physical Chemistry C 2012,
116 (1), 538-545.
3. Grissom T.G., Sirrine G.M., Long T.E., Esker A.R., Morris J.R. Interaction parameters for the uptake
of sulfur mustard mimics into polyurethane films. Progress in Organic Coatings 2017, 107, 14-17.
4. Varady M.J., Pearl T.P., Bringuier S.A., Mantooth B.A. Vapor emission from porous materials with
diffusive transport in the solid-phase. International Journal of Heat and Mass Transfer, 2017, 114,
758-768.
5. Mantooth B.A., Willis M.P., Procell L., Davies J. Transport and Reactivity of Decontaminants to
Provide Hazard Mitigation of Chemical Warfare Agents from Materials. ECBC-TR-1383; U.S. Army
Edgewood Chemical Biological Center: Aberdeen Proving Ground, MD, 2016
6. Williams J.M., Rowland B., Jeffery M.T., Groenewold G.S., Appelhans A.D., Gresham G.L., and
Olson J.E. Degradation Kinetics of VX on Concrete by Secondary Ion Mass Spectrometry. Langmuir
2005,21,2386-2390.
7. Wagner G.W., O'Connor R.J., Edwards J.L., and Brevett C.A.S. Effect of Drop Size on the
Degradation of VX in Concrete. Langmuir 2004, 20, 7146-7150.
8. Leif R., Koester C., Gruidl J., Koester, C. Wipe Sampling Collection Efficiencies and Holding Time
Studies. Lawrence Livermore National Laboratory, LLNL-TR-573693, 2012.
9. Columbus I., Waysbort D., Marcovitch I., Yehezkel L., Mizrahi D.M. VX Fate on Common Matrices:
Evaporation versus Degradation. Environmental Science & Technology 2012, 46, 3921-3927.
10. Brevett, C.A.S., Sumpter, K.B., Pence, J., Nickol, R.G., King, B.E., Giannaras, C.V., Durst, H.D.
Evaporation and Degradation of VX on Silica Sand. The Journal of Physical Chemistry C 2009, 113,
6622-6633.
11. Oudejans L., Chappie D., See D., Lordo R. Natural Attenuation of the Persistent Chemical Warfare
Agent VX on Porous and Permeable Surfaces. U.S. Environmental Protection Agency, Washington,
DC, 2017, EPA/600/R-17/186.
12. U.S. Environmental Protection Agency. Fate and Transport of Chemical Warfare Agents VX and HD
across a Permeable Layer of Paint or Sealant into Porous Subsurfaces. U.S. Environmental Protection
Agency, Washington, DC, 2016, EPA/600/R-16/173.
13. Varady, M.J., Riley, P.C., Mantooth, B.A., Schenning, A.M., Fouse, J.C., Pearl, T.P. Evaporation and
Degradation of A Sessile Droplet of VX on an Impermeable Surface. ECBC-TR-1478; U.S. Army
Edgewood Chemical Biological Center: Aberdeen Proving Ground, MD, 2017.
14. Method 8000D. Determinative Chromatographic Separations. Revision 4, July 2014, Final Update V
to the Third Edition of the Test Methods for Evaluating Solid Waste, Physical/Chemical Methods,
EPA publication SW-846.
85
-------
EPA/600/R-20/May 2020
ATTACHMENT A - ENVIRONMENTAL DATA
Activity
Temperature
Range (°C)
RH Range (%)
Solvent Extraction, Wipe Sampling, and
Waste Sampling Method Development
21.0-23.0
15-16
VX Depth Penetration Assessment
22.0-24.0
40-45
Physical Removal: Concrete Grinding
21.5-22.5
40-45
Physical Removal: Limestone Grinding
21.0-22.0
40-45
Physical Removal: Chemical Stripping
22.0-23.0
42-47
86
-------
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
------- |