EPA/600/R-18/313 | October 2018
www.epa.gov/homeland-security-research
United States
Environmental Protectior
Agency
oEPA
Wet-Vacuum-Based Surface
Sampling Method for
Chemical Agents

Office of Research and Development
Homeland Security Research Program

-------
This page left intentionally blank

-------
Wet-Vacuum-Based Surface Sampling Method
for Chemical Agents
Authors:
Lukas Oudejans, Ph.D.
National Homeland Security Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Barbara Wyrzykowska-Ceradini, Ph.D.
Abderrahmane Touati, Ph.D.
Jacobs Technology, Inc.
Research Triangle Park, NC 27709
Eric Morris
Alexander Korff
Science Systems Applications, Inc.
Hampton, VA 23666
i

-------
Disclaimer
The U.S. Environmental Protection Agency (EPA), through its Office of Research and Development's
National Homeland Security Research Center (NHSRC), funded and managed this investigation through
Contract No. EP-C-15-008, work assignments 0-073 through 2-073 with Jacobs Technology, Inc. (Jacobs)
The report also includes the method development experimental data generated under Contract No. EP-C-
09-027 with ARCADIS, Inc. This report has been peer and administratively reviewed and has been
approved for publication as an Environmental Protection Agency document. It does not necessarily reflect
the views of the Environmental Protection Agency. No official endorsement should be inferred. This report
includes photographs of commercially available products. The photographs are included for purposes of
illustration only and are not intended to imply that EPA approves or endorses the product or its
manufacturer. EPA does not endorse the purchase or sale of any commercial products or services.
Questions concerning this document or its application should be addressed to the principal investigator of
this study:
Lukas Oudejans, Ph.D.
Decontamination and Consequence Management Division
National Homeland Security Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Mail Code E343-06
109 T.W. Alexander Drive
Research Triangle Park, NC 27711
Telephone No.: (919) 541-2973
E-mail Address: Oudejans.Lukas@epa.qov

-------
Acronyms and Abbreviations
2-CEPS	2-chloroethyl phenylsulfide
AEMD	Air and Energy Management Division (EPA)
ANOVA	analysis of variance
CC	continuous calibration
CF	conversion factor
cm	centimeter(s)
cm2	square centimeter(s)
cm3	cubic centimeters)
CMAD	Consequence Management and Advisory Division (EPA)
CRM	Certified Reference Material
CS	control spike
CT	contact time
CWA	chemical warfare agent
D	depth
DCM	dichloromethane
DCMD	Decontamination and Consequence Management Division (EPA)
DF	film thickness
Dl	deionized
DLT	dirty [sample] liquid tank
DQI	data quality indicator
DUP	duplicate injection
EC	end check
EPA	U.S. Environmental Protection Agency
ft2	square foot
g	gram(s)
GC	gas chromatography
GC/FID	gas chromatography/flame ionization detector
GC/MS	gas chromatography/mass spectrometry
h	hour(s)
H	height
HD	sulfur mustard
HPLC	high performance liquid chromatography
HSRP	Homeland Security Research Program (EPA)
HT	holding time
ICAL	initial calibration
ICV	initial calibration verification
IPA	isopropyl alcohol
IR	infrared
IS	internal standard
ISO	International Organization for Standardization
L	liter(s)
LB	laboratory blank

-------
LCS
LCSD
LLE
LOQ
LT
m
M
m2
MAC
mg
M9
min
mL
ML
NHSRC
NIOSH
NIST
NMAM
OEM
OLEM
ORD
OSL
PB
PI
PW
QC
R2
RLV
RPD
RSD
RTP
s
SA
SD
SEE
SL
slpm
S/N
SS
TAT
TC
TEP
TIC
TOT
laboratory control sample
laboratory control sample duplicate
liquid-liquid extraction
limit of quantitation
lapse time
meter(s)
molar
square meter(s)
multiarea composite
milligram(s)
microgram(s)
minute(s)
milliliter(s)
microliter(s)
National Homeland Security Research Center (EPA)
National Institute for Occupational Safety and Health
National Institute of Standards and Technology
NIOSH Manual of Analytical Methods
Office of Emergency Management (EPA)
Office of Land and Emergency Management (EPA)
Office of Research and Development (EPA)
Organic Support Laboratory
procedural blank
principal investigator
plywood
quality control
coefficient of determination
reporting limit verification
relative percent difference
relative standard deviation
Research Triangle Park, North Carolina, USA
second (s)
single [medium-size] area
standard deviation
Senior Environmental Employee
sampling liquid
standard liter(s) per minute
signal-to-noise ratio
stainless steel
turnaround time
test coupon
triethylphosphate
toxic industrial chemical
total operational time
iv

-------
vinyl flooring
width
wetting tank
sampling efficiency

-------
Acknowledg men ts
This research effort is part of the U.S. Environmental Protection Agency's (EPA's) Homeland Security
Research Program (HSRP) to develop novel surface sampling approaches. Here, a wet-vacuum technology
is considered for the sampling of various toxic industrial chemicals from contaminated building material
surfaces. This methodology was optimized for sampling of chemicals with various solubility in water, using a
discrete and a composite sample collection approach. The optimized wet-vacuum sampling technology,
which can be applied to sample for chemicals over a large area, was also compared to standard wipe-based
sampling methods. The results of this work would inform responders, governments, and health departments
in their guidance development for sampling recommendations for medium to large surface areas
contaminated with toxic chemicals.
This effort was directed by the principal investigator (PI) from the Office of Research and Development's
(ORD's) National Homeland Security Research Center (NHSRC), with support of a project team consisting
of staff from across EPA. The contributions of the following individuals have been a valued asset
throughout this effort:
EPA Project Team
Lukas Oudejans, ORD/NHSRC (PI)
Lawrence Kaelin, Office of Land and Emergency Management (OLEM)/Office of Emergency
Management (OEM)/Consequence Management Advisory Division (CMAD)
Leroy Mickelsen, OLEM/OEM/CMAD
Sang Don Lee, ORD/NHSRC
Jacobs Technology Inc. Team
Barbara Wyrzykowska-Ceradini (Jacobs Technology, Inc.)
Abderrahmane Touati (Jacobs Technology, Inc.)
Eric Morris (Science Systems Application, Inc)
Alexander Korff (Science Systems Application, Inc)
U.S. EPA Technical Reviewers of Report
Romy Campisano, ORD/NHSRC
Ahmed Hafez OLEM/OEM/CMAD
U.S. EPA Quality Assurance
Eletha Brady-Roberts, ORD/NHSRC
Ramona Sherman, ORD/NHSRC
Jacobs Technology Inc. Quality Assurance
Zora Drake-Richman
Matthew Allen
U.S. EPA Editorial Review
Joan Bursey, Senior Environmental Employee (SEE)
vi

-------
Executive Summary
A large-area accidental or intentional chemical release may require extensive environmental
sampling for hazard characterization and mapping, and, after the decontamination is completed, may
require reliable post-decontamination clearance sampling to assess if surface cleanup goals were met. Such
complex environmental sampling may necessitate collection of large numbers of samples Traditional
sampling methods can significantly hinder the remediation process because they are time- and labor-
intensive.
This study investigated a novel surface sampling approach based on the use of a commercially-
available wet-vacuum cleaning method as a sampling mechanism. This novel surface sampling approach
utilizes the ability of one piece of equipment to dispense a solution (the wetting agent) onto a surface and to
vacuum the applied wetting agent, including any dissolved chemical, for collection into the same system.
This approach was investigated for collection of various classes of chemicals from indoor building surfaces.
Target chemicals with varying degrees of solubility in water were collected from nonporous to porous and
permeable substrates using water or organic solvent wetting agents. The performance of an optimized wet-
vacuum sampling method using a commercial off-the-shelf wet-vacuum unit was evaluated for multivariate
(chemical and surface type, surface contamination level, and wetting agent type) sampling of medium-size
(approximately 1000 cm2) and large-size (approximately 5000 cm2) areas and compared to existing wipe-
based sampling methods and/or modifications thereof.
The main findings of this study are:
•	Wet-vacuum-based methods, utilizing a commercially available cleaner and isopropyl alcohol (IPA)
wetting solvent provide a better than 75% recovery for sampling of various classes of chemicals
with varied solubility in water at tens of milligrams per square meter (mg/m2) surface concentrations.
•	Wet-vacuum method performance is lower for collection of chemicals from semiporous materials
(wood, vinyl), compared to nonporous materials, and for sampling of surfaces contaminated with
lower microgram per square meter (jjg/m2) contamination levels.
•	The efficiency of aqueous wetting agent-based wet-vacuum sampling is affected by the solubility of
the chemical in water. The addition of surfactant improves recovery of selected water-insoluble
chemicals but generally does not improve the sampling efficiency at lower contaminant surface
concentrations.
•	In comparison to wipe-based sampling methods, the wet-vacuum methodology with IPA as a
wetting solvent offers eight- to tenfold improvement in turnaround time needed to collect and
prepare surface sample for analysis. The aqueous wetting agent methods offers an approximately
twofold reduction of the turnaround time, as compared to wipe-based methodology.
Results indicate that the wet-vacuum sampling method was very efficient for collection of chemicals
from nonporous substrates based upon only a fraction of the chemical mass recovered remaining on the
surfaces post-sampling. However, the vacuum cleaner components were confirmed to contain the chemical
after use suggesting the wet-vacuum units should not be considered for reuse and should be handled as
contaminated waste.
The main limitations of this new technique are: (1) its limited applicability to sampling of highly
porous surfaces (like wipe-based surface sampling) and (2) low sample recoveries when the level of the
chemical surface contamination is significantly below the 1-10 mg/m2 surface concentration level.
vii

-------
Depending on the surface decontamination threshold, this methodology may need to be further optimized
for application to post-contamination sampling. Despite the significant reduction of turnaround time, as
compared to conventional wipe-based sampling, the wet-vacuum method generates relatively expensive
contaminated sampling equipment that cannot easily be reused or decontaminated. In addition to the cost of
equipment (approximately $120-$140 per unit), the cost of proper disposal of post-sampling waste, including
the wet-vacuum units, should be taken into consideration when utilizing commercial vacuum cleaner
devices.
The optimized wet-vacuum sampling method as described here can be considered a good
prospective addition to the wipe-based surface sampling methods when responding to a large-area
chemical release or incident. Outcomes of the systematic testing of this surface sampling method provide
field responders with additional tools to characterize large-area contamination following an accidental
chemical release or terror incident.
viii

-------
Table of Contents
Disclaimer	ii
Acronyms and Abbreviations	iii
Acknowledgments	vi
Executive Summary	vii
Figures	xi
Tables	xiii
1.0	Introduction	16
1.1	Project Objectives	16
2.0 Experimental Approach	18
2.1	Test Facility	18
2.2	Experimental Design	18
3.0 Materials and Methods	19
3.1	Test Materials	19
3.2	Chemicals	20
3.3	Contamination of Coupons	21
3.4	Test Setup	23
3.5	Wet-Vacuum Test Equipment	24
3.6	Surface Wetting Agents	25
3.6.1 Wetting Agent Application Methods	26
3.7	Method Development Tests	26
3.7.1	Extraction of target chemicals from water and water-surfactant solutions	26
3.7.2	Wipe-based methods for surface sampling of target chemicals	28
3.8	Surface Sampling Tests	29
3.8.1	Phase I: Selection of operational parameters for wet-vacuum sampling	29
3.8.2	Phase II: Application of commercial wet-vacuum systems for surface sampling	30
3.8.3	Phase III: Multivariate characterization of wet-vacuum-based sampling methods	31
3.8.3.1 Operational characterization of commercial wet-vacuum cleaner-based sampling	31
3.8.3.1.1	Medium-area sampling	31
3.8.3.1.2	Large-area sampling	32
3.9	Supplementary Surface Concentration Verification Tests	33
3.9.1 Verification of surface concentrations by direct extraction	33
3.10	Supplementary Tests	33
3.10.1	Operational time estimates	33
3.10.2	Flow and temperature profiles of wet-vacuum system	34
3.10.2.1	Measurement of vacuum airflow	34
3.10.2.2	Temperature measurement of wet-vacuum unit	34
4.0 Sampling and Analysis	35
4.1 Surface Sampling Methods	35
4.1.1	Medium-size area wet-vacuum sampling	35
4.1.2	Large-area wet-vacuum sampling	35
4.1.3	Small-area wipe-based sampling	37
4.2. Extraction Methods	38
4.2.1 Extraction of surface wipes	38
ix

-------
4.2.2 Extraction of water and water-based wetting agents	38
4.3	Preparation of Samples for Analysis	40
4.4	Instrumental Analyses	41
4.5	Data Reduction Procedures	44
4.5.1	Chemical concentration in sampling liquid calculations	44
4.5.2	Sampling efficiency calculations	45
5.0 Results	46
5.1	Analysis of Sampling Liquids	46
5.2	Proof of Concept and Initial Optimization of Surface Sampling Method - Phase 1	48
5.3	Initial Evaluation of Commercial Wet-Vacuum Cleaners - Phase II	50
5.3.1	Comparison of commercial wet-vacuum cleaners	50
5.3.2	Optimization of the commercial wet-vacuum cleaner-based sampling	53
5.3.3	Wipe sampling optimization	53
5.4	Operational-scale Testing of Wet-Vacuum Sampling Efficiency - Phase III; Medium-size Area	54
5.4.1	Medium-size area sampling at default surface concentration	54
5.4.1.1 Residual chemical on wet-vacuumed surface and in wet-vacuuming unit	55
5.4.2	Medium-size area sampling at low (below default) surface concentration	56
5.4.3	Surface concentration verification by direct extraction	58
5.4.4	Comparison of wet-vacuum sampling to wipe-based sampling methods	59
5.5	Operational-scale Testing of Wet-Vacuum Sampling Efficiency - Phase III; Large-size (Composite)
Area	60
5.5.1 Operational considerations for large area composite sampling	60
5.6	Operational time comparisons	61
5.7	Supplementary Characterization Tests	62
5.7.1	Bissell wet-vacuum airflow	62
5.7.2	Wet-vacuum temperature profile	62
6.0 Quality Assurance and Quality Control	65
6.1	Test Equipment Calibration	65
6.2	Data Quality Results for Critical Measurements	65
7.0. Summary	67
8.0. References	69
Appendix A: Supporting Information	71
A-1 Wet-Vacuum Sampling Procedure	71
A-2 Parameters and Conditions for Instrumental analysis	73
A-3 Sample-Specific Test Results	76
x

-------
Figures
Figure 2-1. General experimental scheme of wet-vacuum sampling optimization and testing	18
Figure 3-1. Patterns for discrete droplet application of chemicals onto the medium-sized test surfaces: (a)
wet-vacuum sampling, (b) wipe-based sampling, and (c) bulk extraction of positive control sample	22
Figure 3-2. Microdroplet application of chemical (a) and examples of wet (b) and dried out (or post-
weathering) (c) chemical droplet pattern. Example shown is 2-CEPS on stainless steel	22
Figure 3-3. Clean test coupons prepared for testing: (a) stainless steel, (b) laminate flooring, (c) vinyl
flooring, and (d) plywood	23
Figure 3-4. Assembled wet-vacuum sampling apparatus consisting of connector, Nalgene bottle, tubing, and
custom-made nozzle	24
Figure 3-5. Commercial wet-vacuums used in this study: (a) Rug Doctor and (b) Bissell	25
Figure 3-6. Stainless-steel surface before (a) and after (b) wipe sampling	28
Figure 3-7. Surface spike controls ready for extraction	33
Figure 4-1. Experimental design and sample flow for SA sampling. TC - test coupon; PB - procedural blank;
CT - contact time; WT -wetting tank; DLT -dirty liquid tank	35
Figure 4-2. Experimental design and sample flow for MAC sampling. TC - test coupon; PB - procedural
blank; CT - contact time; WT -wetting tank; DLT -dirty liquid tank	36
Figure 4-3. Experimental design and sample flow for small area sampling. TC - test coupon; PB -
procedural blank; CT - contact time	37
Figure 4-4. Dawn Ultra® water SL immediately after conclusion of sampling (a), during settling phase in the
dirty liquid tank (b), and settled liquid (aliquot in beaker) ready for LLE (b)	39
Figure 4-5. Drying of the SL extract	40
Figure 5-1. Results from LLE of TEP and 2-CEPS in control samples of aqueous sample liquids; (a)
SuperSoap®-water LLE, (b) Dawn Ultra®-water LLE, (c) water LLE; HT, analytical HT defined as from control
sample preparation to extraction	47
Figure 5-2. Selection of operational parameters of laboratory-scale wet-vacuum sampling, (a) Selection of
wetting solvent, (b) selection of lapse time for IPA, and (c) results of wet-vacuum sampling of various test
surfaces using 50 mL IPA and LT = 10 s	49
Figure 5-3. Results of the initial evaluation of commercial vacuum cleaners for wet-vacuum sampling of
various test surfaces; sampling was performed using 50 mL of water, LT=10 s; material types: stainless
steel (a); laminate flooring (b); vinyl flooring (c)	51
Figure 5-4. Results of the initial evaluation of commercial vacuum cleaners for wet-vacuum sampling of
various test surfaces as a function of the solubility of the chemical in water	52
Figure 5-5. Comparison of sampling efficiency of SA wet-vacuum and wipe-based method for 2-CEPS (a)
and TEP (b); WS - wipe sampling, WV-wet-vacuum sampling	59
Figure 5-6. Comparison of total operational time for wet-vacuum and wipe-based sampling methods; WS -
wipe sampling; WV - wet-vacuum sampling	61
Figure 5-7. Airflow monitoring of Bissell unit simulated large-area sampling; flow rate is reported at US EPA
standard ambient sampling conditions of 101.325 kPa and 25 °C; slpm - standard liters per minute	62
Figure 5-8. Temperature monitoring of Bissell exhaust during simulated sampling. Sampling performed
under the chemical hood, (a) start of vacuum sampling (t = 0 min); (b) after 1 min of vacuum sampling (t = 1
xi

-------
min); (c) after 2 min of sampling (t = 2 min); (d) after 3 min of sampling (t = 3 min); (e) at the end of sampling
(t = 3 min 24 s); f, 15 min after sampling ended (t = 19 min 24 s)	63
Figure 5-9. Temperature in various zones of Bissell vacuum cleaner after 20 min sampling using water, (a)
Central system temperature near motor; (b) Liquid waste collection tank temperature	64
Figure A-1 Spray nozzle (A) and spray trigger (B) locations	72
xii

-------
Tables
Table 3-1. Specifications of Building Materials	19
Table 3-2. Physicochemical Properties and Functional Surrogate Classification of Target Chemicals	20
Table 3-3. Analytical Standards for Target Chemicals	21
Table 3-4. Chemical Reagents	21
Table 3-5. Surface Wetting Agents for Wet-Vacuum Sampling	26
Table 3-6. Experimental Parameters of Original and Modified Reference Methods for Wipe Sampling and
Extraction of Wipes	28
Table 3-7. Experimental Parameters of Phase I Testing	29
Table 3-8. Experimental Parameters of Phase II Commercial Wet-Vacuum Cleaner Testing - Initial
Evaluation of Rug Doctor and Bissell Vacuum Cleaners	30
Table 3-9. Experimental Parameters of Phase II Commercial Vacuum Cleaner Testing for Improved
Recovery of Organic Solvent-based Wetting Agent	30
Table 3-10. Experimental Parameters of Phase III Multivariate Characterization of Commercial Vacuum
Cleaner Testing - Single Medium-Area Sampling	32
Table 3-11. Experimental Parameters of Phase III Multivariate Characterization of Commercial Vacuum
Cleaner Testing - Large-Area Composite Sampling	32
Table 4-1. Chemical-Specific Experimental Conditions for Extractions of Aqueous SLs	39
Table 4-2. Instrumental Methods Used for Analysis of Target Analytes	41
Table 4-3. QC Checks for Instrumental Analyses Performed by External Laboratory	42
Table 4-4. Initial and Continuing Laboratory Proficiency Results	43
Table 5-1. Results for Wipe Sampling and Extraction Optimization Tests for 2-CEPS and TEP	53
Table 5-2. Test Results of Medium-Sized Area Wet-Vacuum Sampling of 2-CEPS from a Nonporous and
Permeable Material at the Default Surface Concentration	54
Table 5-3. Test Results of Medium-Sized Area Wet-Vacuum Sampling of TEP from Nonporous and a
Permeable Material	55
Table 5-4. Test Results of Medium-Sized Area Wet-Vacuum Sampling of 2-CEPS at Low Concentrations. 56
Table 5-5. Test Results of Medium-Sized Area Wet-Vacuum Sampling of TEP at Low Concentrations	57
Table 5-6. Test Results for Direct Extraction of 2-CEPS from Stainless Steel	58
Table 5-7. Test Results for Direct Extraction of TEP from Stainless steel	58
Table 5-8. Test Results of Large-Sized Area Wet-Vacuum Sampling of 2-CEPS and TEP from Stainless-
steel Material (n=3)	60
Table 6-1. Instrument Calibration Frequency	65
Table 6-2. Acceptance Criteria for Critical Measurements and Corresponding Test Results	66
Table A-1. GC/FID Analyses of Nitrobenzene with Modified NIOSH*2005 (EMSL Analytical, Inc.)	73
Table A-2. GC/FID Analyses of Phenol Modified NIOSH Method 2546 (EMSL Analytical, Inc.)	73
Table A-3. GC/FID Analyses of TEP Using Modified NIOSH 5034 (EMSL Analytical, Inc.)	74
Table A-4. GC/FID Analyses of TEP Modified NIOSH 5038 (EMSL Analytical, Inc.)	74
Table A-5. GC/MS Analyses of TEP and 2-CEPS Samples in IPA (EPA OSL)	75
Table A-6. GC/MS Analyses of TEP and 2-CEPS in Hexane (EPA OSL)	75
xiii

-------
Table A-7. Liquid-Liquid Extraction Efficiency of Selected Chemicals from Wetting Agents	76
Table A-8. Results of Direct Analysis of Phenol in Various Wetting Agents	77
Table A-9. Results of Direct Analysis of TEP in I PA	77
Table A-10. Results of Direct Analysis of2-CEPS in IPA	77
Table A-11. Phase I Operational Parameter Optimization - Selection of Wetting Agent for Sampling of
Phenol from Nonporous Reference Material	78
Table A-12. Phase I Operational Parameters Optimization - Selection of Wetting Agent for Sampling of
Nitrobenzene from Nonporous Reference Material	79
Table A-13. Phase I Operational Parameters Optimization - Selection of Wetting Agent for Sampling of TEP
from Nonporous Reference Material	80
Table A-14. Phase I Operational Parameters Optimization - Selection of the Wetting Agent Surface Contact
Time (Lapse Time) for IPA Sampling of Phenol from Nonporous Reference Material	81
Table A-15. Phase I Operational Parameters Optimization - Selection of the Wetting Agent Surface Contact
Time (Lapse Time) for IPA Sampling of Nitrobenzene from Nonporous Reference Material	82
Table A-16. Phase I Operational Parameters Optimization - Selection of the Wetting Agent Surface Contact
Time (Lapse Time) for IPA Sampling of TEP from Nonporous Reference Material	83
Table A-17. Phase I Operational Parameters Optimization - Evaluation of IPA Wetting Agent for Sampling
of Phenol from Semiporous and Porous Materials	84
Table A-18. Phase I Operational Parameters Optimization - Evaluation of IPA Wetting Agent for Sampling
of Nitrobenzene from Semiporous and Porous Materials	85
Table A-19. Phase I Operational Parameters Optimization - Evaluation of IPA Wetting Agent for Sampling
of TEP from Semiporous and Porous Materials	86
Table A-20. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Phenol from Stainless Steel
	87
Table A-21. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Phenol from Laminate
Flooring	88
Table A-22. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Phenol from Vinyl Flooring 89
Table A-23. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Nitrobenzene from Stainless
Steel	90
Table A-24. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Nitrobenzene from Laminate
Flooring	91
Table A-25. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Nitrobenzene from Vinyl
Flooring	92
Table A.26 Phase II Evaluation of Commercial Wet-Vacuums for Sampling of TEP from Stainless Steel
Coupon	93
Table A-27. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of TEP from Laminate Flooring
	94
Table A-28. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of TEP from Vinyl Flooring	95
Table A-29. Phase II Optimization of Wetting Solvent and Chemical Recovery Using Bissell Little Green
ProHeat Wet-Vacuum - Phenol on Stainless-steel Coupon	96
Table A-30. Phase II Optimization of Wetting Solvent and Chemical Recovery Using IPA and Bissell Little
Green ProHeat Wet-Vacuum - TEP on Stainless-steel Coupon	97
xiv

-------
Table A-31. Phase III Optimized IPA Wet-Vacuum Method Performance for 100% Surface Concentration
Reference Material Sampling Baseline of 2-CEPS, including Mass Balance Tests	98
Table A-32. Phase III Optimized TEP Wet-Vacuum Method Performance for 100% Surface Concentration
Reference Material Sampling Baseline of TEP, including Mass Balance Tests	99
xv

-------
1.0 Introduction
In the event of a large-area chemical incident, many environmental samples may be collected,
preserved, and analyzed to successfully identify and map the hazard and select the appropriate
decontamination strategy. Characterization and clearance sampling procedures and associated analyses
can present a high fiscal and logistical burden to responsible agencies. This is especially the case when
using a statistical sampling design as such design often requires the generation of large numbers of
samples to achieve reasonable confidence in, for example, hotspot detection or, post-decontamination, the
cleanliness of an area. Limited laboratory capacity and costly analysis, especially for chemical warfare
agents (CWAs), make statistically-based sampling strategies difficult to implement as part of the remediation
and clearance process.
Pre- and post-decontamination environmental sampling of chemical agents typically involves the
use of surface wipes for collection of a contaminant from a single discrete location (often an area less than
one square foot) or from multiple locations combined into one sample (composite sampling), followed by
extraction of the sampling medium and instrumental analysis of extracts [1-3], Operationally and logistically,
it is often challenging to sample, extract, and analyze large quantities of such environmental samples
quickly, especially when sample cleanup, fractionation, or solvent exchange is performed. Aside from
revising sampling strategies, for example, via direct in situ analysis, composite multilocation sampling may
be an alternative approach that can present many advantages such as reduced response sampling time,
fewer samples to process, and coverage of large sampling areas that would improve detection of
widespread contamination [1-4], A direct collection of the surface contamination into a solvent can reduce
sampling costs and efforts by eliminating the wipe extraction and preparation steps for such an extract prior
to analysis by readily available instrumental techniques (for example, gas chromatography/mass
spectrometry [GC/MS]). Direct collection of the surface contamination into a solvent also eliminates the use
of a wipe material, which sometimes requires a pre-cleaning cycle and additional Quality Assurance/Quality
Control steps.
In this study, a wet-vacuum cleaning method was evaluated for the surface sampling of nonporous
to porous building surfaces spiked with chemicals exhibiting differing degrees of water solubility. The wet-
vacuum sampling efficiency was studied using a prototype and two commercially available devices. A
relatively mild but moderately volatile organic solvent (isopropyl alcohol [IPA]), water, and water with
surfactant solutions were evaluated as the surface wetting and sampling agents. The optimized wet-vacuum
sampling method efficiency was compared to traditional wipe-based sampling methods and/or modifications
thereof. Research outcomes from this study benefit responders as it will allow them to consider high-
capacity composite sampling approaches for the sampling of chemical agents from environmental surfaces.
1.1 Project Objectives
The purpose of this project was to provide responding agencies with information on the
effectiveness of a wet-vacuum sampling approach for the collection of toxic industrial chemicals (TICs) and
CWA surrogates deposited on building materials with different porosities. This research is part of the U.S.
Environmental Protection Agency's (EPA's) Homeland Security Research Program (HSRP) efforts to
develop time- and cost-effective high-capacity sampling approaches to be used in large areas contaminated
with chemical agents. This new methodology uses commercial, off-the-shelf cleaning equipment (handheld
wet-dry vacuum units) for composite sample collection and composite sample analysis with various wetting
16

-------
solutions and solvents that are pertinent to field sampling in response to an accidental chemical release or
chemical terror incident.
The primary objectives of this research were to:
•	Develop a medium- to large-area surface sampling method that uses an easily accessible, off-
the-shelf device and surface sampling agents that can be used in the field.
•	Determine the feasibility and effectiveness of the proposed method by measuring medium- and
large-area surface sampling efficiencies through measurement of the mass of chemical
deposited on the surface as a function of chemical surface loading, wetting agent, and material
type.
•	Compare wet-vacuum surface sampling efficiencies to the efficiency of conventional (wipe-
based) sampling methods. These comparisons were performed in conjunction with initial
estimates of differences in operational time and sample turnaround time between these two
types of sampling approaches.
The secondary objectives were to:
•	Provide information on residual chemical concentrations that remain inside the wet-vacuum
system after sampling.
•	Provide information on residual chemical concentrations that remain on surfaces after wet-
vacuum sampling.
•	Provide initial information on the general wet-vacuum operational parameters that can be
relevant to field applications (e.g., temperature profile, flow, potential for reuse).
Outcomes of the systematic testing of this surface sampling methodology will provide field
responders with additional tools to characterize large-area contamination following an accidental chemical
release or terror incident. This information provides the scientific basis for a potentially significant reduction
in the time and cost of post-decontamination sampling events of selected chemical agents. This project
supports the strategic goals of improving the capability to respond to chemical incidents that affect buildings
and the outdoor environments as part of EPA's Homeland Security Research Program.
17

-------
2.0	Experimental Approach
2.1	Test Facility
The experimental work was performed at the EPA facility in Research Triangle Park (RTP), NC.
Instrumental analyses of target chemicals in extracts and control samples were performed at an external
accredited chemical analysis laboratory (EMSL Analytical, Inc., Cinnaminson, NJ) and at the in-house EPA
Organic Support Laboratory (OSL), in RTP, NC.
2.2	Experimental Design
This study was performed in three consecutive phases. The first proof-of-concept phase (Phase I)
consisted of the bench-scale determination of basic test parameters such as the selection of type and
volume of the wetting solvent and evaluation of material-specific surface lapse times. This surface lapse
time is defined here as the time between application of the wetting solvent and the wet-vacuum of this
agent. Phase II involved the wet-vacuum method optimization for improved recovery of the selected wetting
solvent and target chemicals using a single discrete area sampling. Phase III involved the operational-scale
deployment of the optimized method for sampling of various surface concentrations of TICs using organic
solvent and water-based wetting agents for medium-sized (929 cm2) and a composite sampled area of five
medium-sized areas (5 x 929 cm2). The general experimental approaches for each phase of this work are
shown in Figure 2-1. Details of each experimental phase are described in Sections 3.8.1 through 3.8.3.
Phase I
¦	Selection of wetting solvent (isopropyl alcohol, water, and water-
surfactant)
¦	Selection of wetting solvent volume
¦	Selection of wetting solvent surface lapse time
Phase II
¦ Initial testing of commercial wet-vacuum systems for single discrete
location (1 x 929 cm2) sampling
» Optimization of IRA wetting solvent recovery for wet-vacuum
sampling (addition of pre-rinse and post-rinse steps)
Phase III
¦	Evaluation of various wetting agents for wet-vacuum sampling of
building materials for single discrete location (1» 929 cm2) and multi-
location (5 x 929 cm2) composite
¦	Chemical mass balance testing
¦	Comparison of wet-vacuum sampling to wipe-based reference method
(sampling efficiency and operational time)
Figure 2-1. General experimental scheme of wet-vacuum sampling optimization and testing
*0# I

18

-------
3.0 Materials and Methods
An ideal surface sampling methodology should generate accurate and reproducible data when used
on environmental surfaces for which it was designed. The analytical data must meet the quantitative and
qualitative detection criteria that are relevant to the applied method. Forthis study, the sampling method
was designed for the defendable determination of a surface chemical contamination level. Wet-vacuum
sampling approaches may consider horizontal surfaces such as flooring materials or countertops or vertical
surfaces such as walls. In this project, the focus was on horizontal surfaces with flooring materials as the
most prevailing material type that would require sampling to characterize the (residual) level of chemical
contamination following a contamination event.
3.1 Test Materials
Several types of building materials with different porosities and permeabilities were selected for
evaluation of sampling procedures (Table 3-1). Multipurpose stainless steel was a surrogate for smooth
nonpermeable building surfaces (e.g., sinks, countertops). Due to its inertness, low porosity, and excellent
corrosion resistance [5], stainless steel was also used as a good reference material for optimization of
sampling approaches. Three types of flooring (vinyl, laminate and plywood) were selected as representative
of semipermeable (laminate) and permeable (vinyl and plywood) building materials. The building material
specifications are given in Table 3-1. A low efficiency of wetting solvent collection was observed when
sampling target chemicals on plywood during the Phase I method development (results are described in
Section 5.2). Hence, this material was not used in the subsequent wet-vacuum sampling optimization and
testing (Phases II and III).
Table 3-1. Specifications of Building Materials
Material
Description
Manufacturer/
Supplier Name/Location/Country
Coupon Size,
LxW (cm)*
Material Preparation
Stainless
steel
Multipurpose stainless steel
(1.2x1.2m),type304,#2B
mill (unpolished), 0.091 cm
thick
McMaster-Carr Douglasville, GA,
USA
35.56 x 35.56
Cut into coupons and remove any
lubricant/grease from shearing with
acetone. Wipe dry.
Immediately before use, remove
particles and dust by wiping clean with
acetone and then water. Wipe dry.
Vinyl
flooring
2.4x3.6mCasa Grande
beige precut sheet vinyl,
residential grade, low gloss,
stain resistant, scratch
resistant, 0.050 cm
Tarkett, Inc.
Whitehall, PA, USA
35.56 x 35.56
Cut into coupons. Remove particles by
wiping clean with water and wipe dry.
Laminate
flooring
Project Source 20.5 x 120
cm natural oak smooth
laminate wood planks
Clarion Laminates LLC
Shippenville, PA, USA
35.56 x 35.56
Attach to the plywood base and cut
into coupons. Remove particles by
wiping clean with water and wipe dry.
Plywood
1.2 x2.4 m Plytanium®
untreated pine plywood,
1.27 cm thick
Georgia-Pacific Building Products
Atlanta, GA, USA
35.56 x 35.56
Cut into coupons. Remove particles by
wiping clean with water and wipe dry.
'Actual effective test area was center 30.48 cm x 30.48 cm (929 cm2) or 12 inches (") x 12".
Stainless-steel and vinyl coupons were cut to the correct length and width from larger sheets using
heavy-duty power hydraulic shears. Plywood panels were precut to desired dimensions using a table saw.
19

-------
All coupons were cleaned prior to testing using the procedures described in Table 3-1. Test boxes
(dimensions: 45.72 cm * 45.72 cm * 15.88 cm; Stor-N-Slide square box with lid, product no. 491530; IRIS
USA, Inc., Pleasant Prairie, Wl, USA) that held one individual coupon were cleaned using laboratory-grade
detergent solution, wiped with acetone and water, and wiped dry.
3.2 Chemicals
The target chemicals used in this study were selected using two criteria: (1) the chemicals selected
must be representative of a wide range of water solubility, from very slightly soluble (range: grams per liter
or less) to very soluble (range: hundreds of grams per liter to completely miscible) and (2) the chemicals
selected must be representative of various classes of TICs, including CWA surrogates.
The relevant physical and chemical properties and functional surrogate classification of target
chemicals are listed in Table 3-2. Information on the neat chemical standard sourcing is summarized in
Table 3-3. Information on internal standard and surrogate compound analytical standards used in this study
is given in Section 4.4. Other chemical reagents used as wetting solvents, drying agents or extraction
solvents are listed in Table 3-4.
Table 3-2. Physicochemical Properties and Functional Surrogate Classification of Target Chemicals
Target Chemical
2-Chloroethyl phenyl
sulfide
(2-CEPS)
Nitrobenzene
Phenol
Triethyl phosphate
(TEP)
CAS registry number
5535-49-9
98-95-3
108-95-2
78-40-0
Other common names
and/or abbreviations
2-CEPS
Nitrobenzol
Hydroxybenzene
TEP

Physicochemical Properties*
Molecular weight
172.7
123.1
94.1
182.2
Chemical formula
CsHgCIS
C6H5NO2
CeHeO
C6H15O4P
Density (g/cm3)
1.174
1.20
1.06
1.072
Vapor pressure (Pa)t
2.53
40
46
52
Solubility in water
(grams (g)/liter (L))t
0.084
2.1t
83
500

Water Solubility and Functional Surrogate Class
Functional surrogate
class
Surrogate of CWA
(sulfur mustard [HD])
[6,7]
Toxic industrial chemical
Toxic industrial chemical
Toxic industrial chemical
*Data from httDs://pubchem.ncbi.nlm.nih.aov: httos://www.siamaaldrich.com
TAt 25 °C
20

-------
Table 3-3. Analytical Standards for Target Chemicals
Target Chemical
2-Chloroethyl phenyl
sulfide
Nitrobenzene
Phenol
Triethyl phosphate
Manufacturer
Sigma Aldrich
St. Louis, MO, USA
Sigma-Aldrich
St. Louis, MO, USA
Sigma-Aldrich
St. Louis, MO, USA
Sigma-Aldrich
St. Louis, MO, USA
Product No.
417602-5ML
48547
P9346-100mL
538728-1L
Purity, %
98
100
>89.0
>99.8
Table 3-4. Chemical Reagents
Chemical Reagent
Purity/Grade
Product No.
Manufacturer
Sodium sulfate cartridge
n/a
12131033
Agilent Technologies, Santa Clara, CA, USA
Sodium chloride
99.7%
S271-500
Fisher Scientific, Fair Lawn, NJ USA
Sodium chloride
99.0%
VW6430-1
VWR Scientific Products, West Chester, PA, USA
1 molar (M) Sodium phosphate buffer pH 7.0
n/a
P2070
Teknova, Hollister, CA, USA
Hexane*
ACS/HPLCt
H303-4
Fisher Scientific, Fair Lawn, NJ, USA
Dichloromethane
ACS/HPLC
300-4
Honeywell International Inc., Muskegon, Ml, USA
IPA
ACS Plus
A4164
Fisher Scientific, Fair Lawn, NJ, USA
'Mixture, as purchased, ofn-hexane (45-60%), hexane (-isomers) (15-40%), and cyclohexane (3%).
tHPLC - High performance liquid chromatography.
The target surface chemical concentrations varied in this study. Most wet-vacuum and wipe-based
sampling tests were performed at a chemical surface concentration challenge level of approximately 300
milligrams (mg) per square meter (m2), which is equivalent to 26-29 mg per medium-sized coupon. For
these tests, the central test area of each medium-sized coupon (30.48 cm x 30.48 cm, or 929 cm2) was
contaminated with twelve uniformly distributed (2 |jl_) droplets of neat chemical. For low surface
concentration and wipe-based (reference method) sampling tests, the coupons were contaminated at lower
levels down to 1% of the default contamination level (26-29 mg/coupon of 929 cm2) using spiking solutions
of the target chemical prepared by dissolving the neat chemical in an appropriate organic solvent. Each
spiking solution was mixed using a vortex mixer and then via sonication for approximately 30 seconds. The
accuracy and precision of the spiked test area was tested along with each experimental batch by analysis of
control spike (CS) samples where the same amount of chemical that was applied to the surface was spiked
directly into the extraction solvent.
3.3 Contamination of Coupons
Chemical solutions were applied to test coupons (TCs) using a discrete droplet (micro)application
method. Prior to chemical application, each 35.56 cm x 35.56 cm TC was placed in a precleaned test box.
Chemical solutions were then applied to the coupons under room temperature conditions within a chemical
safety hood using a separate tip-programmable, electronic, repeatable pipette (Eppendorf Repeater Plus
Single Channel Repeater Pipette, Eppendorf AG, Hamburg, Germany; product no. 22260201) and a
precleaned stainless-steel spiking template placed over the coupon surface. For wet-vacuum testing, 12
droplets (2 microliter (|jL) volume each) were applied within the 30.48 cm x 30.48 cm test area following the
pattern shown in Figure 3-1 a. For wipe-based sampling method evaluation, nine droplets of spiking solution
were applied onto the 10 cm x 10 cm central part of the medium-sized coupon (Figure 3-1 b). For positive
21

-------
control testing, a composite sample comprising 12 sub-coupons of the same material was contaminated
with a discrete droplet spike delivered onto each small coupon (Figure 3-1 c).
30.48 cm
10 cm
QBBBI
~ ~~~
30.48 cm
Figure 3-1, Patterns for discrete droplet application of chemicals onto the medium-sized test surfaces: (a) wet-
vacuum sampling, (b) wipe-based sampling, and (c) bulk extraction of positive control sample
After chemical application, the boxes were closed to allow a 30-minute simulated weathering or
contact time; weathering was performed in the chemical hood under normal ambient laboratory conditions.
Coupons were stored in the closed test boxes during the simulated weathering pre-decontamination phase
to reduce possible evaporation of chemicals due to the high air-flow conditions inside the chemical safety
hood.
Figure 3-2 shows (a) the spiking procedure, (b) examples of the chemical droplet pattern
immediately after spiking, and (c) dried out (or post-weathering; contact time = 30 min) chemical priorto
sampling. The example shown is 2-CEPS on stainless steel.
a
\
K
Figure 3-2. Microdroplet application of chemical (a) and examples of wet (b) and dried out (or post-weathering)
(c) chemical droplet pattern. Example shown is 2-CEPS on stainless steel
The accuracy and precision of spiking solution preparation was assessed for each experimental
batch by analysis of control spike samples.
22

-------
3.4 Test Setup
All proof-of-concept arid methodology optimization tests (Phases I and II, Sections 3.8.1 and 3.8.2)
and simulated field sampling operational-scale experiments (Phase III, Section 3,8.3) were performed within
a chemical safety hood. Each contaminated coupon was placed and handled in a precleaned test box.
Figure 3-3 shows examples of clean stainless-steel, laminate flooring, vinyl flooring, and plywood coupons
readied for testing.
a.
Figure 3-3. Clean test coupons prepared for testing: (a) stainless steel, (b) laminate flooring, (c) vinyl flooring,
and (d) plywood
Experimental details of the sampling approaches that were tested are given in Section 4.1.
23

-------
3.5 Wet-Vacuum Test Equipment
Wet-vacuum sampling method laboratory-scale evaluations were performed using a custom-made
wet-vacuum apparatus made from off-the-shelf components and two different commercial wet-dry vacuum
cleaning units (hereafter referred to as wet-vacuums).
The custom-made sampling apparatus consisted of a Nalgene™ 500-milliliter (ml.) wide-mouth
polypropylene bottle (Thermo Scientific, Waltham, MA, USA) equipped with a wet-vacuum adapter.
Samples were collected using a custom-made acrylic flat nozzle. The nozzle was connected to the wet-
vacuum bottle via 1 meter (m) of latex tubing with a cord-grip fitting. Figure 3-4 shows the wet-vacuum
apparatus readied for testing. The slit nozzle opening was approximately 1 mm (height) x 42 mm (width). A
self-contained service vacuum (OmegaPlus Vacuum Cleaner [Atrix International Inc., Burnsville, MN USA])
pump was used to pull the liquid sample into the bottle. A high-efficiency particulate air filter (Atrix
International Inc., Burnsville, MN USA) was included downstream from the bottle to protect the pump.
Figure 3-4. Assembled wet-vacuum sampling apparatus consisting of connector, Nalgene bottle, tubing, and
custom-made nozzle
A new Nalgene bottle with new tubing was used to collect each sample to avoid cross-
contamination. Grip fittings, wet-vacuum adapters, and nozzles were cleaned by soaking in an activated
hydrogen peroxide-based decontamination solution (EasyDecon® DF200, Intelagard, Lafayette, CO, USA).
After overnight soaking, all parts were triple-rinsed with deionized (Dl) water. Following the rinse step, the
reusable components were air dried.
After the method optimization using the wet-vacuum sampling prototype device (Phase I, Section
3.8.1) was completed, two types of commercial hand-held wet-vacuums were evaluated for sampling
chemicals from different types of building surfaces. The selected units are readily available at hardware
stores. Other models exist that have similar properties. The critical parameters were the need for a single
unit that can dispense the wetting solvent and vacuum this solvent after contact with a surface with the
ability to define the time between solvent application and vacuuming. The requirement for the minimal
solvent / wetting agent collection volume was one liter. The following two models (price $120-$140 per unit)
met these criteria:
• Rug Doctor Portable Spot Cleaner (model no. 93300, Rug Doctor, Inc., Piano, TX) (Figure 3-
5a) is a high-suction power portable carpet cleaner equipped with a handheld motorized brush
that moves 1200 times per minute and is specifically designed to deep-clean carpets and break
down stains without the use of heated water or steam. The maximum capacity of both the clean
24

-------
(wetting) liquid and dirty (cleaning waste) liquid tanks is 1890 mL. The unit weighs
approximately 9.1 kg; dimensions are 45 (H) cm x45 (W) cm x 32 (D) cm.
• Bissell Little Green ProHeat Compact Multipurpose Carpet Cleaner (model no. 14259,
Bissell Corp., Grand Rapids, Ml) (Figure 3-5b) is a high-suction power portable carpet cleaner
holding separate tanks for wetting liquid and cleaning waste, with an approximate capacity of
1420 mL each. The unit is equipped with a built-in (optional-use) water heater for enhanced
cleaning of tough surface stains. The cleaning liquid heater option was not used during this
study. The unit weighs approximately 6.4 kg; dimensions are 32 cm height (H) cm x 44 cm
width (W) x 21 cm depth (D).
Bissell"
LiTTLEsreen
111H111 jj Mil ii iliif	
Figure 3-5. Commercial wet-vacuums used in this study: (a) Rug Doctor and (b) Bissell.
3.6 Surface Wetting Agents
The surface wetting liquid was used to wet the coupon surface and, after the appropriate elapsed
time, was removed from the coupon surface using the vacuum technique. Surfaces were wetted via a spray
nozzle (commercial units) or separate spray bottle (for prototype unit). The wetting liquids considered in this
study represented two general classes: (1) a mild organic solvent, and (2) water-based agents. The organic
solvent used for wet-vacuum sampling was IRA because it is a relatively nonvolatile solvent, typically not
destructive to common building surfaces, and readily available in large quantities at iov/ cost. Other wetting
agents were selected based on previous research efforts [8] or manufacturers' information. The detailed
information regarding composition and preparation of wetting agents is presented in Table 3-5.
25

-------
Table 3-5. Surface Wetting Agents for Wet-Vacuum Sampling
Cleaning Agent
Composition
Manufacturer
Preparation
Utilized in
Phase No.
Aqueous Wetting Agents
Water
H20
Dracor Water Systems,
Durham, NC, USA
Dl water; no preparation
I, II, III
Water-Tween®20
Tween®20 in water
Sigma Aldrich
St. Louis, MO, USA
1 -part Tween®20 to 50 parts
Dl water
I
Water-Dawn Ultra®
Dawn Ultra® dishwashing liquid
in water
P&G
Cincinnati, OH, USA
1-part Dawn Ultra®
dishwashing liquid to 50
parts Dl watert
III
Water-SSDX-12
SuperSoap®
SSDX-12 SuperSoap® in water
Aerosafe
Norcross, GA, USA
1-part SSDX-12
SuperSoap® to 128 parts Dl
water
III
Organic Solvent Wetting Agents
I PA
CH3CHOHCH3, certified ACS
Fisher Chemical
Waltham, MA, USA
No preparation
I, II, III
All aqueous solutions were mixed immediately prior to testing using the preparation procedures
listed in Table 3-5. Wetting agents were prepared in quantities sufficient to perform sampling of all analytical
batches planned for each testing day.
3.6.1 Wetting Agent Application Methods
During Phase I, wetting agents were applied by using a 240-mL spray bottle made of high-density
polyethylene equipped with an adjustable polypropylene sprayer (McMaster-Carr, Elmhurst, IL, USA;
product # 9864T51). Spray application was uniform across the surface in a left to right sweeping motion,
with individual spray patterns overlapping by at least 50%.
Both commercial wet-vacuum units used during Phase II and III are equipped with a built-in spray
tool that can be triggered to apply the solution from the wetting tank to the surface. No modifications were
made to the spray pattern or spray flow rate of each unit. During method development tests only, one wet-
vacuum unit was designated as the unit that sprayed the liquid while a second wet-vacuum unit was used to
collect the sample. For each wet-vacuum sampling test, the initial and end mass of the wetting liquid holding
tank was recorded to check for the accuracy and precision of wetting solvent application (nominal 50 mL).
3.7 Method Development Tests
3.7.1 Extraction of target chemicals from water and water-surfactant solutions
Methods for the liquid-liquid exchange (LLE) of target chemicals sampled using aqueous wetting
agents (Dl water and water-surfactant solutions) were optimized prior to testing. Liquid-based testing was
used for each target chemical wetting agent combination. For this series of tests, one test was performed
with a 1 -hour analytical holding time (HT) and one test was performed with a 24-hour HT, where HT is the
26

-------
period the chemical was in contact with the aqueous (surfactant-containing) solution prior to extraction. The
HTs were designed to determine the ability to recover target chemicals from the surfactant-containing
solutions and to determine the associated stability of target chemicals in surfactant-water solutions. The
spiked sample concentration matched that of samples that originated from vacuum sampling (i.e., 24 |jl_ of
neat chemical was spiked into 50 mL of wetting solvent solution [100% target concentration]). Tests were
performed in triplicate (n = 3) for each chemical-wetting solvent combination. One procedural blank of
wetting solvent did not receive the chemical contamination but did undergo the extraction procedure to
monitor for possible cross-contamination or quantitative interferences that might result from the LLE.
The extraction optimization test parameters are listed in Table 3-6. Optimized LLE procedures are
detailed in Section 4.2. Recoveries of target chemicals for each LLE method and HT tested are described in
Section 5.1. Additives shown in Table 3-6 were used to salt out the surfactant from the aqueous solutions.
Table 3-6. Experimental Parameters for Liquid-Liquid Extraction Optimization Tests
Wetting Agent
Solution Chemical
Combination
Chemical Cone.
Target Tested
Extraction
Solvent Type
Wetting Solvent to
Extraction Solvent
Ratio (v:v)
Additives
Other Steps
TEP
Water
0-100%
DCM
1:1
NaCI, NaH2P04/Na2HP04
buffer (pH 7.0)
Drying with
Na2S04
Water-Tween®
0-100%
DCM
1:1
NaCI, NaH2P04/Na2HP04
buffer (pH 7.0)
Drying with
Na2S04
Water-Dawn Ultra®
0-100%
DCM
1:1
NaCI, NaH2P04/Na2HP04
buffer (pH 7.0)
Drying with
Na2S04
Water-SuperSoap®
0-100%
DCM
1:1
NaCI, NaH2P04/Na2HP04
buffer (pH 7.0)
Drying with
Na2S04
2-CEPS
Water
0-100%
Hexane
1:1
None
Drying with
Na2S04
Water-Dawn Ultra®
0-100%
Hexane
1:1
NaCI
Drying with
Na2S04
Water-SuperSoap®
0-100%
DCM
1:1
NaCI
Drying with
Na2S04
Nitrobenzene
Water
0-100%
Hexane
1:1
None
Drying with
Na2S04
Water-Tween®
0-100%
Hexane
1:1
None
Drying with
Na2S04
27

-------
3.7.2 Wipe-based methods for surface sampling of target chemicals
The effectiveness of the optimized wet-vacuum method for medium- and large-area sampling,
evaluated in experimental Phase III (Section 3.8.3), was compared to the chemical recovery achieved using
the wipe-based reference method. In this study, modified SW-846 Method 3572 [9] for extraction of CWAs
from wipe samples using micro extraction was considered for the surface sampling of CWA surrogates (TEP
and 2-CEPS). During two rounds of method optimization testing (data not shown), the standard wipe
sampling method was modified to include selected sampling strategies (extraction solvent and extraction
solvent volume) outlined in the EPA technical report, Evaluation of Chemical Warfare Agent Wipe Sampling
Collection Efficiencies on Porous, Permeable, or Uneven Surfaces [3], The modifications are listed in Table
3-6.
Table 3-6. Experimental Parameters of Original and Modified Reference Methods for Wipe Sampling
and Extraction of Wipes
Parameter/Method	EPA Method 3572 [3]	Modified Method 3572
Sampling area
10cmx10cm
10cmx10cm
Sampling wipe
5.08 cm. x 5.08 cm cotton gauze wipe
5.08 cm. x 5.08 cotton gauze wipe
Number of wipes per sampling area
1 per 100 cm2
2 per 100 cm2
Sampling (wetting) solvent
I PA
IPA
Sampling (wetting) solvent volume
1 mL per wipe
1.5 mL per wipe
Extraction solvent
10% IPA in DCM
DCM
Extraction solvent volume
4mL
2x15mL
For surface sampling efficiency tests, the reference surface (stainless steel) was contaminated with
solutions of target chemicals using the procedure described in Section 3-3 and placed in the same type of
precleaned test box that was used during decontamination testing. After the prescribed contact time of the
chemical with the surface (30 min), wipe samples were collected and extracted using the procedures
described in Section 4.1. Additional control samples included a wipe spike, a control spike sample, and
procedural blank. Figure 3-6 shows a 2-CEPS-contaminated stainless-steel coupon before and after wipe
sampling. The template was used to establish the 10 cm x 10 cm sampling area,

Figure 3-6. Stainless-steel surface before (a) and after (b) wipe sampling.
The results for the final round of the wipe-sampling method optimization are given in Table 5-1.
28

-------
3.8 Surface Sampling Tests
3.8.1 Phase I: Selection of operational parameters for wet-vacuum sampling
Since the target chemicals were characterized by various degrees of solubility in water, the
selection of the most appropriate, and ideally universal to all chemical-surface combinations wetting
condition was considered first. Phase I, the proof of concept phase, focused on the fundamentals of a new
wet-vacuum-based methodology being developed. In this phase of testing, a small custom-made apparatus
(Section 3-5) was used to allow for cost-effective and environmentally responsible testing (the cost of
equipment and amount of generated waste were minimal compared to Phase II and III operational-scale
testing in which a commercial vacuum was used).
Phase I laboratory-scale testing included the following steps:
1.	Selection of the wetting agent suitable for sampling chemicals with various degrees of solubility
in water.
2.	Selection of the optimal temporal duration between wetting agent application and start of wet-
vacuum sampling (surface residence time, or lapse time), as a function of surface type.
3.	Selection of the optimal wetting agent volume, defined as a minimal volume that offered good
recovery of all target chemicals. Chemical recovery testing was limited to surfaces with low
porosity.
4.	Evaluation of an optimized method for all chemical-material combinations.
The selection of optimal experimental parameters was based on best sampling results (highest
sampling efficiency) achieved in consecutive steps of testing. The test matrix for Phase I testing is provided
in Table 3-7. The results are described in Section 5.2.
Table 3-7. Experimental Parameters of Phase I Testing
Test
Test
Target
Wetting
Wetting Agent
Lapse
Test
Area Sampled*
Parameter
Material*
Chemical
Agent*
Volume
(mL)
Time
(s)
Matrix?
(cm2)
Selection of wetting
agent
Stainless steel
Nitrobenzene
Phenol
TEP
Water
Tween®-Water
IPA
100
100
3xTC
1 x PB
1 xCS
929
Selection of lapse
time
Stainless steel
Vinyl flooring
Plywood
Nitrobenzene
Phenol
TEP
Water
Tween®-Water
IPA
100
1
10
100
3xTC
1 x PB
1 xCS
929
Selection of wetting
agent volume
Plywood
Phenol
IPA
10
50
100
10
3xTC
1 x PB
1 xCS
929
Evaluation of
optimized method
on various surfaces
Stainless steel
Vinyl flooring
Plywood
Nitrobenzene
Phenol
TEP
IPA
50
10
3xTC
1 x PB
1 xCS
929
* Detailed information on test materials and wetting agents in Section 3.0. t Per each test condition, t Per replicate test sample; TC - Test coupon; PB - Procedural
blank; CS - Control spike sample.
29

-------
3.8.2 Phase II: Application of commercial wet-vacuum systems for surface sampling
The wet-vacuum sampling parameters developed during Phase I testing were applied to Phase II
commercial wet-vacuum cleaner-based operational-scale testing. Phase II laboratory-scale testing included
the following steps:
1.	Initial evaluation of the selected commercial vacuum cleaner suitable for sampling chemicals
with various degrees of solubility in water.
2.	Selection and optimization of the commercial vacuum cleaner for operational-scale wet-vacuum
sampling efficiency.
The overall effectiveness of wet-vacuum sampling was determined for two commercial devices and
three wetting agents as a function of the material type (Table 3-8).
Table 3-8. Experimental Parameters of Phase II Commercial Wet-Vacuum Cleaner Testing - Initial
Evaluation of Rug Doctor and Bissell Vacuum Cleaners
Wet
Vacuum Cleaner*
Test
Materials*
Target
Chemical
Wetting
Agents*
Wetting Agent
Volume (mL)
Lapse Time (s)
Test Matrix?
Area
Sampled*
(cm2)
Rug Doctor
Stainless steel
Laminate flooring
Vinyl flooring
Nitrobenzene
Phenol
TEP
Water
IPA**
50
10
5xTC
1 x PB
1 xCS
929
Bissell
Water
IPA**
5xTC
1 x PB
1 xCS
929
* Detailed information on test materials and wetting agents in Section 3.0. t Per each test condition, t Per replicate test sample; TC - Test coupon; PB - Procedural
blank; CS - Control spike sample; "Tested in scoping tests only
Results from this initial evaluation of commercial vacuum cleaners (Section 5.3.1) showed that the
Bissell unit offered better sampling efficiency for all chemical test material-wetting agent combinations
tested. Due to large evaporative losses of IPA observed during the initial scoping tests, the organic solvent-
based sampling procedure was further optimized to include a vacuum conditioning step (a 50-mL pre-rinse
of the same wetting solvent). In addition, different volumes for the introduced post-sampling rinse of the
vacuum cleaner (post-rinse), using clean IPA aspired directly from a beaker, were tested for improved
recovery of target chemicals (Table 3-9). The results of Phase II testing are summarized in Section 5.3.2.
Table 3-9. Experimental Parameters of Phase II Commercial Vacuum Cleaner Testing for Improved
Recovery of Organic Solvent-based Wetting Agent
Wet
Vacuum
Cleaner
Test
Target
Wetting
Wetting Agent
Volume
Lapse
Time
Test
Area
Sampled*
(cm2)
Material
Chemical
Agent*
Pre-rinse
(mL)
Surface Wetting
(mL)
Post-rinse
(mL)
(s)
Matrix?
Bissell
Stainless
steel
Phenol
TEP
IPA
50
50
no post-rinse
100
150
10
5xTC
1 x PB
1 xCS
929
"Detailed i
sample.
nformation on test materials, wetting agents, and commercial wet-vacuum in Section 3.0.* Per each test condition. * Per replicate
30

-------
3.8.3 Phase 111: Multivariate characterization of wet-vacuum-based sampling methods
In Phase III, the Bissell vacuum system was evaluated for collection efficiency of two selected
chemicals (TEP and 2-CEPS) using the optimized method tested in Phase II that included a pre-rinse step
(vacuum system conditioning step) and a post-collection rinse step. Phase III operational testing was
conducted as follows:
1.	Evaluation of one commercial device for wet-vacuum sampling of a medium-sized area, as a
function of chemical surface loading, including a comparison to the wipe-based surface
sampling method.
2.	Evaluation of the commercial device for high-capacity composite-area wet-vacuum sampling.
The various test combinations are shown in Tables 3-10 and 3-11. Each Phase III test was
performed in a sample configuration similar to the sample configuration in the Phase I and II experiments,
three wet-vacuum sampling method replicates (test coupons) accompanied by a procedural blank and
control spike. The same solvent (IPA) sampling and analysis sampling method was further tested for a
discrete single medium-sized area (SA, 929 cm2) and a larger-sized area via multiarea composite (MAC)
sampling of an area approximately 4,645 cm2. Because there were concerns about the feasibility of the IPA-
based method for sampling of semiporous materials (vinyl flooring) and health and safety concerns
(flammability) about the use of IPA in an operational setting, the SA methodology was also tested in the
aqueous wetting solvent configuration. Application of the optimized methodology for low surface
concentrations (1 to 20% of default-loading) sampling of nonporous material was performed using an SA
and IPA as well as a selected water-based sampling agent. Other supplementary tests for the wet-vacuum
system characterization are described in Section 3.9. The results from Phase III sampling tests are given in
Section 5.4.
3.8.3.1 Operational assessment of commercial wet-vacuum cleaner-based sampling
3.8.3.1.1 Medium-area sampling
The medium-area sampling was designed as a systematic study of the effectiveness of the wet-
vacuum method for sampling varying surface loadings of selected CWAs and TICs. The medium-area
sampling was completed for TEP and 2-CEPS, which have noticeably different chemical solubility in water
(see Table 3-4) on nonporous (stainless-steel) and semiporous (vinyl flooring) surfaces. The porous
substrate, plywood, was not tested because the wet-vacuum method has very limited applicability for highly
porous and permeable materials (as demonstrated during the proof of concept Phase I testing). Each
coupon was sampled using a multi-pass vacuum procedure that was used for the medium-sized coupon
sampling method (Appendix A, Section A-1). Both classes of wetting agents (organic solvent- and water-
based, Table 3-10) were tested using this method. Testing at lower surface concentrations were limited to
IPA wetting agent and only the best performing water-based wetting agent (Dawn Ultra®). Results from the
optimized wet-vacuum sampling method were compared to results from the modified standard wipe-based
CWA sampling method. The results from Phase III medium-area testing are provided in Sections 5.4.1 and
5.4.2, and the comparison to wipe method results are in Section 5.4.4.
31

-------
Table 3-10. Experimental Parameters of Phase III Multivariate Characterization of Commercial
Vacuum Cleaner Testing - Single Medium-Area Sampling
Wet
Vacuum
Cleaner
Test
Material
Target
Chemical
Surface
Cone.*
Wetting
Agents*
V
Pre-
rinse
(mL)
letting Ager
Volume
Surface
Wetting
(mL)
it
Post-
rinse
(mL)
Lapse
Time
(s)
Test
Matrix*
Area
Sampled**
(cm2)
Bissell
Stainless
steel
TEP
2-CEPS
1%
2%
10%
20%
100%
Isopropyl
Alcohol
50
50
100
10
3xTC
1 xPB
1 xCS
929
Bissell
Stainless
steel
TEP
2-CEPS
100%
Water
Dawn Ultra®-water
SuperSoap®-water
no
pre-
rinse
50
100
10
3xTC
1 xPB
1 xCS
929
Bissell
Vinyl
flooring
TEP
2-CEPS
100%
Water
Dawn Ultra®-water
SuperSoap®-water
no
pre-
rinse
50
100
10
3xTC
1 xPB
1 xCS
929
Bissell
Stainless
steel
TEP
2-CEPS
2%
10%
20%
Dawn Ultra®-water
no
pre-
rinse
50
100
10
3xTC
1 xPB
1 xCS
929
'With respect to highest loading tested. tDetaiied information on test materials, wetting agents, and commercial wet-vacuum is found in Section 3.0. *Per each test
condition. "Per replicate sample.
3.8.3.1.2 Large-area sampling
For large-area sampling, the organic solvent-based method was used to perform the MAC
sampling, where the same vacuum cleaner was used for composite collection from five coupons (5 x 929
cm2; total area = 4645 cm2) (Table 3-11). Each composite-area coupon was sampled using the multi-pass
procedure that was used for the medium-sized coupon sampling method (Appendix A, Section A-1). The
vacuumed sampling liquids were collected and processed as one composite sample. The results from
Phase III large-area testing are provided in Section 5.5.
Table 3-11. Experimental Parameters of Phase III Multivariate Characterization of Commercial
Vacuum Cleaner Testing - Large-Area Composite Sampling
Wet
Test
Target
Surface
Wetting
Wetting Agent
Volume
Lapse
Time
Test
Area
Vacuum
Cleaner
Material
Chemical
Cone.
Agent*
Pre-rinse
(mL)
Surface
Wetting
(mL)
Post-
rinse
(mL)
(s)
Matrix?
Sampled*
(cm2)
Bissell
Stainless
steel
TEP
2-CEPS
100%
Isopropyl
alcohol
50
5x50
100
10
3xTC
1 x PB
1 xCS
4645
"Detailed information on test materials, wetting agents, and commercial wet-vacuum is found in Section 3.0.
sample.
t Per each test condition, t Per replicate composite
32

-------
3.9 Supplementary Surface Concentration Verification Tests
3.9.1 Verification of surface concentrations by direct extraction
The efficacy of wet-vacuum sampling is a combination of the affinity of the target chemical with the
material, the wetting agent (either aqueous or organic solvent-based) and the mechanical cleaning action of
the vacuum cleaner. During wet-vacuum sampling, both the applied vacuum and the vacuum cleaner
nozzle/brush scrubbing action aid in the removal of the chemical from the contaminated surface. A series of
supplementary tests was performed to investigate the ability of wetting solvents to extract selected
chemicals from the reference material without mechanical scrubbing. These tests were performed to verify
the potential of the wetting solvent for lifting and absorbing chemicals from the surface. Three composite-
area stainless-steel coupon sets (12 [5.1 cm x 5.1 cm dimensions] coupons per set) were spiked individually
with one discrete droplet (2-microliter (jjL) volume) with the chemical at the same spiking concentration as
the wet-vacuum test coupons, i.e., at 100%, 20%, 10%, 2%, and 1% of the default surface concentration.
The spiking approach was like the approach used to spike the medium size area (Figures 3-1 and 3-2) but
now with 12 individual coupons arranged below the template. These small coupons did not undergo wet-
vacuum sampling, but after the 30-min contact time, they were extracted using various aqueous and organic
solvent wetting agents as well as directly in hexane. Extraction of each set of twelve coupons was
performed in 150 mL of solvent using a custom-made extraction coupon holder made of stainless-steel wire
and Teflon (Figure 3-7).
Figure 3-7. Surface spike controls ready for extraction.
After extraction, the resulting solutions underwent the same analytical process as the wetting liquid
samples from the wet-vacuum sampling (Sections 4,1 and 4.2). Samples were processed for analysis as
described in Section 4.2. Test results are provided in Section 5.4.3.
3.10 Supplementary Tests
3.10.1 Operational time estimates
The operational time estimates for wet-vacuum-based versus wipe-based sampling methods were
prepared for each experimental step using laboratory data logs and laboratory notebooks. These estimates
do not include the time needed to clean or contaminate materials for testing, as these steps would not be
part of an actual field-sampling event. Therefore, only four procedural steps were included in the operational
33

-------
time estimates: (1) Preparation of sampling kits; (2) surface sampling; (3) extraction process; and (4)
preparation of samples for analysis. These time estimates are provided in Section 5.6.
3.10.2 Flow and temperature profiles of wet-vacuum system
3.10.2.1	Measurement of vacuum airflow
The airflow rate of the wet-vacuum system was measured at the EPA Metrology Laboratory in
Research Triangle Park, NC. A flow meter (Roots Meter EM175, Dresser, Chagrin Falls, OH, USA) was
placed in line with the wet-vacuum collection hose to record the volume of air moving through the wet-
vacuum. The dispenser tubing was removed from the vacuum hose and penetrations through the vacuum
hose were sealed. The vacuum nozzle was placed on the surface of a medium-sized stainless-steel TC and
clamped in place at a 30-degree (approximate) angle to the coupon surface to simulate the sample
collection procedure. Airflow was measured for 20 min to simulate the timeframe associated with sampling
approximately 20 medium-sized coupons. Results are provided in Section 5.7.1.
3.10.2.2	Temperature measurement of wet-vacuum unit
A series of tests was conducted to determine the maximum temperature that vacuum cleaner
components reach after sampling a large-sized surface area. The tests were performed to locate any high-
temperature areas (hot spots) within the wet-vacuum system during prolonged sampling. A medium-sized
stainless-steel coupon was placed inside a plastic test box under a chemical hood. The vacuum unit holding
tank was filled with Dl water, and the coupon and vacuum cleaner were conditioned (pre-rinsed) with 50 mL
of Dl water. After the pre-rinse was completed, the surface was sampled using a standard wet-vacuum
procedure (Appendix A, Section A-1). The spray and vacuum sampling were repeated from 4 to 20 times to
mimic periods of various lengths of (composite) sampling. Following vacuum sampling, the system was
rinsed with 100 mL of Dl water aspired from a 1000-mL beaker. The wet-vacuum system was filmed and
photo-documented using a thermal imaging camera (FLIR E4, Wilsonville, OR, USA) during the entire
sampling process. This infrared camera is equipped with an uncooled microbolometer detector and offers
infrared (IR) resolution of 80 * 60 pixels combined with multispectral dynamic imaging (MSX®)
enhancements for the spectral range of 7.5-13 jjm. Results are reported in Section 5.7.2.
34

-------
4.0 Sampling and Analysis
4.1 Surface Sampling Methods
4.1.1 Medium-size area wet-vacuum sampling
The experimental design that was used for medium-sized discrete (1-point, 1 x 929 cm2) SA
sampling is shown in Figure 4-1. A detailed description of the wet-vacuum sampling process is provided in
Appendix A, Section A~1. Extraction methods are summarized in Section 4.2.
Weigh Dirty Liquid Tank (DLT)
Weigh Wetting Tank (WT)
30 sec before the end of weathering
condition vacuum (50 mL pre-rinse)
50 mL
pre-rinse
100 mL
post-rinse
Spray coupon with
wetting agent.
Weigh WT.

—4r+



"f



"t"h

r"
-¦+
Spike coupon with target chemical
for discrete medium-size location
sampling:
5 x 1-point TC + 1 x1-pointPB
Allow chemical weathering
(CT = 30 min)
After a 10 sec wetting agent-surface
contact time start vacuum sampling
Vacuum surface with horizontal then
vertical overlapping strokes.
Weigh DLT
Single area discrete sample (929 cm2)
After sampling is
finished rinse vacuum
(100 mL post-rinse)
Weigh DLT
Transfer sampling liquid from DLT
into clean beaker, mix
Transfer an aliquot of sampling
liquid to 40 mL vial for extraction
(aqueous solutions) or preparation
for analysis (IPA)
Figure 4-1. Experimental design and sample flow for SA sampling. TC - test coupon; PB - procedural blank; CT
- contact time; WT - wetting tank; DLT - dirty liquid tank
4.1.2 Large-area wet-vacuum sampling
Large-area sampling was designed to test the efficiency of wet-vacuum sampling for composite-
area (5-point, 5 x 929 cm2) MAC sample collection. This sample compositing approach was developed for
improved turnaround times and cost of analysis. The multistep experimental design that was used for MAC
sampling is shown in Figure 4-2. A detailed description of the wet-vacuum sampling process is provided in
Appendix A, Section A-1. Extraction methods are summarized in detail in Section 4.2.
35

-------
Step 0
Step 1
Weigh Dirty Liquid Tank (DLT)
Weigh Wetting Tank (WT)
30 sec before the end of weathering
condition vacuum (50 mL pre-rinse)
50 mL
pre-rinse
100 mL
post-rinse
Spray each coupon
with wetting agent.
Weigh WT after each
finished rinse vacuum
(100 mL post-rinse)
Weigh DLT
-j"'

M





j


f




~
~
1
-!—i.
|4
i--*—
-1—j-
4-4
j--*—
4-4-
-f-t
Multi-area composite sample (4645 cm2)
¦	Transfer sampling liquid from DLT into
__ clean beaker, mix
¦	Transfer an aliquot of sampling liquid to 40
mL vial for extraction (aqueous solutions)
or preparation for analysis (IPA)
Spike five coupons with target
chemical for multi-location
composite sampling:
3 x 5-point TC + 1 x 5-point PES
Allow chemical weathering
(CT = 30 min)
After a 10 sec wetting agent-surface
contact time start vacuum sampling
Vacuum each surface with
horizontal then vertical overlapping
strokes.
Weigh DLT after sampling of each
coupon
Figure 4-2. Experimental design and sample flow for MAC sampling. TC - test coupon; PB - procedural blank; CT - contact time; WT - wetting tank; DLT
dirty liquid tank
36

-------
4.1.3 Small-area wipe-based sampling
Surface wipe sampling was performed on a small central area (1 point, 100 cm2; Figure 4-3) for
comparison with wet-vacuum method performance and to sample for residual chemical on the surface post-
sampling in mass balance tests. Each wipe was used following a four-step process. A series of horizontal to
vertical strokes followed by diagonal strokes and then perimeter wiping strokes were used, where the wipe
was folded after each step (with the contaminated side folded inward). Types of wiping media, wetting
solvents, and amount of wetting solvent for all material-chemical combinations tested are given in Section
3.7.2. The wipe extraction methods are summarized in Section 4.2.


Prepare sampling kits
Spike central area of each coupon
with target chemical for discrete
small-area location sampling:
5 x 1-point TC + 1 x 1-point PB
Allow chemical weathering
(CT = 30 min)
¦	After a 30 min contact time start wipe sampling
using 2 pre-wetted wipes per 10x10 cm area
¦	Wipe sample with horizontal, vertical, diagonal
overlapping strokes, followed by perimeter
sampling
Single area discrete sample (100 cm2)
I
E
J,
Transfer both sampling wipes to 50 mL
amber glass jar for 2-step sequential
extraction
Figure 4-3. Experimental design and sample flow for small area sampling. TC - test coupon; PB - procedural
blank; CT -contact time
37

-------
4.2. Extraction Methods
This section summarizes the extraction procedures used for all surface-wipe and sample liquid-
chemical combinations that resulted from the wipe-based and wet-vacuum-based sampling procedures. All
wipe and wetting liquid extraction methods were described in Section 3.7.
4.2.1	Extraction of surface wipes
After completion of wipe sampling (Section 4.1.3), the two wipes used for surface sampling were
placed together in a precleaned 40-mL wide-mouth extraction vial with a polytetrafluoroethylene-lined lid for
composite extraction. The extraction vial was filled with 15 mL of dichloromethane (DCM) (Optima™, high-
performance liquid chromatography/spectrophotometry, GC/MS, and pesticide residue analysis grade
Fisher Chemical, product no. H 303-4 UN1208), capped, and transferred to the sonicator for step 1 of the
sequential extraction. Wipe samples were sonicated for 15 minutes. For step 2, the resulting liquid extract
was transferred to a 40-mL glass vial and a second 15-mL aliquot of DCM was added to the two wipes.
Then, the wipe samples were sonicated for another 15 minutes. For step 3, the step 2 extract was removed
and combined with the extract from step 1 and refrigerated at 4 ± 2 °C until further processing. Sample
preparation for instrumental analysis is described in Section 4.3.
4.2.2	Extraction of water and water-based wetting agents
Target chemicals were extracted from the liquid samples generated during wet-vacuum procedures
using a simplified LLE developed based on physicochemical properties of the target analytes, the sampling
solvents, and existing standard methods [1], All sampling liquids were extracted within one hour of sample
collection because method development studies had shown that some chemicals have a limited stability in
water and water-surfactant solutions (Section 5.1).
The recovered sampling liquid volume was measured through the weight measurement of the liquid
waste collection tank before and after the sample collection. Sample liquids were transferred to a clean
extraction beaker. For foaming wetting agents, the collected liquid was allowed to settle (10-15 min period)
until the foam dissipated. This extract settling step was especially important for water-surfactant solutions
with large expansion volumes for which large amounts of foam were observed (e.g., Dawn Ultra® water
wetting agent). Figure 4-4 shows an example of the liquid waste in the collection tank immediately after
sampling and during the settling phase as well as the settled liquid ready for LLE.
38

-------
Figure 4-4. Dawn Ultra® water SL immediately after conclusion of sampling (a), during settling phase in the dirty
liquid tank (b) and settled liquid (aliquot in beaker) ready for LLE (b),
The reagent volumes and amounts of additives to salt out the surfactants used for LLEs are listed in
Table 4-1. For each LLE procedure, a 5- to 10-mL aliquot of sample liquid was transferred to a 40-mL
extraction vial preloaded with the appropriate amount of additive(s) (Table 4-1). After addition of extraction
solvent, each vial was capped, and the contents were manually shaken for 1 minute (min). After the
aqueous and solvent layer separated, the entire extract layer was carefully collected using a Pasteur pipette
and placed into a clean 40-mL vial with graduated markings. For TEP extractions, the procedure was
performed twice, and the resulting extracts were combined. The total extract volume was recorded.
Table 4-1. Chemical-Specific Experimental Conditions for Extractions of Aqueous SLs

Extraction
Reagent Volumes

Other
Wetting Agent
Solvent Type
Wetting
Agent (mL)
Extraction
Solvent (mL)
Additives
Steps
TEP
Water
DCM
10
2x5mL
0.5 g NaCI; 10 pL 1 M
NaH2P04/Na2HP04; buffer pH 7.0
Extract dried with 1
g Na2S04 cartridge
Water-Tween®
DCM
10
2x5 mL
5.0 g NaCI; 100 pL of 1 M
NaH2P04/Na2HP04, buffer pH 7.0
Extract dried with 1
g Na2S04 cartridge
Water-Dawn Ultra®
DCM
10
2x5mL
0.5 g NaCI; 10 |jL of 1 M
NaH2P04/Na2HP04, buffer pH 7.0
Extract dried with 1
g Na2S04 cartridge
Water- SuperSoap®
DCM
10
2x5 mL
0.5 g NaCI; 10 |jL 1 M
NaH2P04/Na2HP04, buffer pH 7.0
Extract dried with 1
g Na2S04 cartridge
2-CEPS
Water
Hexane
10
10
None
Extract dried with 1
g Na2S04 cartridge
Water-Dawn Ultra®
Hexane
10
10
5 g NaCI
Extract dried with 1
g Na2S04 cartridge
39

-------

Extraction
Reagent Volumes

Other
Wetting Agent
Solvent Type
Wetting
Agent (mL)
Extraction
Solvent (mL)
Additives
Steps
Water-SuperSoap®
DCM*
10
10
3 g NaCI
Extract dried with 1
g Na2S04 cartridge
Nitrobenzene
Water
Hexane
10
10
None
Extract dried with 1
g Na2S04 cartridge
Water-Tween®
Hexane
10
10
None
Extract dried with 1
g Na2S04 cartridge
'Unsuccessful with hexane
The extract was then passed through a short drying cartridge prepacked with granular anhydrous sodium
sulfate (Bond Elut sodium sulfate drying cartridges, Agilent Technologies, Santa Clara, CA, USA; product
no. 12131033) to remove residual surfactant water from the organic extract. The eluent was collected in a
graduated centrifuge tube (Figure 4-5). After the drying step, the sodium sulfate cartridge was rinsed with
solvent until the final sample volume was 10 mL.
Figure 4-5. Drying of the SL extract
After drying, extracts were transferred to a 12-mL vial and refrigerated at 4 ± 2 °C until preparation
for analysis (Section 4.3).
4.3 Preparation of Samples for Analysis
Extracts generated from sample liquids (Section 4.2.2), surface wipes (Section 4.2.1), and IPA
wetting solvent were prepared for analysis in 1,8-mL amber glass gas chromatography (GC) vials.
Depending on the type of sample, extracts underwent up to 100-fold dilution. Briefly, an aliquot of raw
extract was drawn using an appropriate-sized micropipette and added to a GC vial filled with a premeasured
amount of hexane or IPA. The control spike samples were also diluted up to 25-fold, If analytical results
were outside the calibration range, the analytical laboratory performed necessary dilutions and reported
dilution factors along with quality control (QC) data. The samples were refrigerated at 4 ± 2 °C or below prior
40

-------
to delivery to the local EPA laboratory or shipment to the external chemical analysis laboratory. All analytical
batches were accompanied by a chain of custody form and inspected at the analytical laboratory upon
receipt.
4.4 Instrumental Analyses
Instrumental analyses were performed at an accredited external laboratory (EMSL Analytical, Inc.,
Cinnaminson, NJ) or by the EPA Organic Support Laboratory (OSL). The standard methods used for each
analyte are listed in Table 4-2. The instrumental parameters and conditions for instrumental analyses are
listed in Appendix A (Tables A-1 through A-6). Analyses by the EPA OSL were limited to Phase III samples
associated with the wet-vacuum tests described in Section 3.8.3.
Table 4-2. Instrumental Methods Used for Analysis of Target Analytes
Target Analyte
Experimental Phase
Instrumental Method
Reference Method
Analytical
Laboratory
Phenol
Phase I, Phase II
GC/FID*; Spectrophotometry?
NIOSH 2546* [10]; EPA420.1 [11]
EMSL Analytical, Inc.
Nitrobenzene
Phase I, Phase II
GC/FID
NIOSH 2005* [12]
EMSL Analytical, Inc.
TEP
Phase I, Phase II
GC/FID
NIOSH 5034* [13], NIOSH 5038* [14]
EMSL Analytical, Inc.
Phase III
GC/MS
Not available
EPA OSL
2-CEPS
Phase III
GC/MS
Not available
EPA OSL
*ln IPA. Tin aqueous samples; * Modified gas chromatography/flame ionization detector (GC/FID); GC/MS - gas chromatography/mass
spectrometry.
For EMSL Analytical, Inc. analyses, a calibration range of 1-100 jjg/mL for all target analytes (7-
point calibration curve; 1-10-20-40-60-80-100 jjg/mL) was used for initial calibration (ICAL), with reporting
limit verification (RLV) and initial calibration verification (ICV) analyses performed at the lowest and mid-
calibration level, respectively, prior to each analytical run. In addition, prior to each analytical run, a
laboratory control sample (LCS) and the laboratory control sample duplicate (LCSD) were analyzed. A
continuous calibration (CC) standard at a mid-level concentration was analyzed for every 10 samples, with a
calibration end check (EC) performed at the end of each analytical run. Additional QC samples included
duplicate instrument injections (DUPs) of test samples and laboratory blanks. Samples with results below
the lowest calibration point (i.e., 1 jjg/mL) were reported as less than the limit of quantitation (
-------
Table 4-3. QC Checks for Instrumental Analyses Performed by External Laboratory
QC Check
Acceptance Limits
Initial Calibration (ICAL)
7-point initial calibration prior to analysis*
Reporting limit verification at lowest point concentration (RLV)
60-140% of lowest ICAL
Initial calibration verification at midpoint concentration (ICV mid)
80-120% of midpoint ICAL
Laboratory control sample at midpoint concentration (LCS)
70—130%of expected concentration
Laboratory Control sample duplicate at midpoint concentration (LCSD)
<25% RPD
Continuous calibration (CC) at midpoint concentration
80-120% of midpoint ICAL
End check of calibration (EC) at midpoint concentration
80-120% of midpoint ICAL
Duplicate injections (DUP)
<25% RPD
Laboratory (solvent) blank (LB)
0.995). The continuous calibration was performed using a mid-concentration
calibration standard, that is, approximately every 10 test samples and at the end of the analytical run, with
an acceptance control limit of 80-120% of the ICAL concentration. If QC criteria were not met, the
instrument was recalibrated, and any affected samples were reanalyzed. Additional QC samples included
duplicate injections of test samples (one duplicate injection per analytical run; acceptance criteria: relative
percent difference [RPD] <20%) and analysis of blanks (procedural blank and laboratory solvent blank [LB]).
Prior to testing, an initial laboratory proficiency evaluation was performed for both laboratories.
Accuracy and precision were determined by analysis of multiple measurements of control spike solutions (n
= 3-5 for each concentration level; single analytical run). Control spike samples were generated by spiking
the target chemical or target chemical solution used during testing directly into the injection solvent (e.g.,
hexane, IPA, water). All control spikes were sonicated for 10 minutes and then diluted as needed. Each
control spike set was accompanied by one laboratory blank sample (1 mL of solvent used for preparation of
samples for analysis). These control spike experiments were used as independent verifications of the
results obtained from the outside chemical analysis laboratory. The initial and continuing laboratory
proficiency results are listed in Table 4-4.
42

-------
Table 4-4. Initial and Continuing Laboratory Proficiency Results
Target Chemical and Analytical
Laboratory
Spike Control Sample Analysis Results

Accuracy and Precision
Number of Samples Analyzed
Solvent Blank

(% of true value ± 1 SD; RSD [%])
(n)

EMSL Analytical, Inc.
Nitrobenzene (initial*)
81.6%± 14.7% SD; RSD=18%
2

-------
4.5 Data Reduction Procedures
4.5.1 Chemical concentration in sampling liquid calculations
The final sample concentration results (in jjg/mL) for liquid samples that did not undergo extraction
(IPA as the wetting liquid) were converted to total mass of chemical per sample (mg/sample or sample
composite) by multiplying by the collected sample volume and dilution factor, if applicable:
where:
Ms(ipa) = mass of chemical in sample (mg)
Cs = concentration (|jg/mL) from an individual replicate sample of IPA sampling liquid (SL)
Vt = volume of SL collected (mL)
Df = sample dilution factor prior to analysis (if any)
For aqueous solutions that were extracted using liquid extraction, a volumetric fraction conversion
factor (CFV), considering a fraction of the total SL collected (Vt) to volume of SL extracted (Ve), was applied
to chemical mass calculations as follows:
Ms(aqueous) = Cs * Vt * Df * 1,000 * OFv	(^Q- ^)
where:
Ms(aqueous) = mass of chemical in sample (mg)
Cs = concentration (|jg/mL) from an individual replicate sample of extracted aqueous SL
Vt = volume of SL collected (mL)
Df = sample dilution factor prior to analysis (if any)
CFV= volumetric fraction conversion factor calculated according to Eq. 3:
where:
CFV = volumetric fraction conversion factor
Ve = volume of SL extracted (mL)
Vt = volume of SL collected (mL)
The percent recovery of the chemical from the QC samples (e.g., CSs) was calculated against the
theoretical chemical amount spiked into the solution as follows:
Ms(ipa) - Cs x Vt x Df x 1,000
(Eq. 1)
CFv = 1/(Ve/Vt)
(Eq. 3)
44

-------
%Rqc = Cqc/(Vsp x ScA/t/Df) x 100%	(Eq. 4)
where:
%Rqc = percent recovery for an individual QC sample (versus theoretical)
Cqc = concentration (|jg/mL) from an individual replicate QC sample
Vsp = volume of spike (mL)
Sc = concentration of chemical in spiking solution (|jg/mL)
Vt = total sample volume (mL)
Df = sample dilution factor prior to analysis (if any)
The chemical mass (Ms) results used for sampling efficiency calculations were not adjusted for QC
sample recovery (%Rqc).
4.5.2 Sampling efficiency calculations
The sampling efficiency was calculated using the mean of the chemical mass recovered from the
replicate samples compared to the theoretical amount spiked onto a coupon surface:
XSE = x(TC/TS)n x 100%	(Eq. 5)
where:
XSE = mean sampling efficiency (%)
TC = mean chemical amount recovered from replicate TCs or coupon composites (mg)
TS = mean chemical amount spiked onto coupons or coupon composite surface (mg)
The mean sampling efficiency (xSE) was calculated as the arithmetic mean for each set of three to
five replicates (n= 3 to 5), along with the associated standard deviation (SD) and coefficient of variation
(relative standard deviation, %RSD). If the sample (or sample composite) concentration was found to be
below the lowest point of the calibration curve, but the signal-to-noise (S/N) ratio was greater than 10, the
results were reported and flagged according to National Functional Guidelines [15], The samples where
analytes were not detected (not found or S/N <3) were reported as ND (non-detect). The analysis of
variance (ANOVA) was used to check if the observed differences in sampling efficiencies of various
methods tested are statistically significant. The p-values are reported at significance level of 95% (a=0.05).
45

-------
5.0 Results
The wet-vacuum method assessment results are discussed in the order of the method
development/proof of concept steps in Phase I, utilizing the prototype wet-vacuum system; Phase II,
optimization of wet-vacuum method using two commercially available wet-vacuum systems and Phase III,
the demonstration of the optimized commercial wet-vacuum system. Each phase has in common that the
collected sample liquid should be analyzed based on validated analytical protocols. All analytical methods
(Table 4-1) for the analysis of organic solvent-based and water-based sampling agents were verified prior to
the collection of any wet-vacuum samples. These results are summarized in Section 5.1.
5.1 Analysis of Sampling Liquids
The wet-vacuum methodology was tested in two wetting solvent configurations: organic solvent
(IPA) and aqueous solution. As discussed previously, the resulting sampling liquids were either directly
analyzed using the appropriate analytical method (Section 4.4) or extracted using a straightforward LLE
procedure (Section 4.2.2).
The ability to directly analyze the IPA sampling liquid was considered one of the main advantages
of the wet-vacuum sampling method that was developed in this study, primarily due to the ease of
downstream treatment post-sampling. There was no extraction step, and the sample preparation for
analysis was limited to dilution of samples and addition of surrogate or IS compounds, when applicable
(Section 4.3). The instrumental methods used for direct analysis of IPA (GC/MS, GC/FID; Table 4-1) were
selected based on the general applicability/compatibility of target chemicals with IPA, accessibility, and the
cost of sample analysis. All phenol-containing sampling liquids were directly analyzed using a
spectrophotometric method that is used for drinking, surface, and saline waters, as well as domestic and
industrial wastes [11], The results of direct analysis of simulated liquids samples that were submitted to the
analytical laboratories prior to testing were all above 90% of the amount of chemical spiked into the solvent.
Target analyte-specific results are given in Appendix A (Tables A-8 through A-10).
Apart from the above-mentioned analysis of phenol, LLE procedures were necessary for
preparation of water-based sample liquids for instrumental analysis. The LLE procedures were based on the
fundamental phenomenon of the partition of the analyte between two immiscible (aqueous and organic)
phases, one being the sample liquid and the other the organic extraction solvent. ForTEP, the extraction
was performed with addition of phosphate buffer to adjust the pH to a range where the analyte is nonionized
and more easily migrates into the organic phase. Sodium chloride was added to salt out the surfactant from
some of the aqueous SL solutions. Target analyte-specific recoveries of the optimized extraction method are
provided in Appendix A (Table A-7). A summary of the results for the 1-hour and 24-hour holding times from
control sample preparation (a simulated sample liquid) to extraction for these water-based samples is
presented in Figure 5-1. A 24-hour HT was tested only for chemicals that were evaluated during operational-
scale testing to obtain initial information on stability of the sample liquid from the wet-vacuum process.
During field operations, it is likely that the sample handling time prior to analysis would be at least 1 day,
equivalent to overnight shipping and a 1-day analytical turnaround time (meaning same-day extraction and
analysis by the analytical laboratory). Results for nitrobenzene extractions from water and Tween®-20 (HT =
1 hour) are given in Table A-7 (Appendix A).
46

-------
Results of liquid-liquid extraction
of SuperSoap® sampling liquids - control samples
100%
80%
| 60%
o
a>
g. 40%
20%
0%
in
TEP, HT=1 h
2-CEPS, HT=1 h
0TEP, HT=24 h
2-CEPS, HT =24 h
Results of liquid-liquid extraction
of DawnUltra® sampling liquids - control samples
100%
80%
40%
20%
o%
i TEP, HT=1 h
2-CEPS, HT=1 h
0TEP, HT=24 h
2-CEPS, HT =24 h
Results of liquid-liquid extraction
of water sampling liquids - control samples
100%
80%
60%
40%
20%
0%
¦
¦ TEP, HT=1 h
2-CEPS, HT=1 h
0TEP, HT=24 h
2-CEPS, HT =24 h
Figure 5-1. Results from LLE of TEP and 2-CEPS in control samples of aqueous sample liquids; (a) SuperSoap®-
water LLE, (b) Dawn Ultra®-water LLE, (c) water LLE; HT, analytical HT defined as from control sample
preparation to extraction.
47

-------
The results show that LLE method extracted more than 90% of target analytes from all types of
water-based wetting agents. The lower recoveries of 2-CEPS in all aqueous sampling liquids types
analyzed with a 24-hour HT emphasizes the importance of liquid sample preservation immediately after
collection to ensure stability of the target analyte prior to extraction. The LLE method is relatively simple and
inexpensive. However, the LLE was also time-consuming and involved the use of relatively hazardous
solvents (dichloromethane, hexane). Another drawback was that the solvent choice must be optimized for
each analyte to be sampled using the wet-vacuum method, even for chemicals with similar physicochemical
properties. Therefore, the LLE can never be specific to a particular sampling liquid, as the target chemical
polarity will determine the ability to extract the solvent from the wetting agent matrix, making the procedure
not easily expandable to possible further automation. Moreover, other chemicals co-collected during wet
sampling may affect the effectiveness of the LLE procedure. In this laboratory study, all samples were
extracted within 1-hour post-collection. Hence, the addition of preservatives to samples that contained 2-
CEPS (e.g., glacial acetic acid and sodium chloride recommended for preservation of aqueous samples that
contain CWAs [16]), was not considered.
5.2 Proof-of-Concept and Initial Optimization of Surface Sampling Method -
Phase I
The target chemicals were characterized by various degrees of solubility in water. Hence, the
selection of the most appropriate, and ideally universal to all chemical-surface combinations, wetting
condition was considered one of the first and most important steps of the method optimization. In the initial
test, 100 mL was chosen as the wetting volume for a 926 cm2 surface area, and 100 seconds was chosen
as the initial wetting agent lapse time on the surface. Testing was performed on reference materials
(stainless steel) using the wet-vacuum prototype system. The wetting solvent selection experiment results
are summarized in Figure 5-2a, and test-specific results are given in Appendix A.
Recoveries of the wetting agent were the highest for the water-based agents (90%) while only 50%
of the IPA was recovered in part due to evaporative IPA losses on the surface and within the prototype wet-
vacuum system (Figure 5-2a). Nevertheless, the IPA wetting agent offered the highest recoveries of all
target analytes, on average >60% of initial surface loading for chemicals that represented all solubility
classes. Since it was also determined to be the easiest to handle because sample preparation was not
needed prior to analysis, it was chosen to subsequently select the optimal lapse time (time between wetting
and wet-vacuuming of the surface). Despite differences in average solvent recoveries between tests (Figure
5-2b), there was no statistically significant difference in chemical recovery (p=0.12) for three lap times tested
(1-100 sec). Then, an optimal amount of solvent for wetting the test surfaces was determined (using water)
in a series of gravimetric tests; no significant difference was observed in recovery for 1 and 10 s lapse times
(average 66-68%; p=0.35), the average recovery for 100 s lapse time was lower at average 52% (± 2.7%
SD). The 10-sec long lapse time was considered a reasonable amount of time between wetting agent
application and sampling while 50 mL of wetting solvent volume was considered to provide good surface
coverage without excessive runoff (which occurred for the 100-mL wetting volume). This conservative
amount of solvent was considered for the more volatile wetting agent such as IPA. Consequently, these
parameters (lapse time = 10 sec; 50 mL wetting volume) were selected for evaluation of the laboratory-scale
wet-vacuum method on three building materials of different porosities and permeabilities: stainless steel
(SS), vinyl flooring (VF), and plywood (PW). The recoveries from different materials are shown in Figure 5-
2c.
48

-------
100%
80%
^ 60%

O 40%
O

O
O

-------
recovery of target chemicals from plywood was performed; IPA wetting solvent volumes of up to 100 mL per
coupon were tested. Because no statistically significant improvement was observed in the sampling
efficiency of plywood using larger volumes of IPA (data not shown), the permeable material was excluded
from the next stages of experimental testing. A smooth and relatively nonporous laminate flooring material
was used instead during the initial evaluation of commercial vacuum cleaners for wet-vacuum sampling
(Phase II, section 5.3.1). Shorter lapse times (less than 1 s) would potentially have overcome some of the
low wetting agent recoveries for plywood. However, this should be balanced against the specific time for a
chemical to dissolve into the applied wetting solvent.
5.3 Initial Evaluation of Commercial Wet-Vacuum Cleaners - Phase II
The commercial vacuum cleaners used in the initial evaluation of wet-vacuum sampling of
chemicals were previously successfully deployed for environmental sampling of biological threats [17], Both
units are designed for cleaning tough surface stains and, hence, were expected to effectively lift
contamination from surfaces and efficiently collect the wetting agent.
5.3.1 Comparison of commercial wet-vacuum cleaners
The initial wetting liquids selected for the commercial wet-vacuum systems testing were IPA and
water. Due to large IPA losses observed during initial commercial wet-vacuum testing (data not shown), the
sampling performance of these wet-vacuum units was initially tested using water only. The results for all
chemical-material type commercial vacuum cleaners tested are shown in Figure 5.3. Sample-specific results
are given in Appendix A (Tables A-20 through A-28).
The Bissell unit offered better sampling efficiency for all classes of chemicals and all types of
surfaces tested (Figure 5-3). The unit was more efficient at collecting the wetting solvent, 59-68%,
depending on the type of surface sampled, outperforming the Rug Doctor cleaner (33-53%) on the same
type of surfaces.
50

-------
D
80%
Phase II: Evaluation of commercial vacuum cleaners
- stainless steel
60%
g> 40%
o
CJ
 40%
o
o
(1)
or
20%
0%
Bissell
Rug Doctor
i Nitrobenzene
Phenol
i TEP
Solvent recovery
Phase II: Evaluation of commercial vacuum cleaners
- vinyl flooring
80%
60%
a:
20%
0%
h
Nitrobenzene
Phenol
bTEP
I - Solvent recovery
Bissell	Rug Doctor
Figure 5-3. Results of the initial evaluation of commercial vacuum cleaners for wet-vacuum sampling of various
test surfaces; sampling was performed using 50 mL of water, LT=10 s; material types: stainless steel (a);
laminate flooring (b); vinyl flooring (c)
51

-------
Like the prototype testing, one of the most important factors influencing the sampling efficiency by
commercial cleaners was the water solubility of the chemical. Figure 5-4 shows the chemical recovery
achieved during water-based wet-vacuum sampling using Bissell and Rug Doctor cleaners as a function of
the solubility of the chemical in water. The very low recovery of very slightly water-soluble nitrobenzene
(Table 3-2) indicates that the solubility of the chemical in water is likely to be a limiting factor for collection
efficiency during water-based wet-vacuum sampling. In the case of nitrobenzene, the mass delivered to the
medium size coupon was 29 mg which, if recovery of nitrobenzene in water as the wetting solvent was
100%, would have yielded a 0.60 mg/mL sample concentration. This concentration is a factor of 3.2 below
the 1.9 g/L solubility of nitrobenzene in water. Similar ratios of solubility over theoretical sample
concentration are much larger for phenol and TEP, namely 160 and 914, respectively. The rate at which
chemicals dissolve in water is expected to slow down when the solubility concentration is approached as is
the case for nitrobenzene. Hence, a longer contact time of water as the wetting agent on the surface may
have improved the recovery for nitrobenzene. One other approach to improve on recovery may have been a
repeated application of the wetting agent on the same surface area. Neither was further investigated in this
study considering the high recoveries using IPA as the wetting agent.
Sampling efficiency of commercial wet-vacuum systems
using water as wetting agent vs. target solubility in water
II
tl
TEP
100	200	300	400	500	600
Solubility (g/L)
•	Bissell - Stainless Steel	BRug Doctor - Stainless Steel
•	Bissell - Laminate	¦ Rug Doctor - Laminate
Figure 5-4. Results of the initial evaluation of commercial vacuum cleaners for wet-vacuum sampling of various
test surfaces as a function of the solubility of the chemical in water
Based on solvent and chemical recovery results using water as the wetting solvent, the focus of the
next experimental phase of testing was on improving the efficiency of IPA-based sampling, which offered
better chemical recoveries across different levels of water solubility, as demonstrated in the prototype wet-
vacuum testing (Figure 5.2a). Since the IPA solvent recovery appeared to be an important factor
52

-------
contributing to overall sampling efficiency, the commercial unit-based IPA sampling was further optimized
using the Bissell cleaner only, as described in Section 5.3.2.
5.3.2	Optimization of the commercial wet-vacuum cleaner-based sampling
The last step of the wet-vacuum sampling optimization using a commercially available unit included
addition of a pre-sampling conditioning step of the wet-vacuum unit (pre-rinse) and a post-sampling rinse
step of the wet-vacuum unit (post-rinse), both performed with IPA as the wetting solvent. Results of this
optimization step are listed in Appendix A (Table A-29 and A-30). The pre-rinse step was designed to
mitigate solvent evaporation related to losses of IPA; as observed during the initial stages of wet-vacuum
sampling when the system was not pre-wetted (i.e., "primed") with the IPA solvent. The optimal volume of
solvent for the prewetting step was 50 mL of IPA for the test Bissell unit. A 100 mL prewetting volume
provided similar improvement in absolute recovery of IPA (data not shown) but resulted in significant
carryover of solvent from this pre-rinse step into the dirty liquid tank. The post-rinse volume selected for
further testing was 100 mL, aspired directly from a clean beaker containing IPA after sampling. A post-rinse
volume of 100 mL was chosen over a 200-mL rinse to maintain a more conservative final sample volume.
Additional chemical mass balance tests, performed during operational-scale testing (Phase III),
demonstrated that the amount of chemical remaining in the vacuum cleaner system after application of the
100-mL rinse was relatively minimal, about 10% of the total mass of chemical recovered during wet-vacuum
sampling (results from mass balance tests are discussed in Section 5.4.1.1).
5.3.3	Wipe sampling
The last test performed before the operational testing of the wet-vacuum method was related to the
wipe sampling method development. The surface wipe sampling method was used in a direct comparison
with the wet-vacuum method that had been developed. As discussed before, the wipe-based method was
adopted from established methods used for sampling CWAs [9], Results are given in Table 5-1.
Table 5-1. Results for Wipe Sampling and Extraction Optimization Tests for 2-CEPS and TEP
Target
Chemical
Concentration
(mg)
Control Spike
Recovery
Mass % Rec.
(mg)
Wipe Spike
Recovery
Mass % Rec.
(mg)
Mass
(mg)
Chemical Recovery from
Wipe Sampling Method*
STD % Rec. STD
(mg) (%)
RSD
(%)
2-CEPS/Stainless Steel

2.8
100
2.9
98
2.00
0.014
70
0.49
0.71
TEP/Stainless Steel

2.4
89
1.9
71
0.55
0.54
21
20
96
'Per test area of 10 cm x 10 cm (100 cm2), n = 3.
The sampling of 2-CEPS from nonporous reference material (stainless steel) was characterized by
good average recovery (70% ± 0.49% SD of the theoretical target) and reproducibility (%RSD <1%) (Table
5-1). However, the average recovery of TEP with wipe-based surface sampling was only 21% of the
theoretical concentration delivered onto the stainless-steel surface and was characterized by large intra-test
variation (RSD = 96%; n= 3) (Table 5-1). The high recovery of a control wipe sample that was spiked with
53

-------
the target chemical prior to extraction (71% of target concentration) suggests that the wipe extraction
method (as described in Section 4.2.1) was effective for extracting TEP from the gauze wipes used in
sampling. As the focus of this study was development of novel wet-vacuum cleaner-based sampling
approaches, the wipe-based method was not further improved for TEP recovery. The wipe sampling method
may improve for TEP recovery when considering a different type of wipe and/or wetting solvent.
5.4 Operational-scale Testing of Wet-Vacuum Sampling Efficiency - Phase III;
Medium-size Area
The operational-scale research addressed multiple aspects of discrete wet-vacuum sampling
applications: the overall sampling effectiveness (Section 5.4.1) at the default surface concentration and at
lower surface concentrations (Section 5.4.2); the direct comparison to the surface concentration via
extraction of applied amounts (Section 5.4.3); and the wet-vacuum methods comparison to conventional
wipe-based technique (Section 5.4.4)
5.4.1 Medium-size area sampling at default surface concentration
Medium-sized area (929 cm2) sampling focused on validation of the wet-vacuum method for single-
location sampling. The sampling efficiency achieved for nonporous reference material (stainless steel) at
100% chemical surface loading (equivalent to approximately 26-28 mg/coupon or 0.28-0.30 g/m2) was
considered the method performance baseline, compared to the wet-vacuum sampling effectiveness for one
other material, several chemical challenge levels, and to the performance of the single-area wipe sampling
method. The results forthe medium-sized area sampling of a nonporous (stainless steel) and a more
permeable, yet nonporous, flooring material (vinyl) are provided in Tables 5-2 and 5-3.
Table 5-2. Test Results of Medium-Sized Area Wet-Vacuum Sampling of 2-CEPS from a Nonporous
and Permeable Material at the Default Surface Concentration
2-CEPS: Medium-Size Area Wet-Vacuum Sampling



Chemical Recovered
Wetting Liquid
Recovered
Proc.
Sampling
Material
Wetting
Agent
Surface
Loading*
Mean
±SD

Mean
SD

Blank
Efficiency



(mg)
%RSD
(%)
%RSD
(mg)
%
SD

IPA
100%
21.8
1.5
7.3%
71
3.7
5.2

-------
Table 5-3. Test Results of Medium-Sized Area Wet-Vacuum Sampling of TEP from Nonporous and a
Permeable Material
TEP: Medium-Size Area Wet-Vacuum Sampling
Material
Wetting
Agent
Surface
Loading*
Chemical Recovered
Mean ±SD
Wetting Liquid
Recovered
Mean ±SD
Proc.
Blank
Sampling
Efficiency



(mg)
/oKoU
(%)
/oKoU
(mg)
%
SD

IPA
100%
18.7
0.93
4.7%
72
1.7
2.4%

-------
respectively. However, the final sample volume would also increase from 150 mL to 350 mL, i.e., by over
100%. Such an increase would worsen the analytical method detection limits.
5.4.2 Medium-size area sampling at low (below default) surface concentration
The results for the medium-sized area sampling at lower surface loadings (1 to 20% of default
surface concentration) are provided in Tables 5-4 (2-CEPS) and 5-5 (TEP). These tests were conducted
using IPA and one water based wetting agent. Results from the default surface concentration tests (Tables
5-2 and 5-3) indicated that water without surfactant is less efficient than either of the two water-surfactant
wetting agents. The Dawn Ultra® - water wetting agent was selected over the SuperSoap® - water wetting
agent considering its better performance for sampling of 2-CEPS.
Table 5-4. Test Results of Medium-Sized Area Wet-Vacuum Sampling of 2-CEPS at Low
Concentrations
2-CEPS: Medium-Size Area Wet-Vacuum Sampling - Low surface concentration threshold



Chemical Recovered
Wetting Liquid
Recovered
Proc.
Sampling
Material
Wetting
Agent
Surface
Loading*
Mean
±SD

Mean
±SD

Blank
Efficiency



(mg)
%RSD
(%)
%RSD
(mg)
%
SD
Stainless steel
IPA
20%
3.5
0.07
2.1%
72
2.4
3.4%

-------
Table 5-5. Test Results of Medium-Sized Area Wet-Vacuum Sampling of TEP at Low Concentrations
TEP: Medium-Size Area Wet-Vacuum Sampling - Low surface concentration threshold



Chemical Recovered
Wetting Liquid
Recovered
Proc.
Sampling
Material
Wetting
Agent
Surface
Loading*
Mean
±SD

Mean
±SD

Blank
Efficiency



(mg)
%RSD
(%)
%RSD
(mg)
%
SD
Stainless
steel
IPA
20%
0.66
0.36
53%
74
2.8
3.8%

-------
5.4.3 Surface concentration verification by direct extraction
Additional supplemental tests were conducted to confirm the surface concentration via direct
extraction, allowing for verification of whetherthe observed losses in wet-vacuum sampling efficiencies at
lower surface concentrations (Section 5.4.2) could be attributed solely to the loss in sampling efficiency of
the wet-vacuum sampling method or to an inherent lower recovery of the targeted chemical when spiked
onto a surface in a diluted solution due to accelerated evaporation of the chemical in the presence of a
volatile solvent or a combination of both processes. Table 5-6 and 5-7 provide the data on the amount
recovered of 2-CEPS and TEP, respectively, by direct extraction in the identified extraction agent as
described in Section 3.9.1 and the associated control spike recovery efficiency.
Table 5-6. Test Results for Direct Extraction of 2-CEPS from Stainless Steel
2-CEPS: Medium-Size Area Direct Extraction Sampling


Surface
Chemical Recovered
Extraction Efficiency
Control
Spike
Recover
ed
Control Spike
Material
Extraction
Agent
Loading
(mg)
[% of
default]
Mean
±SD
Mean
SD
%RSD
Recovery
Efficiency


(mg)
(%)

(mg)
%
SD

Hexane
28.2, [100]
33.1
0.5
117
1.7
1.5



Stainless steel
Hexane
5.64, [20]
4.6
N/A
82
N/A
N/A
0.90
94
N/A
Hexane
2.82, [10]
1.9
N/A
66
N/A
N/A
0.45
94
N/A

Hexane
0.56, [2]
0.002(J)
N/A
0.4
N/A
N/A
0.09
90
N/A
WAT - Dl water; DUW - Dawn Ultra®-water; SSW -
applicable (single test)
SuperSoap®-water; J - estimated value, data reported was below lowest point of the calibration curve. N/A: Not
Table 5-7. Test Results for Direct Extraction of TEP from Stainless steel
TEP: Medium-Size Area Direct Extraction Sampling


Surface
Chemical Recovered
Extraction Efficiency
Control
Spike
Recover
ed
Control Spike
Material
Extraction
Agent
Loading
(mg)
[% of
default]
Mean
±SD
Mean
SD
%RSD
Recovery
Efficiency


(mg)
(%)

(mg)
%
SD

Hexane
25.7, [100]
24.3
0.4
94
1.5
1.6



Stainless steel
Hexane
5.15, [20]
0.08036(J>
N/A
1.6
N/A
N/A
0.94
112
N/A
Hexane
2.57, [10]
0.00385(J)
N/A
0.16
N/A
N/A
0.50
117
N/A

Hexane
0.51,[2]
0.00095(J)
N/A
0.19
N/A
N/A
0.06
72
N/A
WAT - Dl water; DUW - Dawn Ultra®-water; SSW -
applicable (single test)
SuperSoap®-water; J - estimated value, data reported was below lowest point of the calibration curve. N/A: Not
Control spike recoveries for both 2-CEPS and TEP were high (90-94% for 2-CEPS and 72-117%
for TEP), indicating that extraction of these two chemicals from the hexane solvent is feasible and that
prepared spiked diluted solutions are correct in absolute concentration. However, extraction efficiencies of
these two diluted chemicals when applied to stainless-steel surfaces followed by a short contact time of 30
58

-------
min were strongly dependent on the solution concentration, with the lowest efficiency noted for the lowest
applied concentration. This dependence on concentration can be caused by a much poorer extraction
efficiency for these chemicals from a surface at lower concentrations and/or by a lower delivery of the
chemical due to an enhanced evaporation from surfaces of less volatile chemicals in the presence in a more
volatile solvent. Without further research, neither of the phenomena can be verified. Nevertheless, the
recovered amount is negatively biased, which will translate into lower recoveries when other sampling
methods such as wipe sampling or wet-vacuum sampling are considered. The observed poorer wet-vacuum
efficiency when sampling lower surface concentrations may be linked to this negative bias and would not
reflect an inherently poor quality of the proposed sampling method.
5.4.4 Comparison of wet-vacuum sampling to wipe-based sampling methods
As discussed previously, surface wipe sampling is one of the most common techniques used for
assessing environmental surface contamination with chemicals. The wipe sampling method used in this
study was a modification to an existing EPA reference method [9] and included selected modifications from
a large wipe-sampling evaluation study performed by EPA NHSRC [3], The method was compared to
efficiency of the single coupon IPA-based wet-vacuum sampling of a smooth nonporous material (stainless
steel), which was considered the performance baseline for the new method developed in the current study.
Both sampling methods were deployed against the same chemical surface loadings (280-300 mg per
square meter, or 100% default surface concentration target), but were deployed at different surface areas,
as described in detail in Section 4.1. The sampling efficiency results for medium-size area wipe sampling
against the results of the wet-vacuum sampling using IPA are summarized below in Figure 5-5.
2-CEPS: Comparison of sampling efficacy
Wipe
Wet-Vacuum
0% 20% 40% 60% 80% 100%
TEP: Comparison of sampling efficacy
Wipe
Wet-Vacuum
0% 20% 40% 60% 80% 100%
Figure 5-5. Comparison of sampling efficiency of SA wet-vacuum and wipe-based method for 2-CEPS (a) and
TEP (b); WS - wipe sampling, WV - wet-vacuum sampling
The wipe sampling method was outperformed by the wet-vacuum sampling method, offering a
statistically significant improvement in sampling efficiency (p=0.00067 and p=0.000012 for 2-CEPS and TEP
tests, respectively). As mentioned previously, the wipe sampling method was adopted from CWA sampling
methods and, therefore, had a limited applicability for sampling of TEP from surfaces. At the same time, the
IPA-based wet-vacuuming method was deployed in an identical (nonchemical-specific) configuration,
suggesting a much broader applicability to different classes of chemical compounds without chemical-
specific optimization requirements.
59

-------
5.5 Operational-scale Testing of Wet-Vacuum Sampling Efficiency - Phase III;
Large-size (Composite) Area
Due to relatively low recovery of target chemicals at low surface concentration as described in
Section 5.4.2, the large area evaluation was performed only at the 100% default contamination level (26-28
mg/coupon) on the nonporous stainless-steel material. The theoretical sample volume for five coupons was
expected to be approximately double the total sample liquid volume collected during a single coupon
sampling, namely, 5 x 50 mL of wetting solvent + 100-mL rinse (equals 350 mL) versus 1 x 50 mL of wetting
solvent and 100-mL rinse (equals 150 mL). Wetting liquid recoveries and chemical specific recoveries are
provided in Table 5-8.
Table 5-8. Test Results of Large-Sized Area Wet-Vacuum Sampling of 2-CEPS and TEP from
Stainless-steel Material (n=3)
Stainless
steel
2-CEPS
TEP
Large Size Area Wet-Vacuum Sampling
High surface concentration threshold
Chemical	Wetting Liquid
~ . . Wetting Surface
Chemical . . . ..
Agent Loading
Chemical
Recovered
Mean ±SD
Recovered
Mean SD
I PA
I PA
100%
100%
95.7
99.8
1.1
1.0
1.1%
1.0%
47
52
3.3
2.5
7.1%
4.7%
Proc. Sampling
Blank Efficiency
(mg) % SD

-------
5.6 Operational time comparisons
As part of this project, it was important to assess not only the test method feasibility for hazard
mapping and identification of various levels of contamination zones but also the time and cost-effectiveness
of proposed wet-vacuum approaches. The comparative analysis of the total operational time (TOT)
performed for each method included all major operational steps of the complete sampling procedure,
including the preparation for sampling step (consisting of preparation of sampling kits for wipe sampling
methodology, and preparation of cleaners and wetting agents for wet-vacuum based methodologies). The
TOT estimates were normalized for a surface area of 929 cm2, a minimum area evaluated for sampling
efficiency using wet-vacuum sampling methodologies. Collection of 10 SA wipe samples would be
necessary and the MAC method would allow for one composite sample characterization of five medium-size
(929 cm2) areas. Results of TOT comparison between wipe and wet-vacuum-based sampling methods are
shown in Figure 5-6 and suggest a large TOT reduction for the wet-vacuum methodology versus the wipe-
based methodology. A noticeable 8- to 10-fold reduction in TOT observed for IPA-based wet-vacuum
methodologies indicated that the same-solvent sampling and analysis should be considered as a critical
factor for improving analytical turnaround times, followed by sample-compositing during sampling.
Comparision of total operational time:
wet vaccum and wipe-based sampling methods




SAWS







SA AQUEOUS WV




MAC IPAWV
SA IPAWV
¦ I
¦I
1



0	50	100	150	200
Time (min)
¦ Preparation	Sampling ¦ Extraction and preparation for analysis
Figure 5-6. Comparison of total operational time for wet-vacuum and wipe-based sampling methods; WS - wipe
sampling; WV- wet-vacuum sampling; SA - single area; MAC - multiarea composite.
61

-------
5.7 Supplementary Characterization Tests
5.7.1 Bissell wet-vacuum airflow
The wet-vacuum flow rate of the Bissell wet-vacuum system that was tested was measured by
installing a reference flow meter in the vacuum hose between the operational [sampling]-position nozzle and
the body of the wet-vacuum (details in Section 3.10.2.1). The results of the operational flow rate
measurements - collected at approximately two-minute intervals throughout the duration of continuous, 20-
min long, wet-vacuum operation - are found in Figure 5-7.
Flow rate (slpm)
CO CO CO CO CO CO CO
co cn ct> co cd
o o o o o o o
_ooooooo
o
Operational flow rate of Bissell wet vacuum
during simulated sampling














¦
¦ ¦
¦ ¦
¦ ¦
¦ ¦
¦






















30 0:04 0:08 0:12 0:17 0:21 0:25
Elapsed time (min)
Figure 5-7. Airflow monitoring of Bissell unit simulated large-area sampling; flow rate is reported at US EPA
standard ambient sampling conditions of 101.325 kPa and 25 °C; slpm - standard liters per minute.
The average flow rate was recorded at 366.3 standard liters per minute (slpm) with an SD of 1.6
slpm, with a slight (<2%) flow rate drop observed throughout the 20-min long operation, which suggests that
robust and stable wet-vacuum airflow conditions were attainable for both SA and MAC methods.
5.7.2 Wet-vacuum temperature profile
During preliminary testing, the Bissell exhaust port, which is located near the vacuum motor, was
identified as a possible "hot spot" during large-area sampling. The temperature of the Bissell unit exhaust
port was thermally imaged throughout the duration of a simulated sampling test (with and without sampling
of surface). Figure 5-8 shows the heating of the wet-vacuum system during the surface sampling with water
as the wetting solvent. At the start of the test, the exhaust port temperature was approximately 26 °C (Figure
5-8a). After one minute of vacuum sampling, the temperature had risen to 34.9 °C (Figure 5-8b), and after
two or three minutes the temperature was above 41 °C (Figure 5-8c and 5-8d, respectively). At the end of
the sampling event, at approximately four minutes, the vacuum exhaust port temperature had reached 47.8
°C (Figure 5-8e); 15 minutes later, the vacuum port had cooled to 26.6 °C or close to the starting
temperature). Internal components of the vacuum (pump and motor) were still at approximately 30 °C
following the 15-minute cooling period (thermal imaging data not shown in Figure 5-8).
62

-------
47.8 °c
Figure 5-8. Temperature monitoring of Bissell exhaust during simulated sampling. Sampling performed under
the chemical hood, (a) start of vacuum sampling (t = 0 min); (b) after 1 min of vacuum sampling (t = 1 min); (c)
after 2 min of sampling (t = 2 min); (d) after 3 min of sampling (t = 3 min); (e) at the end of sampling (t = 3 min 24
s); f, 15 min after sampling ended (t = 19 min 24 s).
The simulated sampling described above was performed using only four medium-size coupons, the
number of coupons that can be accommodated in the chemical hood during operational-scale testing and
indicated that multiarea sampling quickly increases the temperature of the vacuum cleaner exhaust port.
From an operational safety perspective, the most critical consideration was to determine if the wetting liquid
holding tank or liquid waste collection tank temperatures were rising substantially during sampling,
especially while working with a flammable solvent like IPA. The temperature of major vacuum cleaner
components was monitored continuously for 20 min, with and without wetting solvent (water). Thermal
images of all general vacuum components during this test indicated the wetting liquid holding tank and liquid
waste collection tank were cooler than the exhaust port and the central vacuum system (Figure 5-9a).
Remarkably, the liquid waste collection tank temperature felt below 20 °C during 20-min long sampling
(Figure 5-9b). As the effect of the evaporative cooling was going to be even greater during IPA testing, the
Bissell unit was deemed usable for IPA-based SA and MAC wet-vacuum sampling.
63

-------
~18.8 °c

¦ 31.6
L
A '
ill

m.
b.

¦
CO J
VI


Figure 5-9. Temperature in various zones of Bissell vacuum cleaner after 20 min sampling using water, (a)
Central system temperature near motor; (b) Liquid waste collection tank temperature.
64

-------
6.0 Quality Assurance and Quality Control
6.1 Test Equipment Calibration
All equipment was verified as calibrated at the time of use. Instruments were calibrated at the
frequency shown in Table 6-1. In case of any deficiencies, instruments were adjusted to meet calibration
tolerances and/or recalibrated prior to testing. In the case of the GC/MS instrument, any initial calibration
deficiencies were noted. The GC/MS instrument was recalibrated prior to analysis. If the tolerances for
continuous calibration were not met, the GC/MS instrument was recalibrated and affected samples were
reanalyzed.
Table 6-1. Instrument Calibration Frequency
Equipment
Calibration/Certification
Expected Tolerance
Results
Thermometer
Compare to independent NIST thermometer (a thermometer that is
recertified annually by either NIST or an ISO-17025 facility) value once per
quarter.
±1 °C
100%
Stopwatch
Compare to official U.S. time at time.gov every 30 days.
± 1 min/30 days
100%
Micropipettes
Certified as calibrated at time of use; recalibrated by gravimetric evaluation
of performance to manufacturer's specifications every year.
±5%
100%
Scale
Certified as calibrated at time of use; calibration verified yearly by the
AEMD Metrology Laboratory.
±1 g
100%
Graduated cylinder
Certified by manufacturer at the time of use.
± 1 mL
100%
Solvent dispenser
Certified by manufacturer at the time of use; rechecked volume delivered
using graduated cylinder prior to use.
± 1 mL
100%
Gas
chromatography/mass
spectrometer
5-point calibration prior to analysis; continuous calibration prior to each
analytical run; recalibrate when continuous calibration fails acceptance
criteria and/or after system maintenance; details in Section 4.4.
± 20% at mid-point
100%
Gas
chromatography/flame
ionization detector.
6-point calibration prior to analysis; continuous calibration prior to each
analytical run; recalibrate when continuous calibration fails acceptance
criteria and/or after system maintenance; details in Section 4.4.
± 20% at mid-point
100%
NIST - National Institute of Standards and Technology; ISO - International Organization for Standardization;
6.2 Data Quality Results for Critical Measurements
The following measurements were deemed critical to accomplishing project objectives:
Surface concentration of target chemicals as determined by instrumental analysis
Chemical concentration in the SL
Chemical concentration in the SL or wipe extracts
Contact time
Lapse time
Volume of wetting agent applied onto coupon surfaces
The data quality indicators (DQIs) for test measurements are provided in Table 6-2. The limited
number of results/tests that were not within acceptance criteria (as determined in the project-specific quality
assurance project plan) were not indicative of any systematic error introduced into the experimental results
and do not change the general findings of this study.
65

-------
Table 6-2. Acceptance Criteria for Critical Measurements and Corresponding Test Results
Critical Measurement
Target Value and Acceptance
Criteria
Results
Contact/weathering time
30 min ± 1 min
All contact times within 30 min ± 1 min from spiking
Lapse time
10min±30s
All lapse times within 10 min ± 30 s from application of wetting
solvent to sampling
Delivery of target surface concentration
of chemical*
80-120% of target
All mean spike controls for decontamination tests were within
acceptance criteria with coefficients of variation <30% between
tests (all target chemicals); results are in Table 4-6.
Recovery of chemical from surface
samples^
<30% coefficient of variation for
identical test set
Selected low-concentration tests had >30% coefficient of
variation for surface samples resulting from identical test set;
test-specific results are in Tables 5-2 and 5-3.
Procedural blank
<5% of the analyte amount
recovered from the positive control.
All procedural blank samples within acceptance criteria; all
reported 
-------
7.0. Summary
A novel wet-vacuum-based chemical sampling methodology was evaluated for laboratory-scale and
operational scale application to sample four types of building materials (stainless steel, vinyl flooring,
laminate flooring, and plywood). A first step of method development involved the evaluation of a wetting
agent suitable for wet-vacuum sampling of chemicals with varied solubility in water, followed by
measurement of the amount of organic solvent or water-based wetting agent to be dispensed to a surface,
and selection of the wetting agent surface residence time (or lapse time). After wetting solvent and lapse
time evaluation, IPA- and water-based sampling methods were evaluated for sampling of target chemicals
using two types of commercially available cleaners. The cleaner that offered better efficiency of solvent
collection (Bissell Multi-Purpose Cleaner) and better overall chemical sampling efficiency was further
optimized for improved recovery of IPA and chemical recovery. The optimized method was evaluated in
operation-scale testing for multivariate (chemical and surface type, surface contamination level, and wetting
agent type) sampling of medium-size (<1000 cm2) and large-size (approximately 5000 cm2) areas and
compared to existing wipe-based sampling methods and/or modifications thereof.
The main findings of this study are:
•	Wet-vacuum-based methods, utilizing a commercially available cleaner and IPA wetting
solvent for sample collection, can be considered for sampling of various classes of chemicals
with varied solubility in water, but the methodology is prone to the evaporation-related losses
of IPA from the surface and - when larger-volume composite samples are collected - also in
the liquid waste collection tank.
•	Wet-vacuum method performance is lower for collection of target chemicals from semiporous
materials (wood, vinyl), and for sampling of surfaces contaminated with lower (jjg/m2) levels
of chemicals.
•	The efficiency of aqueous wetting agent-based wet-vacuum sampling is affected by the
solubility of the chemical in water. The addition of surfactant improves recovery of selected
water-insoluble chemicals but generally does not improve the sampling efficiency when lower
surface concentrations of chemicals are targeted.
•	The newly developed methodology used with IPA as wetting solvent offers eight- to tenfold
improvement in turnaround time needed to collect and prepare surface sample for analysis.
The aqueous wetting agent methods - that include the step of liquid-liquid extraction of
sampling liquid- offers an approximately twofold reduction of the TOT, as compared to wipe-
based methodology deployed in a discrete-area configuration.
•	Both variations of wet-vacuum methods (single area and multiarea composite) used for
collection of mg/m2 surface concentrations of chemicals, offer a similar (or better) sampling
efficiency when compared to wipe-based methods.
•	The vacuum cleaner components were confirmed to contain the chemical after use
suggesting the wet-vacuum units should not be considered for reuse and should be handled
as contaminated waste.
The novel sampling method developed in this study can be considered as a promising analytical tool for
the medium- to large-area chemical contamination sampling. However, a careful selection of the wetting
solvent that is appropriate for target chemical-surface characteristics combinations is needed for obtaining
67

-------
reliable results. Further testing should be considered to find the maximum area the method can be deployed
for while using composite-sampling techniques and water-based wetting solvents). Such optimization would
improve the applicability of this method to field sampling of CWAs and other toxic chemicals of interest.
Future work to improve a wet-vacuum sampling method should consider use of more inert materials
which may eliminate the use of a pre-sampling conditioning step and may allow for reuse of the same unit
resulting in a lower sampling cost and less waste generated.
68

-------
8.0. References
1.	U.S. EPA. (2012) Selected Analytical Methods for Environmental Remediation and Recovery (SAM)
-2012. EPA/600/R-12/555, July 2012. Cincinnati, OH: U.S. Environmental Protection Agency.
2.	U.S. EPA. (2007) A Literature Review of Wipe Sampling Methods for Chemical Warfare Agents and
Toxic Industrial Chemicals. EPA/600/R-11/079, January 2007. Cincinnati, OH: U.S. Environmental
Protection Agency.
3.	U.S. EPA. (2016) Evaluation of Chemical Warfare Agent Wipe Sampling Collection Efficiencies on
Porous, Permeable, or Uneven Surfaces, EPA/600/R-16/189, September 2016. Office of Research
and Development Homeland Security Research Program, US Environmental Protection Agency,
4.	U.S. EPA. (2016) Guidance on Choosing a Sampling Design for Environmental Data Collection for
Use in Developing a Quality Assurance Project Plan. EPA/240/R-02/005, December 2002 Office of
Environmental Information. U.S. Environmental Protection Agency. Washington, DC
5.	Baldenebro-Lopez FJ, Gomez-Esparza CD, Corral-Higuera R, Arredondo-Rea SP, Pellegrini-
Cervantes MJ, Ledezma-Sillas JE, Martinez-Sanchez R, Herrera-Ramirez JM. (2015) Influence of
Size on the Microstructure and Mechanical Properties of an AISI 304L Stainless steel—A
Comparison between Bulk and Fibers. Ivey D, ed. Materials. 2015;8(2):451-461.
6.	Bartelt-Hunt SL, Knappe DRU, Barlaz MA. (2008) A Review of Chemical Warfare Agent Simulants
forthe Study of Environmental Behavior, Crit. Rev. Environ. Sci. Technol. 38(2), 112-136.
7.	Oudejans L, Wyrzykowska-Ceradini B, Williams C, Tabor D, Martinez J. (2013) Impact of
Environmental Conditions on the Enzymatic Decontamination of a Material Surface Contaminated
with Chemical Warfare Agent Simulants. Ind. Eng. Chem. Res. 52 (30), 10072-10079.
8.	Oudejans L, O'Kelly J, Evans AH, Wyrzykowska-Ceradini B, Touati A, Tabor D, Gibb Snyder E.
(2016) Decontamination of Personal Protective Equipment and Related Materials Contaminated
with Toxic Industrial Chemicals and Chemical Warfare Agent Surrogates. J. Environ. Chem.
Engineering 4(3), 2745-2753.
9.	U.S. EPA. (2007) Method 3572 (SW-846). Extraction of Chemical Agents from Wipe Samples
Using Micro-extraction, Rev 1, July 2014, U.S. Environmental Protection Agency.
10.	NIOSH (1994) Method 2546. Cresol (all Isomers) and phenol. NIOSH Manual of Analytical Methods
(NMAM), Fourth Edition, 8/15/94
11.	U.S. EPA. (1978) Method 420.1: Phenolics (Spectrophotometric, Manual 4-AAP With Distillation).
U.S. Environmental Protection Agency. Washington, DC
69

-------
12.	NIOSH (2005) Method 2546. Nitroaromatic Compounds. NIOSH Manual of Analytical Methods
(NMAM), Fourth Edition, 8/15/94
13.	NIOSH (2005) Method 5034. Tributyl phosphate. NIOSH Manual of Analytical Methods (NMAM),
Fourth Edition, 8/15/94
14.	NIOSH (2005) Method 5038. Triphenyl phosphate. NIOSH Manual of Analytical Methods (NMAM),
Fourth Edition, 8/15/94
15.	U.S. EPA. (1999) USEPA Contract Laboratory Program National Functional Guidelines for Organic
Data Review. OSWER EPA540/R-99/008, October 1999. Washington, DC: U.S. Environmental
Protection Agency
16.	U.S. EPA. (2007) Method 3571 (SW-846): Extraction of Solid and Aqueous Samples for Chemical
Agents. Revision 0. Washington, DC: U.S. Environmental Protection Agency
17.	U.S. EPA (2017) Field Application of Emerging Composite Sampling Methods. DC, EPA/600/R-
17/212, 2017. Washington, D.C. U.S. Environmental Protection Agency,
70

-------
Appendix A: Supporting Information
A-1 Wet-Vacuum Sampling Procedure
This procedure describes vacuum sampling using the Bissell SpotClean Portable Carpet Cleaner,
model 5207A device. This procedure was optimized for sampling of areas from 929 cm2 to 4,645 cm2.
Preparation of the vacuum
To assemble the vacuum cleaner, slide the flex hose clip into the opening on the front of the cleaner
until you hear it click into place. Then snap the hose grip bracket into the back of machine on the clean tank
side. Wrap the flex hose around the unit and snap the hose grip into the bracket. Attach the cord wrap by
snapping it into place on the dirty water tank (referred to as dirty liquid tank below) side of the machine.
Wrap the power cord around the cord wrap. The machine is now assembled.
Sampling
Step 1
| Connect the cleaning tool/nozzle and peripheral equipment per manufacturer instructions
and secure. Tool easily snaps onto hose. For this study, the 3-in. Tough Stain Brush Tool was used.
Step 2
Remove tank from tool side of unit. Pull on the black tab to remove rubber stopper and
reveal tank opening.
Step 3
| Tank is marked with lines for a large stain or a small stain. For this project, fill the tank with
IPA to the small stain fill line.
Step 4
Step 5
Replace rubber stopper, then replace the tank on the machine.
| Twist the quick-release cord-wrap to release the cord. The vacuum is now ready to clean
with the spot-clean option.
Step 6
Prime the sprayer by compressing the trigger and discharging the IPA into a waste beaker
for approximately 5 seconds. Weigh and record the wetting solvent tank mass. Weigh the liquid waste
collection tank prior to the start of vacuum sampling.
Step 7
If using IPA wetting agent for sampling, condition the vacuum by vacuuming 50 mL of IPA
from a 1-L beaker. No conditioning is needed for any aqueous wetting agents (e.g., water, water-surfactant
solutions). Immediately proceed to wetting the surface with the wetting agent.
Step 8
| To wet the sampling area of 929 cm2, spray the coupon surface for 8 seconds with the wet-
vacuum sprayer. Hold the spray nozzle approximately 6 in. above the coupon surface. Press the spray trigger
and spray the test surface of the coupon for eight seconds in an S-pattern covering the test area. Let IPA dwell
on the coupon surface for 10 seconds. Figure A-1 illustrates the locations of the spray nozzle and spray
trigger.
71

-------
Figure A-1 Spray nozzle (A) and spray trigger (B) locations
Vacuum each TC starting in the top left corner the coupon. Position the nozzle at a 45-
degree angle. Pull the nozzle toward the right side of the coupon. When the edge of the coupon is reached,
start the next vacuuming pass on the left side overlapping approximately 50% of the previous pass.
Continue this vacuuming pattern until the bottom of the coupon is reached. Next, vacuum the coupon
starting in the top left corner. Position the nozzle at a 45-degree angle. Pull the nozzle toward the bottom
edge of the coupon. When the bottom edge of the coupon is reached, start the next vacuuming pass on the
top side overlapping approximately 50% of the previous pass. Continue this vacuuming pattern until the right
edge of the coupon is reached. For composite-area sampling, repeat the procedure for all coupons.
liquid tank to a clean Nalgene or equivalent archival bottle and store the sample at 4 ± 2 °C until preparation
for analysis.
After surface sampling is completed, rinse the wet-vacuum with 100 mL of IPA from a
clean beaker.
Weigh the liquid waste collection tank and then transfer the entire contents of the dirty
72

-------
A-2 Parameters and Conditions for Instrumental analysis
Table A-1. GC/FID Analyses of Nitrobenzene with Modified NIOSH* 2005 (EMSL Analytical, Inc.)
Parameter
Description/Conditions
Instrument
Agilent 7890A with dual FID (Agilent Technologies, Santa Clara, CA)
Autosampler
Agilent 7693 automatic sampler (Agilent Technologies, Santa Clara, CA)
Column
RTX 624sil Restek #13870 S/N 1208834 30 m, x 0.32 mm ID, 1.8 pm df (Restek Corporation,
Bellefonte, PA)
GC column program
40 °C initial temperature, hold 0 min, 10 °C/min to 190 °C, hold 2 min
Carrier gas flow rate
3.3 mL/min
Injection volume/type
1.0 [jL/splitless
Inlet temperature
225 °C
FID heater temperature
250 °C
Carrier
Helium
df: film thickness; *National Institute for Occupational Safety and Health
Table A-2. GC/FID Analyses of Phenol Modified NIOSH Method 2546 (EMSL Analytical, Inc.)
Parameter
Description/Conditions
Instrument
Agilent 7980A gas chromatograph equipped with FID (Agilent Technologies, Santa Clara, CA)
Autosampler
7693 Automatic Sampler (Agilent Technologies, Santa Clara, CA)
Column
RTX™-VGC capillary column, 60 m x 0.530 |im ID, 0.3 |jm df; part no. 19488 (Restek Corporation,
Bellefonte, PA)
GC column program
70 °C initial temperature, hold 0 min, 15 °C/min to 250 °C, hold 2 min
Carrier gas flow rate
11.258 mL/min
Injection volume/type
1.0 [jL/splitless
Inlet temperature
225 °C
FID heater temperature
250 °C
Hydrogen flow rate
40 mL/min
Air flow rate
400 mL/min
df: film thickness
73

-------
Table A-3. GC/FID Analyses of TEP Using Modified NIOSH 5034 (EMSL Analytical, Inc.)
Parameter
Description/Conditions
Instrument
Agilent 7980 Gas Chromatograph equipped with FID (Agilent Technologies, Santa Clara, CA)
Autosampler
7693 Automatic Sampler (Agilent Technologies, Santa Clara, CA)
Column
RTX™-VGC Capillary Columns, 60 m x 0.530 |im ID, 0.3 |jm df; part no. 19488 (Restek Corporation,
Bellefonte, PA)
GC column program
70 °C initial temperature, hold 0 min, 15 °C/min to 250 °C, hold 2 min
Carrier gas flow rate
11.258 mL/min
Injection volume/type
1.0 [jL/splitless
Inlet temperature
225 °C
FID heater temperature
225 °C
Hydrogen flow rate
40 mL/min
Air flow rate
400 mL/min
df: film thickness
Table A-4. GC/FID Analyses of TEP Modified NIOSH 5038 (EMSL Analytical, Inc.)
Parameter
Description/Conditions
Instrument
Agilent 7890A with dual FID (Agilent Technologies, Santa Clara, CA)
Autosampler
Agilent 7693 Automatic Sampler (Agilent Technologies, Santa Clara, CA)
Column
RTX 624 sil Restek #13870 S/N 1208834 30 m x 0.32 mm ID, 1.8 pm df (Restek Corporation,
Bellefonte, PA)
GC column program
40 °C initial temperature, hold 0 min, 10 °C/min to 190 °C, hold 2 min
Carrier gas flow rate
3.3 mL/min
Injection volume/type
1.0 [jL/splitless
Inlet temperature
225 °C
FID heater temperature
250 °C
Carrier
Helium
df: film thickness
74

-------
Table A-5. GC/MS Analyses of TEP and 2-CEPS Samples in IPA (EPA OSL)
Parameter
Description/Conditions
Instrument
Thermo Trace 1300 GC ISO™ Mass Spectrometer (Thermo Fisher Scientific, Inc., Waltham, MA)
Autosampler
AS/A11310 Autosampler (Thermo Fisher Scientific, Inc., Waltham, MA)
Column
DB-5,20 m x 0.25 mm ID, 0.25 |jm df (Agilent, Santa Clara, CA)
GC column program
60 °C initial temperature, hold 0 min, 8 °C/min to 260 °C, hold 8 min
Carrier gas flow rate
1.3 mL/min
Injection volume/type
1.0 [jL/splitless
Inlet temperature
150 °C
MS source temperature
155 °C
MS transfer line
150 °C
df: film thickness
Table A-6. GC/MS Analyses of TEP and 2-CEPS in Hexane (EPA OSL)
Parameter
Description/Conditions
Instrument
Thermo Trace 1300 Gas Chromatograph GC ISO™ Mass Spectrometer (Thermo Fisher Scientific, Inc.,
Waltham, MA)
Autosampler
AS/A11310 Autosampler (Thermo Fisher Scientific, Inc., Waltham, MA)
Column
DB-5,20 m x 0.25 mm ID, 0.25 |jm df (Agilent, Santa Clara, CA)
GC column program
60 °C initial temperature, hold 0 min, 8 °C/min to 260 °C, hold 8 min
Carrier gas flow rate
1.3 mL/min
Injection volume/type
1.0 [jL/splitless
Inlet temperature
250 °C
MS source temperature
250 °C
MS transfer line
250 °C
df: film thickness
75

-------
A-3 Sample-Specific Test Results
Table A-7. Liquid-Liquid Extraction Efficiency of Selected Chemicals from Wetting Agents
Chemical:
2-CEPS
TEP
Nitrobenzene
HT tested:
1 h
24 h
1 h
24 h
1 h
Amount spiked:
5.6 mg

2.4 mg
Sample recovery:
Mass
(mg)
Recovery
(%)
Mass
(mg)
Recovery
(%)
Mass
(mg)
Recovery
(%)
Mass
(mg)
Recovery
(%)
Mass
(mg)
Recovery
(%)
Extraction from SSDX-12®-water solution*
PB-1
ND
NA
ND
NA
ND
NA
ND
NA
Not tested
TL-1
4.82
86%
1.76
31%
4.98
97%
4.92
96%
TL-2
5.10
90%
1.66
29%
5.28
103%
5.04
98%
TL-3
5.10
90%
1.67
30%
5.25
102%
5.11
100%
Average
4.99
89%
1.70
30%
5.17
101%
5.02
98%
Std Dev
0.12
2.2%
0.045
0.79%
0.13
2.6%
0.081
1.6%
% RSD
2.4%
2.6%
2.6%
1.6%
Extraction from Dawn Ultra®-water solution*
PB-1
ND
NA
ND
NA
ND
NA
ND
NA
Not tested
TL-1
5.04
90%
3.06
54%
4.81
94%
5.21
101%
TL-2
5.35
95%
3.01
53%
4.77
93%
5.18
101%
TL-3
5.40
96%
3.15
56%
5.05
98%
5.16
100%
Average
5.27
93%
3.07
55%
4.87
95%
5.18
101%
Std Dev
0.16
2.8%
0.060
1.1%
0.12
2.4%
0.020
0.4%
% RSD
3.0%
1.9%
2.53%
0.38%
Extraction from water*
PB-1
ND
NA
ND
NA
ND
NA
ND
NA
ND
NA
TL-1
4.86
86%
2.48
44%
4.80
93%
4.98
97%
2.10
85%
TL-2
4.92
87%
2.78
49%
5.16
100%
5.03
98%
2.70
109%
TL-3
4.94
88%
2.80
50%
4.99
97%
5.05
98%
2.20
89%
Average
4.91
87%
2.69
48%
4.98
97%
5.02
98%
2.33
94%
Std Dev
0.034
0.61%
0.15
2.6%
0.15
2.8%
0.029
0.56%
0.32
13%
% RSD
0.70%
5.4%
2.9%
0.58%
14% |
Extraction from Tween® 20-water solution*
PB-1
Not tested
Not tested
NA
NA
TL-1
4.3
75%
TL-2
4.4
76%
TL-3
1.2(E)
21 %(E)
Average
4.4
76%
Std Dev
0.05
1%
% RSD
I 1% I
*1:1 (v:v) extraction with DCM or hexane, chemical target-wetting agent combination-specific additives are listed in Table 3-6; E- excluded from arithmetic
mean (average) calculations due to suspected reporting error identified during data validation; TL- test liquid sample (wetting liquid spiked with target
chemical); PB - procedural blank sample (not spiked wetting liquid that underwent the entire analytical procedure)
76

-------
Table A-8. Results of Direct Analysis of Phenol in Various Wetting Agents
Chemical:
Phenol
HT tested:
1 h
Sample recovery:
Mass
Recovery
Wetting Agent
(mg)
(%)
Water
2.0
94%
2%Tween-water
2.1
99%
IPA
2.1
99%
Average
2.1
97%
Std Dev
0.06
2.9%
% RSD
2.8%
Table A-9. Results of Direct Analysis of TEP in IPA
Chemical:
TEP
HT tested:
1 h
Sample recovery:
Recovery
(%)
Standard Deviation
(%)
Phase
Phase I, phase II
105.3
4.7
Phase III
96.1
7.14
Average recovery (all phases)
100.7%
Std Dev
4.6%
% RSD
4.6%
Table A-10. Results of Direct Analysis of 2-CEPS in IPA
Chemical:
2-CEPS
HT tested:
1 h
Sample recovery:
Recovery
(%)
Standard Deviation
(%)
Phase
Phase II
92.9
7.2
Average
92.9%
Std Dev
7.2%
% RSD
7.7%
77

-------
Table A-11. Phase I Operational Parameter Optimization - Selection of Wetting Agent for Sampling
of Phenol from Nonporous Reference Material


Phenol


Material Type
and Wetting Agent
Type
Surface
Loading
(mg)
Volume of Wetting
Solvent Collected
(mL)
Wetting Liquid
Recovered
(%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless Steel - Water
TC-1
26
37.8
87%
15.5
60%
TC-2
26
43.0
86%
15.9
62%
TC-3
26
40.4
87%
15.7
61%
TC-4
26
41.7
87%
16.2
63%
TC-5
26
42.3
88%
16.9
66%
PB
Not spiked
44.9
89%
ND
NA
Average
41.7
87%
16.0
62%
Std Dev
2.42
1.0%
0.55
2.3%
% RSD
5.8%
1.2%
3.4%
3.7%


Stainless Steel
- TweenO-water


TC-1
26
41.0
87%
17.6
69%
TC-2
26
41.1
89%
17.7
69%
TC-3
26
42.4
89%
16.5
64%
TC-4
26
42.0
88%
16.4
64%
TC-5
26
41.5
86%
17.8
69%
PB
Not-spiked
42.4
89%
ND
NA
Average
41.7
88%
17.2
67%
Std Dev
0.63
1.3%
0.69
2.7%
% RSD
1.5%
1.4%
4.0%
4.1%
Stainless Steel - IPA
TC-1
26
26.1
43%
16.5
64%
TC-2
26
27.8
56%
15.6
61%
TC-3
26
25.0
46%
17.0
66%
TC-4
26
30.4
63%
17.9
70%
TC-5
26
23.2
42%
15.8
61%
PB
not-spiked
35.3
58%
ND
NA
Average
28.0
51%
16.6
64%
Std Dev
4.35
8.8%
0.93
3.8%
% RSD
15.6%
17.1%
5.6%
5.9%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from
noncontaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable.
78

-------
Table A-12. Phase I Operational Parameters Optimization - Selection of Wetting Agent for Sampling
of Nitrobenzene from Nonporous Reference Material


Nitrobenzene


Material Type
and Wetting Agent
Type
Surface
Loading
(mg)
Volume of Wetting
Solvent Collected
(mL)
Wetting Liquid
Recovered
(%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless Steel - Water
TC-1
29
44.4
92%
4.0
14%
TC-2
29
46.8
91%
5.6
20%
TC-3
29
41.8
92%
4.1
14%
TC-4
29
45.4
93%
3.9
13%
TC-5
29
48.3
91%
4.8
17%
PB
Not spiked
44.4
91%
ND
NA
Average
45.2
92%
4.5
16%
Std Dev
2.24
0.8%
0.72
2.9%
% RSD
5.0%
0.9%
16.0%
18.5%


Stainless Steel
- TweenO-water


TC-1
29
35.0
88%
4.2
15%
TC-2
29
25.3
85%
2.8
10%
TC-3
29
33.8
89%
3.1
11%
TC-4
29
44.6
87%
4.9
17%
TC-5
29
37.3
88%
22.0 (S,E)
77% (S,E)
PB
Not spiked
32.2
84%
ND
NA
Average
34.7
87%
3.8
13%
Std Dev
6.33
1.9%
0.97
3.3%
% RSD
18.2%
2.2%
26.0%
24.9%
Stainless Steel - IPA
TC-1
29
28.2
51%
17.5
61%
TC-2
29
13.6 (L)
28% (L)
9.2(E)
32% (E)
TC-3
29
28.9
57%
18.8
65%
TC-4
29
21.6
46%
17.3
60%
TC-5
29
33.0
62%
20.5
71%
PB
Not spiked
26.2
50%
ND
NA
Average
27.6
53%
18.5
64%
Std Dev
4.16
6.3%
1.48
5.0%
% RSD
15.1%
11.8%
8.0%
7.8%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from
noncontaminated (procedural blank) stainless-steel coupon; S - Possible sample preparation error identified; L - low solvent collection efficiency observed;
E -value excluded from statistical calculations; ND - Not detected; NA- Not applicable.
79

-------
Table A-13. Phase I Operational Parameters Optimization - Selection of Wetting Agent for Sampling
of TEP from Nonporous Reference Material


TEP


Material Type
Surface
Volume of Wetting
Wetting Liquid
Mass of Chemical
Chemical
and Wetting Agent
Loading
Solvent Collected
Recovered
Collected
Recovery
Type
(mg)
(mL)
(%)
(mg)
(%)
Stainless Steel - Water
TC-1
26
45.4
89%
11.1
43%
TC-2
26
46.7
89%
11.1
43%
TC-3
26
48.6
92%
13.0
51%
TC-4
26
47.9
90%
13.1
51%
TC-5
26
52.9
93%
8.2
32%
PB
Not spiked
46.2
91%
ND
NA
Average
48.0
91%
11.3
44%
Std Dev
2.69
1.6%
1.99
7.8%
% RSD
5.6%
1.8%
17.6%
17.8%


Stainless Steel
- TweenO-water


TC-1
29
41.0
88%
12.4
48%
TC-2
29
44.3
90%
13.3
52%
TC-3
29
44.1
90%
13.5
53%
TC-4
29
43.7
88%
11.2
44%
TC-5
29
46.8
88%
14.5
56%
PB
Not spiked
45.3
88%
ND
NA
Average
44.2
89%
13.0
51%
Std Dev
1.92
1.0%
1.24
4.7%
% RSD
4.3%
1.2%
9.6%
9.2%
Stainless Steel - IPA
TC-1
26
30.4
53%
17.6
69%
TC-2
26
25.5
49%
16.1
63%
TC-3
26
27.8
51%
22.3
87%
TC-4
26
22.0
42%
11.0
43%
TC-5
26
38.3
76%
18.4
72%
PB
Not spiked
31.2
55%
ND
NA
Average
29.2
54%
17.1
67%
Std Dev
5.58
11.5%
4.10
16.0%
% RSD
19.1%
21.2%
24.0%
23.9%
ItC — SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non- I
pontaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA- Not applicable.


80

-------
Table A-14. Phase I Operational Parameters Optimization - Selection of the Wetting Agent Surface
Contact Time (Lapse Time) for IPA Sampling of Phenol from Nonporous Reference Material


Phenol


Material Type-
Wetting Agent -
Surface Lapse Time
Surface
Loading
(mg)
Volume of Wetting
Solvent Collected
(mL)
Wetting Liquid
Recovered
(%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless Steel - IPA
- LT = 1 s




TC-1
26
31.2
59%
17.8
69%
TC-2
26
31.4
60%
16.6
65%
TC-3
26
42.8
72%
18.0
70%
TC-4
26
36.8
70%
17.7
69%
TC-5
26
43.0
76%
16.8
65%
PB
Not spiked
33.4
65%
ND
NA
Average
36.4
67%
17.4
68%
Std Dev
5.40
6.8%
0.63
2.4%
% RSD
14.8%
10.2%
3.6%
3.6%


Stainless Steel
IPA - LT = 10 s


TC-1
26
34.8
68%
18.5
72%
TC-2
26
34.6
65%
19.7
77%
TC-3
26
32.9
59%
13.2
51%
TC-4
26
31.7
64%
17.1
67%
TC-5
26
42.9
73%
17.6
69%
PB
Not spiked
34.4
65%
ND
NA
Average
35.2
66%
17.2
67%
Std Dev
3.95
4.6%
2.45
9.8%
% RSD
11.2%
7.1%
14.3%
14.6%


Stainless Steel -
IPA-LT = 100 s


TC-1
26
26.1
43%
16.5
64%
TC-2
26
27.8
56%
15.6
61%
TC-3
26
25.0
46%
17.0
66%
TC-4
26
30.4
63%
17.9
70%
TC-5
26
23.2
42%
15.8
61%
PB
Not spiked
35.3
65%
ND
NA
Average
28.0
53%
16.6
64%
Std Dev
4.35
10.2%
0.93
3.8%
% RSD
15.6%
19.5%
5.6%
5.9%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from
noncontaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface
contact time).
81

-------
Table A-15. Phase I Operational Parameters Optimization - Selection of the Wetting Agent Surface
Contact Time (Lapse Time) for IPA Sampling of Nitrobenzene from Nonporous Reference Material
Material Type-
Wetting Agent -
Surface Lapse Time
Surface
Loading
(mg)
Nitrobenzene
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless Steel - IPA
- LT = 1 s

TC-1
29
17.1
34%
15.2
53%
TC-2
29
39.7
75%
27.8
97%
TC-3
29
30.1
63%
24.4
85%
TC-4
29
34.4
65%
26.2
91%
TC-5
29
36.0
70%
23.0
80%
PB
Not spiked
36.4
62%
ND
NA
Average
32.3
62%
23.3
81%
Std Dev
8.07
14.3%
4.89
17.0%
% RSD
25.0%
23.3%
21.0%
20.9%
Stainless Steel
IPA - LT = 10 s

TC-1
29
26.6
61%
29.2
102%
TC-2
29
34.1
66%
29.0
101%
TC-3
29
30.9
65%
25.9
90%
TC-4
29
37.8
74%
26.1
91%
TC-5
29
32.4
70%
27.5
96%
PB
Not spiked
41.5
77%
ND
NA
Average
33.9
69%
27.5
96%
Std Dev
5.24
6.0%
1.55
5.5%
% RSD
15.5%
8.7%
5.6%
5.8%
Stainless Steel -
IPA-LT = 100 s

TC-1
29
28.2
51%
17.5
61%
TC-2
29
13.6
28%
9.2
32%
TC-3
29
28.9
57%
18.8
65%
TC-4
29
21.6
46%
17.3
60%
TC-5
29
33.0
62%
20.5
71%
PB
Not spiked
26.2
50%
ND
NA
Average
25.3
49%
16.7
58%
Std Dev
6.81
11.7%
4.36
15.1%
% RSD
27.0%
23.9%
26.2%
26.0%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
82

-------
Table A-16. Phase I Operational Parameters Optimization - Selection of the Wetting Agent Surface
Contact Time (Lapse Time) for IPA Sampling of TEP from Nonporous Reference Material


TEP


Material Type-
Wetting Agent -
Surface Lapse Time
Surface
Loading
(mg)
Volume of Wetting
Solvent Collected
(mL)
Wetting Liquid
Recovered
(%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless Steel - IPA
- LT = 1 s




TC-1
26
36.7
72%
20.9
81%
TC-2
26
35.4
68%
22.0
86%
TC-3
26
39.0
71%
23.4
91%
TC-4
26
40.2
70%
22.9
89%
TC-5
26
35.5
65%
21.3
83%
PB
Not spiked
37.9
64%
ND
NA
Average
37.5
68%
22.1
86%
Std Dev
1.94
3.3%
1.05
4.1%
% RSD
5.2%
4.8%
4.8%
4.8%


Stainless Steel
IPA - LT = 10 s


TC-1
26
37.0
69%
20.0
78%
TC-2
26
42.4
73%
21.6
84%
TC-3
26
35.5
68%
23.8
93%
TC-4
26
41.3
73%
24.8
96%
TC-5
26
38.7
70%
21.7
84%
PB
Not spiked
38.7
69%
ND
NA
Average
38.9
70%
22.4
87%
Std Dev
2.58
2.2%
1.91
7.3%
% RSD
6.6%
3.1%
8.5%
8.4%


Stainless Steel -
IPA-LT = 100 s


TC-1
26
30.4
53%
17.6
69%
TC-2
26
25.5
49%
16.1
63%
TC-3
26
27.8
51%
22.3
87%
TC-4
26
22.0
42%
11.0
43%
TC-5
26
38.3
76%
18.4
72%
PB
Not spiked
31.2
55%
ND
NA
Average
29.2
54%
17.1
67%
Std Dev
5.58
11.5%
4.10
16.0%
% RSD
19.1%
21.2%
24.0%
23.9%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
83

-------
Table A-17. Phase I Operational Parameters Optimization - Evaluation of IPA Wetting Agent for
Sampling of Phenol from Semiporous and Porous Materials
Material Type-
Wetting Agent -
Surface Lapse Time
Surface
Loading
(mg)
Phenol
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Vinyl flooring - IPA -
LT = 10 s

TC-1
26
19.2
42%
9.4
37%
TC-2
26
20.3
40%
7.7
30%
TC-3
26
23.9
48%
9.1
35%
TC-4
26
20.0
40%
8.2
32%
TC-5
26
17.8
38%
6.8
26%
PB
Not spiked
22.3
44%
ND
NA
Average
20.6
42%
8.2
32%
Std Dev
2.19
3.6%
1.05
4.3%
% RSD
10.7%
8.5%
12.8%
13.4%
Plywood - IPA - LT = 10 s
TC-1
26
14.6
25%
0.8
3%
TC-2
26
10.6
20%
0.6
2%
TC-3
26
11.2
19%
0.4
2%
TC-4
26
11.2
19%
0.4
1%
TC-5
26
11.3
21%
0.6
2%
PB
Not spiked
11.6
23%
ND
NA
Average
11.8
21%
0.6
2%
Std Dev
1.43
2.4%
0.17
0.7%
% RSD
12.2%
11.3%
29.9%
35.4%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
84

-------
Table A-18. Phase I Operational Parameters Optimization - Evaluation of IPA Wetting Agent for
Sampling of Nitrobenzene from Semiporous and Porous Materials
Material Type-
Wetting Agent -
Surface Lapse Time
Surface
Loading
(mg)
Nitrobenzene
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Vinyl flooring - IPA-
LT = 10 s

TC-1
29
22.0
48%
3.7
13%
TC-2
29
19.5
53%
2.7
10%
TC-3
29
29.2
39%
4.1
14%
TC-4
29
20.0
59%
3.6
13%
TC-5
29
26.7
40%
4.0
14%
PB
Not spiked
22.8
48%
ND
NA
Average
23.4
48%
3.6
13%
Std Dev
3.84
7.6%
0.55
1.6%
% RSD
16.4%
15.9%
15.3%
12.8%
Plywood - IPA - LT = 10 s
TC-1
29
10.7
20%
1.2
4%
TC-2
29
8.7
17%
0.9
3%
TC-3
29
9.6
19%
1.1
4%
TC-4
29
19.7
33%
1.1
4%
TC-5
29
14.1
26%
1.2
4%
PB
Not spiked
12.6
22%
ND
NA
Average
12.6
23%
1.1
4%
Std Dev
4.01
5.8%
0.12
0.4%
% RSD
31.9%
25.6%
11.1%
11.8%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
85

-------
Table A-19. Phase I Operational Parameters Optimization - Evaluation of IPA Wetting Agent for
Sampling of TEP from Semiporous and Porous Materials
Material Type-
Wetting Agent -
Surface Lapse Time
Surface
Loading
(mg)
TEP
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Vinyl flooring - IPA -
LT = 10 s

TC-1
26
24.7
48%
6.4
25%
TC-2
26
31.0
61%
9.3
36%
TC-3
26
20.7
42%
5.8
23%
TC-4
26
27.6
53%
7.2
28%
TC-5
26
26.0
50%
7.5
29%
PB
Not spiked
23.6
46%
ND
NA
Average
25.6
50%
7.2
28%
Std Dev
3.53
6.5%
1.33
5.0%
% RSD
13.8%
13.1%
18.4%
17.6%
Plywood - IPA - LT = 10 s
TC-1
26
11.7
22%
1.3
5%
TC-2
26
18.6
31%
1.0
4%
TC-3
26
17.3
31%
0.9
4%
TC-4
26
16.1
28%
1.5
6%
TC-5
26
10.4
20%
0.8
3%
PB
Not spiked
15.9
33%
ND
NA
Average
15.0
28%
1.1
4%
Std Dev
3.23
5.3%
0.29
1.1%
% RSD
21.6%
19.3%
26.5%
25.9%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND- Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
86

-------
Table A-20. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Phenol from Stainless
Steel
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
Phenol
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless Steel - Bissell Little Green ProHeat - IPA- LT = 10 s
TC-1
26
26.6
59%
9.8
38%
TC-2
26
40.1
69%
12.8
50%
TC-3
26
34.6
67%
10.7
42%
TC-4
26
29.4
62%
10.3
40%
TC-5
26
27.5
58%
9.9
39%
PB
Not spiked
36.1
66%
ND
NA
Average
32.4
64%
10.7
42%
Std Dev
5.38
4.5%
1.23
4.8%
% RSD
16.6%
7.1%
11.5%
11.5%
Stainless Steel -Rug Doctor Portable Spot Cleaner - IPA- LT = 10 s
TC-1
26
17.1
32%
4.8
19%
TC-2
26
15.2
70%
4.9
19%
TC-3
26
19.8
39%
5.1
20%
TC-4
26
24.1
42%
5.8
23%
TC-5
26
22.5
43%
5.4
21%
PB
Not spiked
23.5
41%
ND
NA
Average
20.4
45%
5.2
20%
Std Dev
3.63
13.1%
0.41
1.7%
% RSD
17.8%
29.4%
7.8%
8.2%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
87

-------
Table A-21. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Phenol from Laminate
Flooring
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
Phenol
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Laminate flooring -Bissell Little Green ProHeat - IPA- LT = 10 s
TC-1
26
31.2
64%
10.9
43%
TC-2
26
34.7
64%
11.5
45%
TC-3
26
20.2
51%
9.1
35%
TC-4
26
21.0
53%
9.2
36%
TC-5
26
23.1
57%
10.2
40%
PB
Not spiked
29.7
61%
ND
NA
Average
26.7
58%
10.2
40%
Std Dev
6.02
5.6%
1.05
4.3%
% RSD
22.6%
9.6%
10.3%
10.9%
Laminate flooring - Rug Doctor Portable Spot Cleaner -1
PA-LT = 10s

TC-1
26
19.8
33%
5.9
23%
TC-2
26
18.3
35%
4.0
16%
TC-3
26
16.7
32%
3.7
14%
TC-4
26
19.6
36%
5.7
22%
TC-5
26
16.9
28%
5.2
20%
PB
Not spiked
12.8
26%
ND
NA
Average
17.4
32%
4.9
19%
Std Dev
2.58
3.9%
1.00
3.9%
% RSD
14.9%
12.4%
20.4%
20.4%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
88

-------
Table A-22. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Phenol from Vinyl
Flooring
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
Phenol
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Vinyl flooring - Bissell Little Green ProHea
t-IPA-LT = 10s

TC-1
26
19.0
51%
8.2
32%
TC-2
26
15.6
48%
6.9
27%
TC-3
26
27.1
60%
8.7
34%
TC-4
26
22.9
39%
5.7
22%
TC-5
26
25.8
51%
7.2
28%
PB
Not spiked
22.4
43%
ND
NA
Average
22.1
49%
7.3
29%
Std Dev
4.27
7.3%
1.17
4.7%
% RSD
19.3%
15.0%
16.0%
16.3%
Vinyl floorin
g - Rug Doctor Portable Spot Cleaner - IPA- LT = 10 s

TC-1
26
17.1
32%
3.1
12%
TC-2
26
19.4
38%
4.7
18%
TC-3
26
21.3
38%
4.7
18%
TC-4
26
12.0
25%
3.0
12%
TC-5
26
20.7
40%
4.6
18%
PB
Not spiked
18.6
38%
ND
NA
Average
18.2
35%
4.0
16%
Std Dev
3.38
5.7%
0.89
3.3%
% RSD
18.6%
16.1%
22.1%
21.1%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
89

-------
Table A-23. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Nitrobenzene from
Stainless Steel
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
Nitrobenzene
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless Steel - Bissell Little Green ProHeat - IPA - LT = 10 s
TC-1
29
31.4
59%
1.9
6%
TC-2
29
30.5
58%
2.0
7%
TC-3
29
34.7
68%
2.3
8%
TC-4
29
28.2
59%
1.5
5%
TC-5
29
38.5
66%
2.3
8%
PB
Not spiked
24.0
54%
ND
NA
Average
31.2
61%
2.00
7%
Std Dev
5.04
5.3%
0.33
1.3%
% RSD
16.1%
8.7%
16.6%
19.2%
Stainless Steel - Rug Doctor Portable Spot Cleaner - IPA - LT = 10 s
TC-1
29
10.0
25%
0.0
0%
TC-2
29
21.0
39%
0.5
2%
TC-3
29
19.0
36%
0.4
1%
TC-4
29
25.3
44%
0.7
2%
TC-5
29
25.8
46%
0.8
3%
PB
Not spiked
23.9
40%
ND
NA
Average
20.8
38%
0.48
2%
Std Dev
5.91
7.4%
0.31
1.1%
% RSD
28.4%
19.4%
64.9%
71.3%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB -SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
90

-------
Table A-24. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Nitrobenzene from
Laminate Flooring
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
Nitrobenzene
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Laminate flooring - Bissell Little Green ProHeat - IPA- LT = 10 s
TC-1
29
34.5
72%
1.4
5%
TC-2
29
22.4
82%
1.5
5%
TC-3
29
41.9
80%
2.7
9%
TC-4
29
49.0
82%
3.2
11%
TC-5
29
37.9
79%
2.3
8%
PB
Not spiked
33.5
80%
ND
NA
Average
36.5
79%
2.22
8%
Std Dev
8.94
3.7%
0.77
2.6%
% RSD
24.5%
4.7%
34.8%
34.3%
Laminate flooring - Rug Doctor Portable Spot Cleaner -1
PA-LT = 10s

TC-1
29
21.1
42%
0.5
2%
TC-2
29
22.6
41%
0.6
2%
TC-3
29
22.1
14%
0.6
2%
TC-4
29
21.5
40%
0.5
2%
TC-5
29
17.9
34%
0.6
2%
PB
Not spiked
13.3
30%
ND
NA
Average
19.8
34%
0.56
2%
Std Dev
3.57
10.6%
0.05
0.0%
% RSD
18.1%
31.7%
9.8%
0.0%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
91

-------
Table A-25. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of Nitrobenzene from
Vinyl Flooring
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
Nitrobenzene
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Vinyl flooring - Bissell Little Green ProHea
t-IPA-LT = 10s

TC-1
29
24.4
59%
ND
NA
TC-2
29
35.6
64%
ND
NA
TC-3
29
41.7
69%
ND
NA
TC-4
29
28.7
64%
ND
NA
TC-5
29
31.7
65%
ND
NA
PB
Not spiked
31.9
62%
ND
NA
Average
32.3
64%
ND
NA
Std Dev
5.92
3.3%
ND
NA
% RSD
18.3%
5.2%
ND
NA
Vinyl floorin
g - Rug Doctor Portable Spot Cleaner - IPA- LT = 10 s

TC-1
29
25.7
52%
ND
NA
TC-2
29
32.1
65%
ND
NA
TC-3
29
25.9
94%
ND
NA
TC-4
29
34.1
67%
ND
NA
TC-5
29
36.1
74%
ND
NA
PB
Not spiked
42.2
74%
ND
NA
Average
32.7
71%
ND
NA
Std Dev
6.31
13.9%
ND
NA
% RSD
19.3%
19.5%
ND
NA
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
92

-------
Table A.26 Phase II Evaluation of Commercial Wet-Vacuums for Sampling of TEP from Stainless
Steel Coupon


TEP


Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
Volume of Wetting
Solvent Collected
(mL)
Wetting Liquid
Recovered
(%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless-steel coupon - Bissell Little Green ProHeat - IPA- LT
= 10s


TC-1
26
26.0
69%
10.7
42%
TC-2
26
38.7
70%
5.8
23%
TC-3
26
37.1
72%
15.2
59%
TC-4
26
34.4
66%
8.9
35%
TC-5
26
40.6
80%
11.8
46%
PB
Not spiked
40.0
78%
ND
NA
Average
36.1
73%
10.48
41%
Std Dev
5.44
5.4%
3.48
13.3%
% RSD
15.1%
7.5%
33.2%
32.5%

Stainless-steel coupon - Rug Doctor Portable Spot Cleaner
-IPA-LT = 10s

TC-1
26
25.2
52%
3.5
14%
TC-2
26
36.7
62%
5.1
20%
TC-3
26
29.7
64%
5.0
20%
TC-4
26
58.0
69%
4.6
18%
TC-5
26
34.7
63%
4.5
18%
PB
Not spiked
39.1
68%
ND
NA
Average
37.2
63%
4.54
18%
Std Dev
11.34
6.1%
0.63
2.4%
% RSD
30.5%
9.6%
14.0%
13.6%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
93

-------
Table A-27. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of TEP from Laminate
Flooring
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
TEP
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Laminate flooring - Bissell Little Green ProHeat - IPA- LT = 10 s
TC-1
26
33.0
60%
8.3
32%
TC-2
26
37.8
72%
8.7
34%
TC-3
26
27.2
68%
9.5
37%
TC-4
26
30.9
64%
5.6
22%
TC-5
26
26.9
66%
7.0
27%
PB
Not spiked
34.3
69%
ND
NA
Average
31.7
67%
7.82
30%
Std Dev
4.23
4.2%
1.54
5.9%
% RSD
13.4%
6.3%
19.6%
19.5%
Laminate flooring-Rug Doctor Portable Spot Cleaner - IPA- LT = 10 s
TC-1
26
23.3
42%
7.0
27%
TC-2
26
23.9
41%
4.1
16%
TC-3
26
18.9
28%
7.7
30%
TC-4
26
20.3
35%
4.3
17%
TC-5
26
20.8
37%
6.0
23%
PB
Not spiked
8.3
18%
ND
NA
Average
19.3
34%
5.82
23%
Std Dev
5.68
9.1%
1.60
6.1%
% RSD
29.5%
27.1%
27.5%
27.0%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time).
94

-------
Table A-28. Phase II Evaluation of Commercial Wet-Vacuums for Sampling of TEP from Vinyl
Flooring
Material and Vacuum
Type-Wetting
Agent - Surface
Lapse Time
Surface
Loading
(mg)
TEP
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Vinyl flooring - Bissell Little Green ProHea
t-IPA-LT = 10s

TC-1
26
20.1
60%
5.0
20%
TC-2
26
30.4
59%
5.8
22%
TC-3
26
31.8
63%
2.3
9%
TC-4
26
36.3
66%
6.5
25%
TC-5
26
44.4
73%
6.2
24%
PB
Not spiked
31.7
73%
ND
NA
Average
32.5
66%
5.16
20%
Std Dev
7.94
6.2%
1.69
6.4%
% RSD
24.5%
9.4%
32.8%
32.2%
Vinyl floorin
g - Rug Doctor Portable Spot Cleaner - IPA- LT = 10 s

TC-1
26
30.2
62%
2.0
8%
TC-2
26
28.8
52%
2.2
9%
TC-3
26
23.8
47%
1.7
7%
TC-4
26
31.5
61%
2.7
11%
TC-5
26
36.7
70%
3.7
14%
PB
Not spiked
10.1
22%
ND
NA
Average
26.9
52%
2.46
10%
Std Dev
9.20
16.9%
0.78
2.8%
% RSD
34.3%
32.3%
31.8%
28.3%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND- -Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time)

-------
Table A-29. Phase II Optimization of Wetting Solvent and Chemical Recovery Using Bissell Little
Green ProHeat Wet-Vacuum - Phenol on Stainless-steel Coupon
Phenol
Surface loading: 85 mg
Wetting agent: IPA
Procedure:
Vacuum conditioning -
Chemical
Recovered
Wetting Solvent
Recovered
(%)
Recovery
of Rinse
(%)
surface wetting-
vacuum rinse (volumes,
mL)
Mass
(mg)
Recovery
(%)
Volume
(mL)
Recovery
(%)
Volume
(mL)
Recovery
(%)
50 mL + 50 mL + no rinse
20.8
24%
27
57%
NA
no rinse
19.7
23%
26
50%
NA
no rinse
35.1
41%
33
68%
NA
no rinse
Average
25.2
30%
28.7
59%
NA
NA
SD
8.60
10%
3.63
9.1%
NA
NA
%RSD
34%
13%
16%
NA
NA
50mL + 50mL+100mL
67
79%
32
65%
76
76%
72
84%
30
62%
85
85%
66
77%
27
54%
79
79%
Average
68.1
80%
29.6
61%
80.3
80%
SD
3.10
3.6%
2.49
5.5%
4.68
4.7%
%RSD
4.5%
8.4%
9.1%
5.8%
50 mL + 50 mL +200 mL
59
69%
30
62%
133
66%
60
70%
31
67%
135
67%
56
66%
29
63%
126
63%
Average
58.1
68%
30.1
64%
131
66%
SD
2.02
2.4%
1.27
2.2%
4.39
2.2%
%RSD
3.5%
4.2%
3.5%
3.3%
96

-------
Table A-30. Phase II Optimization of Wetting Solvent and Chemical Recovery Using IPA and Bissell
Little Green ProHeat Wet-Vacuum - TEP on Stainless-steel Coupon
Surface loading: 27 mg
Wetting agent: IPA
Procedure:
Chemical
Recovered
Wetting Solvent
Recovered
Recovery
of Rinse
wetting-
vacuum rinse (volumes, mL)
Mass
(mg)
Recovery
(%)
Volume
(mL)
Recovery
(%)
Volume
(mL)
Recovery
(%)
50 mL + 50 mL + no rinse
8.94
34%
24.2
54%
NA
no rinse
5.40
20%
24.6
53%
NA
no rinse
8.87
26%
28.6
58%
NA
no rinse
Average
7.07
27%
25.8
55%
NA
NA
SD
1.78
6.7%
2.47
2.9%
NA
NA
%RSD
25%
10%
5.2%
NA
NA
50mL + 50 mL+100 mL
11.3
43%
27.4
60%
75.8
76%
13.1
50%
30.2
58%
79.4
79%
Average
11.2.
46%
28.8
59%
77.6
78%
SD
1.27
4.8%
1.98
0.8%
2.52
2.5%
%RSD
10%
6.9%
1.4%
3.2%
50mL + 50 mL + 100 mL
38
143%
27.1
56%
173
86%
20
74%
29.9
65%
167
84%
Average
28.8
109%
28.5
60%
170
85%
SD
12.9
45%
1.98
6.1%
3.6
1.8%
%RSD
45%
6.9%
10.2%
2.1%
97

-------
Table A-31. Phase III Optimized IPA Wet-Vacuum Method Performance for 100% Surface
Concentration Reference Material Sampling Baseline of 2-CEPS, including Mass Balance Tests
Material and Vacuum
Type-Wetting
Agent-Vol-Surface
Lapse Time - Rinse
Vol
Surface
Loading
(mg)
2-CEPS
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless-steel coupon - Bissell Little Green ProHeat - 50 mL IPA (LT = 10 s) -100 mL post rinse
Target concentration: 100%
TC-1
28
100
66%
21.4
76%
TC-2
28
118
73%
23.5
83%
TC-3
28
111
73%
20.6
73%
PB
not-spiked
111
73%
ND
NA
Average
110
71%
21.8
77%
Std Dev
7.6
3.7%
1.5
5.1%
% RSD
6.9%
5.2%
6.9%
6.9%
Mass balance samples
TC-1-VW
NA
NA
0.0011 (J)
0.004%
TC- 2-VW
NA
NA
0.0011 (J)
0.004%
TC- 3-VW
NA
NA
0.0010 (J)
0.004%
TC-1-AR1
89.7
89%
3.58
13%
TC-1-AR2
89.9
90%
0.70
2.5%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB - SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND - Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time); VW - Post-vacuum sampling surface wipe; AR - Additional post-rinse; (J) - Estimated, reported value below lowest point of calibration curve.
98

-------
Table A-32. Phase III Optimized TEP Wet-Vacuum Method Performance for 100% Surface
Concentration Reference Material Sampling Baseline of TEP, including Mass Balance Tests
Material and Vacuum
Type-Wetting
Agent-Vol-Surface
Lapse Time - Rinse
Vol
Surface
Loading
(mg)
TEP
Volume of Wetting Wetting Liquid
Solvent Collected Recovered
(mL) (%)
Mass of Chemical
Collected
(mg)
Chemical
Recovery
(%)
Stainless-steel coupon - Bissell Little Green ProHeat - 50 mL IPA (LT = 10 s) -100 mL post-rinse
Target concentration: 100%
TC-1
26
109
70%
18.5
72%
TC-2
26
111
74%
17.9
70%
TC-3
26
115
73%
19.7
77%
PB
not-spiked
109
73%
ND
NA
Average
111
72%
18.7
73%
Std Dev
2.9
1.7%
0.93
3.6%
% RSD
2.7%
2.4%
4.7%
4.7%
Mass balance samples
TC-1-VW
NA
NA
ND
NA
TC- 2-VW
NA
NA
ND
NA
TC- 3-VW
NA
NA
ND
NA
TC-1-AR1
89.7
90%
1.73
6.7%
TC-1-AR2
89.9
90%
0.32
1.2%
TC - SL collected using wet-vacuum apparatus from contaminated stainless-steel coupon; PB SL collected using wet-vacuum apparatus from non-
contaminated (procedural blank) stainless-steel coupon; ND- Not detected; NA - Not applicable; LT - Surface lapse time (wetting agent surface contact
time); VW - Post-vacuum sampling surface wipe; AR - Additional post-rinse; (J) - Estimated, reported value below lowest point of calibration curve

-------
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

-------