EPA/600/R-11/156C
November 2012
Laboratory Study of Poly chlorinated Biphenyl (PCB) Contamination and
Mitigation in Buildings
Part 4. Evaluation of the Activated Metal Treatment System (AMIS) for On-site
Destruction ofPCBs
Xiaoyu Liu and Zhishi Guo
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
and
Corey A. Mocka, R. Andy Stinson, Nancy F. Roache, and Joshua A. Nardin
ARCADIS, US Inc.
4915 Prospectus Dr., Suite F
Durham, NC 27709
-------
NOTICE
This document has been reviewed internally and externally in accordance with the U.S. Environmental
Protection Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-------
Executive Summary
E.I Background
Polychlorinated biphenyls (PCBs) were once used as a plasticizer in certain building materials such as
caulking, sealants, and paints from the 1950s through the late 1970s. Because PCBs have a variety of
adverse health effects in animals and human, federal regulations have specific requirements for use and
disposal of PCB-containing materials (U.S. EPA, 2005; 2009). Briefly, building materials that contain 50
ppm or more PCBs are not authorized for use and must be disposed of as PCB bulk product waste according
the Code of Federal Regulations 40 CFR §761.3 and §761.62. If PCBs have contaminated eitherthe
surrounding building materials or adjacent soil, these materials are considered PCB remediation waste,
which is subject to the cleanup and disposal requirements according 40 CFR §761.61.
There are many methods for mitigating PCB contamination in buildings. Source removal is one of the most
commonly used methods. It can be accomplished either by physically removing the sources (such as caulk
and light ballasts) or by chemically transforming PCBs into non-hazardous reaction products (EH&E,
2012). One of the chemical destruction techniques for reductive dechlorination of PCBs is the use of zero-
valent metals, a technique that uses the Bimetallic Treatment System (BTS) or the Activated Metal
Treatment System (AMTS). Both treatment systems are paste-like materials that contain a reducing agent,
organic solvents/hydrogen donor, and thickening agent. The BTS method uses elemental magnesium (Mg)
powder coated with a small amount of a palladium (Pd) complex as the reducing agent while the AMTS
method uses only magnesium powder. According to Krug et al. (2010), the organic solvents in the BTS or
AMTS paste can penetrate into the PCB-containing material and extract the PCBs from the material. Then
the reducing agent in the paste degrades the extracted PCBs rapidly and thoroughly. Scientists from the
National Aeronautics and Space Administration (NASA) and the University of Central Florida have refined
the technique and applied it to the Department of Defense (DoD) facilities that had PCB levels as high as
11,000 mg/kg in the painted surfaces of concrete and greater than 50 mg/kg in the painted surfaces of
steel (Krug et al., 2010). According to the authors, the demonstration of the technique at the DoD facilities
showed rapid dechlorination of PCBs to less than 50 ppm in painted steel and concrete surfaces in about a
week.
To expand this technique to other buildings, such as school buildings, several questions must be
addressed, including:
• What is the efficiency of this technique for removing PCBs from indoor sources such as PCB-
containing caulk, masonry materials, and paint?
• For thick sources such as PCB-containing caulk and concrete, how deep can the treatment system
penetrate into the substrate to effectively remove PCBs?
• After PCBs are removed from the top layer of thick material, the PCBs in the deep layer may slowly
migrate into the top layer, which is sometimes referred to as "bleed-back." Will PCB bleed-back occur
after treatment?
m
-------
• Do the waste products from application of this technique cause any environmental concerns?
This study was intended to address these questions by using a combination of laboratory testing and
mathematical modeling approaches. The results will be used to support the development of guidelines and
decision-making tools for PCB mitigation.
The BTS and AMTS methods are closely related. Only the AMTS method was evaluated in this study
because it was developed more recently and it is more cost- effective than the BTS method.
The AMTS paste can be applied to the surface of a PCB source either using spray-on or wipe-on techniques.
The solvents in the mixture extract the PCBs from the source. The extracted PCBs are then destroyed by
reacting with the reducing agent. There are two types of AMTS pastes: the "active paste" and the "inactive
paste". The former contains a reducing agent (magnesium powder) whereas the inactive paste does not.
When the active paste is used, the extraction and chemical reaction take place at the same time. When the
inactive paste is used, the paste needs to be collected from the source after treatment and then placed in a
container where the extracted PCBs are destroyed by reacting with the reducing agent.
E.2 Objectives
The main goal of this study was to evaluate the effectiveness and usefulness of the AMTS method for
decontamination of PCBs in buildings. The objectives were to:
• Evaluate the performance of the AMTS method for laboratory-mixed PCB paint, caulk, and
concrete, as well as field caulk, under laboratory conditions.
• Evaluate the dechlorination mechanism of the AMTS method.
• Use modeling tools to predict PCB bleed-back from the AMTS-treated PCB sources.
E.3 Methods
E. 3.1 Preparation of the AMTS Pastes
The active and inactive AMTS pastes (Figure E. 1) were made prior to use according to the formulas
provided by the NASA scientists. The active paste contains magnesium powder, ethanol, glacial acetic acid,
limonene, and several other chemicals. The inactive paste does not contain magnesium powder. The exact
formulas of the AMTS pastes and the preparation procedures are proprietary information owned by NASA.
IV
-------
Figure E.I. AMTS Pastes (left - inactive paste; right - active paste)
E. 3.2 Application of the AMTS Method to PCB-Contaminated Materials
The AMTS pastes were tested on the following PCB-containing materials:
• Three types of coatings, i.e., oil-based primer ("primer"), oil-based alkyd paint ("alkyd"), and
solvent-free epoxy coating ("epoxy"). Separate samples of each type of coating were spiked with
either a high or a low concentration of Aroclor 1254, approximately 0.7% and 0.3% by weight,
respectively (Figure E.2)
• Two types of PCB-containing caulk from buildings (Figure E.3) and two laboratory-made
polysulfide caulk samples spiked with different concentrations of Aroclor 1254
• Laboratory-made concrete substrates spiked with Aroclor 1254 (Figure E.4)
-------
Figure E.2. Application of the inactive paste to a coating coupon
Figure E.3. Field caulk covered with the active paste
VI
-------
Figure E.4. Concrete pieces in a stainless steel mold for treatment by the AMTS method
The test coupons were treated for 5 to 10 days. After treatment, the AMTS paste was removed from the
coupons. Then, the test coupons were placed in scintillation vials for sonication extraction and GC/MS
analysis.
E. 3.3 Using a Barrier Model
After treatment by the AMTS, PCBs are eliminated from the top layer of the source. This layer acts as a
barrier to the migration of PCBs from the interior of the source to the surface. To evaluate the protective
effect of this barrier layer, a fugacity-based, multi-layer model (Yuan et al, 2007) was used to predict the
"bleed-back" of PCBs (i.e., migration of PCBs from the deep layer of the source into the treated top layer).
The material/air partition coefficients and solid-phase diffusion coefficients, two parameters required by the
model, were estimated from other experiments, and the effective penetration depths for field caulk and
laboratory-mixed concrete obtained from this study were used as inputs to the model. The outputs from the
model included the concentration profiles of a selected PCB in the source and barrier layers as functions of
time and depth and the contribution of the AMTS treated source to indoor air pollution as a function of time.
The modeling results allowed the calculation of the average PCB concentrations in the barrier layer and the
concentrations of the PCB on the exposed surfaces.
E.4 Findings
E. 4.1 Efficiency of the AMTS Method
The test results presented in Section 4 demonstrate that the AMTS method removed PCBs from the primer
and the alkyd paint effectively. The removal efficiencies were greater than 80% after seven days of
treatment. These results are consistent with those reported by Krug et al. (2010). The removal efficiency
was higher for the primer (> 95%) than for the alkyd paint (> 81%).
VII
-------
The removal efficiencies for the caulk and concrete samples were much lower than those for the coating
materials, i.e., 12 to 36% of PCB congeners (4 to 36% in terms of Aroclor 1254) for the field caulk and 39
to 54% of PCB congeners (27 to 52% in terms of Aroclor 1254) for the laboratory-made concrete (Figure
E.5).
The AMTS method was evaluated at two different PCB concentration levels for each type of material. The
results showed that there was not much difference in the PCB removal efficiencies for the two
concentrations for paint, field caulk and laboratory-made concrete.
100%
Figure E.5. Efficiency of the AMTS method in removing PCBs from different materials
(AP = active paste; IP = inactive paste; H = high PCB concentrations; L = low PCB concentrations; FC
field caulk; LC = laboratory-mixed concrete)
E.4.2 Effective Penetration Depths
The percent removal efficiency is a good indicator of the effectiveness of the AMTS method for coating
materials, but the removal efficiency may be misleading for thick sources such as caulk and concrete
because of the limited ability of the solvents in the pastes to penetrate the source substrate completely. The
concept of "effective penetration depth" was introduced to determine the effectiveness of the AMTS for
thick sources. The effective penetration depth is the thickness of a layer of the source material near the
treated surface in which all PCBs are removed and beyond which the PCBs remain intact. This parameter is
independent of the thickness of the source that is being treated (Section 5.2). The test results yielded an
average effective penetration depth of approximately 1 mm for field caulk with 50% relative standard
deviation (RSD) and approximately 3 mm for the laboratory-made concrete with 16% RSD.
vm
-------
E. 4.3 Mathematical Modeling
After the AMTS treatment, the top layer of a thick source from which the PCBs has been removed acts as a
temporary barrier to the migration of the PCBs from the untreated portion of the source substrate to the
treated surface. It is important to know whether this barrier can effectively reduce the PCB concentration in
room air and how long the protective effect lasts. To obtain a basic understanding of these issues, a fugacity-
based barrier model was used to predict the "bleed-back" of PCBs for caulk and concrete after treated by
AMTS. This model requires the material/air partition coefficient and solid-phase diffusion coefficient for the
substrates. These parameters were obtained from the authors' previous studies (Guo et al., 2011, 2012a).
The simulation results demonstrated that, when the AMTS method is used to treat thick sources that contain
have very high concentrations of PCBs, the potential effect of PCB bleed-back on the PCB concentrations at
the treated surface and in room air must be considered. Figures E.6 shows the simulation results for the
concentration profiles of congener #110, the most abundant congener in Aroclor 1254, for treated concrete
as a function of time and the depth of the test materials. It was assumed that the thickness of the PCB-
containing layer of the concrete is 10 mm, that the initial concentration of congener #110 in the concrete is
196 ug/g, and that the AMTS creates an effective penetration depth of 3 mm.
250
200
Bottom layer
(C01 = 196 ug/g)
Top layer
(C02 =
10 days
•100 days
•500 days
1000 days
•5000 days
Figure E.6. Concentration profiles for congener #110 in the PCB-containing concrete as a function
of time and depth, C(x,t), after the AMTS treatment
(Coi and Co2 are the initial concentrations in the bottom and top layers, respectively; the effective
penetration depth is 3 mm, i.e., between x = 7 and x = 10 mm in the x-axis.)
In general, the effect of PCB bleed-back for thick sources treated by AMTS on the PCB concentrations in
room air is dependent on several factors, including the initial PCB concentration in the source, the effective
penetration depth, the resistance of the substrate to PCB migration, and time. It could be a limiting factor for
treating sources with very high PCB content. Therefore, post-treatment environmental monitoring, such as
periodical air and wipe sampling, is necessary. Mathematical modeling can help understand the general
behavior of the bleed-back phenomenon. Details are presented in Section 5.3 of the main body.
IX
-------
E. 4.4 Summary of Findings
Overall, the AMTS method has the potential to become a viable method for mitigating PCB contamination
in buildings. This method is promising for treating contaminated masonry materials near the expansion
joints after the caulking material is removed because the AMTS can treat sources that contain several
thousand ppm PCBs, for which the encapsulation method may not be effective. However, more research is
needed because the current method has limited effective penetration depths for thick sources, including
masonry materials. In addition, the reaction mechanism needs to be verified. The potential effect of the
method on indoor air quality during the treatment also needs further evaluation.
E.5 Study Limitations
This study was conducted in a relatively short period of time with a limited scope. The main focus was the
performance of the AMTS method as expressed in PCB removal efficiency and effective penetration depth.
There are several important areas that this study did not investigate, including (1) the effectiveness of
multiple treatments; (2) the properties and conditions of the source materials after being treated with AMTS;
(3) the PCB residual concentration in the active paste after treatment. Furthermore, the study on the reaction
mechanism of the AMTS method is inconclusive, and the modeling results for PCB bleed-back are semi-
quantitative.
This study evaluated the AMTS technology that was available in early 2011. Since then, the developer of
this method has modified the formulation and application procedure aimed to improve the performance of
the method for contaminated masonry materials. This study did not evaluate this new method.
-------
TABLE OF CONTENTS
Executive Summary iii
E.I Background iii
E.2 Objectives iv
E.3 Methods iv
E.3.1 Preparation of the AMTS Pastes iv
E.3.2 Application of the AMTS Method to PCB-Contaminated Materials v
E.3.3 Using a Barrier Model vii
E.4 Findings vii
E.4.1 Efficiency of the AMTS Method vii
E.4.2 Effective Penetration Depths viii
E.4.3 Mathematical Modeling ix
E.4.4 Summary of Findings x
E.5 Study Limitations x
List of Tables xiii
List of Figures xv
Acronyms and Abbreviations xvii
1. Introduction 1
1.1 Background 1
1.2 Goals and Objectives 2
1.3 About This Report 3
2. Experimental Methods 4
2.1 Activated Metal Treatment System Pastes 4
2.1.1 Preparation of AMTS Pastes 4
2.1.2 Application and Removal of the AMTS Pastes 5
2.2 Test Procedures 5
2.2.1 Tests for Coating Materials 8
2.2.2 Tests for Field Caulk 10
2.2.3 Tests for Laboratory-made Caulk 13
2.2.4 Tests for Laboratory-made Concrete 15
2.2.5 Reaction Mechanism Study 17
2.3 Sample Extraction and Analysis 18
2.3.1 Sample Extraction 18
2.3.2 Analytical Instrument and Reagents 18
XI
-------
2.3.3 Sample Analysis for AMTS Efficiency Tests 19
2.3.4 Sample Analysis for the Reaction Mechanism Study 19
2.4 Environmental Parameters 21
2.5 Measurement of the Dimensions of Test Coupons 21
3. Quality Assurance and Quality Control 23
3.1 Data Quality Indicator Goals for Critical Measurements 23
3.2 GC/MS Instrument Calibration 23
3.3 Detection Limits 25
3.4 Quality Control Samples 26
3.5 Daily Calibration Check 29
3.6 Recovery Check Standards 30
4. Results 32
4.1 Terminology and Definitions 32
4.2 Tests for Coating Materials 32
4.3 Tests for Caulk Materials 36
4.4 Tests for Concrete 41
4.5 Partial Dechlorination Investigation 44
5. Discussion 46
5.1 General Performance of the AMTS Method 46
5.2 Effective Penetration Depth 47
5.3 Predicting the "Bleed-back" of PCBs from Treated Sources 49
5.3.1 Model Description 49
5.3.2 Simulations for Treated Caulk 51
5.3.3 Simulations for Treated Concrete 54
5.3.4 Summary of Mathematical Modeling 57
5.3.5 Limitations of Mathematical Modeling 57
5.4 Reaction Mechanisms 58
5.5 Organic Solvents 59
5.6 Study Limitations 59
6. Conclusions 60
7. Recommendations 62
Acknowledgments 63
References 64
Xll
-------
List of Tables
Table 2.1. Components of the AMTS paste 4
Table 2.2. Substrate materials tested 6
Table 2.3. Summary of AMTS evaluation tests 7
Table 2.4. Laboratory preparation of Aroclor 1254-spiked coating 8
Table 2.5. Preparation of Aroclor 1254-spiked laboratory-mixed caulk 14
Table 2.6. Preparation of Aroclor 1254-spiked laboratory-made concrete 15
Table 2.7. Preparation of PCB-209-spiked paint for the reaction mechanism study 17
Table 2.8. Chemical names and CAS Registration Numbers for the internal standards and recovery
check standards 19
Table 2.9. Chemical names and CAS Registration Numbers for the PCB congeners analyzed 19
Table 2.10. Chemical names and CAS Registration Numbers for EPA Method 680 20
Table 2.11. Operating conditions for the Agilent 6890/5973N GC/MS for EPA Method 680 20
Table 2.12. SIM acquisition parameters for the Agilent 6890/5973N GC/MS for the analysis of PCB
homologues 2 1
Table 3.1. GC/MS calibration for PCB congeners 24
Table 3.2. IAP results for each calibration of PCB congeners 25
Table 3.3. Instrument detection limits (IDLs) for PCB congeners on GC/MS (ng/mL) 26
Table 3 .4. Summary of method blank for tests (ng/sample) 27
Table 3.5. Summary of extraction method blanks for tests (ng/sample) 28
Table 3.6. Summary of solvent blank for tests (ng/sample) 28
Table 3.7. Summary of duplicate samples for PCB congeners 29
Table 3.8. Average recoveries of DCCs for AMTS method evaluation tests 30
Table 3.9. Results summary of the recovery check standards 3 1
Table 4.1. Weight fractions (Fj) of selected PCB congeners used to calculate Aroclor 1254
concentration 32
Table 4.2. Summary of the concentrations of PCB congener ((ig/g) and Aroclor 1254 (w/w) measured
by GC/MS in the untreated test materials 33
Table 4.3. Percent removal efficiencies for the sum of the target PCB congeners for coating materials
after the AMTS treatment 3 4
Table 4.4 Percent removal efficiencies for Aroclor 1254 after AMTS treatment of coatings 35
Table 4.5. Summary of the concentrations of PCB congeners ((^g/g) and Aroclor 1254 (w/w)
measured by GC/MS in the original caulk 37
Table 4.6. Percent removal efficiencies for PCB congeners for field caulk after AMTS treatment 38
Table 4.7. Percent removal efficiencies for Aroclor 1254 for field caulk after AMTS treatment 39
Xlll
-------
Table 4.8. Summary of depth (height) measurements of pieces of field caulk 40
Table 4.9. Summary of the concentrations of the PCB congeners and Aroclor 1254 in the untreated,
laboratory-made concrete measured by GC/MS 41
Table 4.10. Average percent removal efficiencies of the AMTS treatment for PCB congeners in
laboratory-made concrete 42
Table 4.11. Removal efficiencies of Aroclor 1254 after AMTS treatment of laboratory-made concrete 43
Table 4.12. Concentrations of PCB homologues in the untreated primer measured by GC/MS 44
Table 5.1. Comparison of removal efficiencies expressed as percent efficiency and effective
penetration depth using hypothetical values 48
Table 5.2. Effectiveness of the AMTS method expressed as effective penetration depth for the field
caulk 49
Table 5.3. The effectiveness of the AMTS method expressed as effective penetration depth for the
laboratory-made concrete 49
Table 5.4. Classroom scenario for the simulation using the fugacity-based barrier model 50
Table 5.5. Input parameters for the fugacity-based barrier model for PCB caulk (PCB -110) 51
Table 5.6. Input parameters for the fugacity-based simulation model for PCB-containing concrete
(PCB-110) 54
XIV
-------
List of Figures
Figure E.I. AMTS Pastes (left - inactive paste; right - active paste) v
Figure E.2. Application of the inactive paste to a coating coupon vi
Figure E.3. Field caulk covered with the active paste vi
Figure E.4. Concrete pieces in a stainless steel mold for treatment by the AMTS method vii
Figure E.5. Efficiency of the AMTS method in removing PCBs from different materials viii
Figure E.6. Concentration profiles for congener #110 in the PCB-containing concrete as a function of
time and depth, C(x,t), after the AMTS treatment ix
Figure 2.1. AMTS Pastes (left - inactive paste; right - active paste) 5
Figure 2.2. Coupons for paint tests (left: alkyd; center- epoxy; right: primer) 9
Figure 2.3. Application of inactive AMTS paste to a paint coupon 9
Figure 2.4. Paint coupon covered with the Dupli-Color® Undercoat 10
Figure 2.5. Field caulk in the holder before treatment 11
Figure 2.6. Field caulk covered with AMTS active paste 11
Figure 2.7. AMTS paste covered with the Dupli-Color® Undercoat 12
Figure 2.8. Caulk being cut into slices for extraction 12
Figure 2.9. Caulk slices for extraction 13
Figure 2.10. Laboratory-made caulk (left: caulk curing in the mold; right: caulk used in tests) 14
Figure 2.11. Laboratory-made caulk after AMTS treatment 14
Figure 2.12. Concrete pieces (left: in the stainless steel mold; right: individual coupons) 15
Figure 2.13. Application of inactive AMTS paste on the laboratory-made concrete 16
Figure 2.14. Laboratory-made concrete covered with AMTS paste and coating 16
Figure 2.15. Crushed concrete ready for extraction 17
Figure 4.1. Caulk slices for extraction and analysis 36
Figure 4.2. Chromatograms (SIM mode) of compounds detected in the reduction of PCB-209 by the
AMTS method 45
Figure 5.1. Efficiency of the AMTS method used on different materials 46
Figure 5.2. Schematic diagram of effective penetration depth (Dp) 47
Figure 5.3. Schematic classroom scenario for the fugacity-based barrier model (Yuan etal., 2007) 50
Figure 5.4. Concentration profiles [Ci(x)] for congener #110 in the bottom layer (i.e., below the
effective penetration depth) of the PCB caulk after treatment by the active AMTS paste 51
Figure 5.5. Concentration profiles for congener #110 as a function of depth in the top layer of the
caulk from which the congener had been removed by the AMTS treatment 52
Figure 5.6. Average concentration of congener #110 as a function of time (C2) in the top layer of the
caulk from which the congener had been removed by treatment with AMTS 53
-------
Figure 5.7. Concentrations of congener #110 at the air exposed surface of the caulk [C2 (x=L2)] as a
function of time 53
Figure 5.8. Concentrations of congener #110 in bulk room air as a function of time due to emissions
from the treated PCB caulk assuming that the air is well mixed 54
Figure 5.9. Concentration profiles for congener #110 in the PCB-containing concrete as a function of
time (t) and depth (x), [C(x,t)], after AMTS treatment 55
Figure 5.10. Average concentration of congener #110 as a function of time (C2) in the top layer of the
concrete from which the congener had been removed by treatment with AMTS 56
Figure 5.11. Concentrations of congener #110 at the surface of the concrete that is exposed to air
[C2(x = L2)] as a function of time 56
Figure 5.12. Concentration of congener #110 in bulk room air as a function of time due to emissions
from the treated concrete assuming that the air is well mixed 57
Figure 5.13. PCB destruction mechanisms proposed in the literature 58
XVI
-------
Acronyms and Abbreviations
AMTS Activated Metal Treatment System
BTS Bimetallic Treatment System
CAS# Chemical Abstracts Service Registry Number
CFR Code of Federal Regulations
DCC daily calibration check
DoD Department of Defense
DQI data quality indicator
EH&E Environmental Health & Engineering, Inc.
EPA U. S. Environmental Protection Agency
GC gas chromatography
GC/MS gas chromatography/mass spectrometry
IAP internal audit program
IDL instrument detection limit
IS internal standard
IUPAC International Union of Pure and Applied Chemistry
MSD mass selective detector
NASA National Aeronautics and Space Administration
NIST National Institute of Standards and Technology
NRMRL National Risk Management Research Laboratory
ORD Office of Research and Development
PCB polychlorinated biphenyl
ppm parts per million
PQL practical quantification limit
QA quality assurance
QAPP quality assurance project plan
QC quality control
RCS recovery check standard
RRF relative response factor
RSD relative standard deviation
SIM selected ion monitoring
TMX tetrachloro-ra-xylene
TSCA Toxic Substances Control Act
XVll
-------
1. Introduction
1.1 Background
Polychlorinated biphenyls (PCBs), marketed as "Aroclors" and other trade names, were commonly used in
public and commercial building-construction materials from the 1950s through the late 1970s in the United
States (Herrick et al, 2004; Robson et al., 2010; Erickson & Kaley, 2011). Caulk, sealants, paint,
fluorescent light ballasts, and other products manufactured with PCBs are the primary sources of PCBs in
buildings. Because of their persistence (Brown et al., 1994), toxicological effects (ATSDR, 2000; U.S. EPA,
2008a), and the presence of high concentrations of PCBs in certain building materials in a wide range of old
buildings, PCB-containing building materials are regulated by the Code of Federal Regulations 40 CFR
§761 (U.S. EPA, 2005; 2009). Building materials with PCB concentrations greater than 50 ppm are not
authorized for use and must be disposed of as PCB bulk product waste according to §761.3 and §761.62 or,
otherwise, must be approved by EPA under a risk-based disposal according to §761.62(c). If PCBs have
contaminated either the surrounding building materials or adjacent soil, these materials are considered PCB
remediation waste, which is subject to the cleanup and disposal requirements according to §761.61.
The main goal of mitigating PCB contamination in buildings is to reduce human exposure to PCBs caused
by emissions from and contact with PCB sources. The EPA recommended public health levels of PCBs in
school indoor air are between 70 and 600 ng/m3 depending on the age groups (U.S. EPA, 2012). In general,
the existing methods for managing and remediating PCBs in buildings include: (1) physical removal of the
sources, such as bulk removal, blasting, and cutting; (2) source modification, such as chemical extraction
and chemical degradation; and (3) management solutions, such as encapsulation, physical barriers,
ventilation, air cleaning, and administrative controls (EH&E, 2012 and references therein). Mitigating PCB
contamination in buildings is complex and costly. The process generally involves: (1) obtaining site-specific
information; (2) understanding regulations, regulatory implications, and social and economic impacts; (3)
analyzing cost-effectiveness, cost-benefits, and short-term vs. long-term benefits; (4) assessing and
managing risk and establishing health and safety protocols; (5) establishing acceptance criteria and
verification methods; (6) pre-mitigation and post-mitigation monitoring; and (7) managing waste. The cost
of mitigation ranges from $0.85 to $18 per square foot of building space and from $30 to more than $100
per linear foot of caulk, depending on the mitigation methods that are used (EH&E, 2012).
One of the chemical destruction techniques for PCBs is the use of zero-valent metals for reductive
dechlorination. The technique uses the Bimetallic Treatment System (BTS) or the Activated Metal
Treatment System (AMTS) to treat the PCB-contaminated materials (Agarwal et al., 2007; Kume et al.,
2008; DeVor et al., 2008; DeVor et al., 2009; Maloney et al., 2011). Both treatment systems are paste-like
materials that contain a reducing agent, organic solvents/hydrogen donor, and thickening agent. The BTS
method uses elemental magnesium (Mg) powder coated with a small amount of a palladium (Pd) complex
as the reducing agent while the AMTS method uses only magnesium powder. According to Krug et al.
(2010), the organic solvents in the BTS or AMTS paste can penetrate into the PCB-containing material and
extract the PCBs from the material. Then the reducing agent in the paste degrades the extracted PCBs
rapidly and thoroughly. Scientists from National Aeronautics and Space Administration (NASA) and the
University of Central Florida have refined the technique and applied it to actual contaminated structures
(Krug etal., 2010).
-------
The BTM and AMTS are prepared prior to use, and are in the form of "pastes." There are two types of
pastes: the active paste and the inactive pastes. Their formulations are the same except that the latter does
not contain the reducing agent. Consequently, the extraction and degradation of PCBs can be accomplished
in either a one-step process by using the active paste or a two-step process by using the inactive paste. In the
latter case, the inactive paste is applied to the surface of the PCB-containing material to extract and remove
the PCBs from the source and then the extracted PCBs are degraded by the active metal reducing agent in a
separate container (Krug et al., 2010).
The proposed reaction mechanism for PCB dechlorination is shown below:
Mg + 2 ROH -» Mg (RO)2 + H2 (1.1)
n H2 + 2 Ci2Hio-nCln -» 2 Ci2Hi0 + n HC1 (1.2)
In the reactions, Mg is a reducing agent and ROH is the proton donor (alcohol, acid, or water). The formula
Ci2H10.nCln represents the PCB molecule, and Ci2Hio is biphenyl, anon-hazardous reaction product. The
first reaction generates hydrogen, and the second reaction transforms the PCBs into non-chlorinated
biphenyl through a dechlorination process. The method has been claimed to be applicable to painted
structures, concrete surfaces contaminated by PCB-laden transformer oil, caulks and other adhesives,
electrical equipment, soils, and other PCB-contaminated debris (Krug et al., 2010).
The BTS and AMTS methods have been applied in Department of Defense (DoD) facilities with PCB
levels as high as 11,000 mg/kg in painted surfaces of concrete and greater than 50 mg/kg in painted
surfaces of steel (Krug et al., 2010). The methods reduced the PCB concentration in treated paint to less
than 50 mg/kg after one week of treatment. During this treatment process at DoD, a protocol was
developed for formulating BTS and AMTS to enhance their applicability to various PCB-containing
materials throughout DoD facilities, and their effectiveness was demonstrated on a wide range of
structures that were contaminated with different PCB concentrations. According to the authors, the
potential advantages of the chemical degradation techniques, including both the BTS and AMTS
methods, are that it can be conducted on-site, that it is non-destructive, efficient, safe, and economically
competitive, and that it eliminates long-term liabilities. However, it does have some limitations, i.e.: (1) it
may be difficult to apply to irregular surfaces; (2) the treated surfaces may require reapplication of an
additional coating after the first application; and (3) the active paste was unable to degrade all the PCB
concentrations to less than 50 mg/g when the original PCB concentrations were very high (> 20,000 mg/kg).
Also, an extremely simplified cost model was used to conduct the cost analysis. Application of BTS/AMTS
to a concrete building prior to demolition appears to be more cost effective than to demolish the building
and dispose of the waste in a Toxic Substances Control Act (TSCA) landfill.
The BTS and AMTS methods are closely related. Only the AMTS method was evaluated in this study
because it is more recent and more cost-effective than the BTS method.
1.2 Goals and Objectives
The main goal of this study was to evaluate the effectiveness and usefulness of the AMTS method for
treating PCB sources in buildings. The study was designed to: (1) evaluate the performance of the ATMS
method for laboratory-mixed PCB-containing paint, caulk, and concrete, as well as PCB-containing caulk
-------
obtained from buildings; (2) use modeling tools to predict the PCB bleed-back (i.e., migration of PCBs from
the deep layer of the source into the treated top layer); and (3) evaluate the AMTS reaction mechanism. The
performance of the AMTS method was evaluated based on the PCB removal efficiency and the effective
penetration depth. The information and data collected from this study can be used to support the
development of risk-reduction strategies and decision-making tools regarding further recommendations for
long-term measures to eliminate available PCBs on building surfaces and to minimize exposure to protect
public health. The results should be useful to mitigation engineers, building owners and managers, decision-
makers, researchers, and the general public.
1.3 About This Report
This is the fourth report in the publication series entitled Laboratory Study of Poly chlorinated Biphenyl
(PCB) Contamination and Mitigation in Buildings, produced by EPA's Office of Research and
Development (ORD), National Risk Management Research Laboratory (NRMRL). The first report (Guo et
al., 2011) was a characterization of the primary sources of PCBs that was focused on PCB-containing
caulking materials and light ballasts. The second report (Guo et al., 2012a) focused on the transport of PCBs
from secondary sources to interior surfaces and settled dust. The third report (Guo et al., 2012b) evaluated
the encapsulation method as an interim measure for remediation of PCB contamination in buildings. This
report summarizes the research results for evaluation of a chemical removal method for PCB sources. The
study was limited to a laboratory investigation.
-------
2. Experimental Methods
2.1 Activated Metal Treatment System Pastes
2.7.7 Preparation of AMTSPastes
The exact formulas of the AMTS pastes and their preparation procedures are proprietary information that is
owned by NASA. Active AMTS paste was prepared using the chemicals shown in Table 2.1. The inactive
AMTS paste made up of all of the components shown in the table except the magnesium powder. Figure 2.1
shows the final active and inactive pastes.
Table 2.1. Components of the AMTS paste
Component
Ethanol
Glacial acetic acid
Limonene
Calcium Stearate
Carbomax PEG 8000
Glycerin (glycerol)
Sodium Polyacry late (5 '100)
Magnesium Powder
CAS#
1634-01-4
~
5989-27-5
~
~
~
9003-04-7
7439.95.4
Supplier
Pharmco-APPER
Fisher Scientific
Sigma Aldrich
ACROS
Fisher Scientific
Sigma Aldrich
Sigma Aldrich
Sigma Aldrich
-------
Figure 2.1. AMTS Pastes (left - inactive paste; right - active paste)
2.1.2 Application and Removal of the AMTS Pastes
The paste was evenly applied to the substrates using a flexible putty knife. After application, a coating
material (Dupli-Color® Professional Undercoat and Sound Eliminator, Product No. UC102) was sprayed
over the paste and the substrate. This rubberized Dupli-Color® Undercoat encapsulates the paste, thereby
preventing the evaporation of the ethanol (Krug et al., 2010). As the coating material dries, it may contract
and form cracks. The substrates were inspected several times throughout the day on which the applications
were applied, and the coating was reapplied, if necessary.
2.2 Test Procedures
The materials selected for tests are listed in Table 2.2. Mention of trade names is only for product
identification. The laboratory tests that were conducted to evaluate the AMTS method are summarized in
Table 2.3.
-------
Table 2.2. Substrate materials tested
Product Name
All-Surface Enamel Oil-Based
Primer
All-Surface Enamel Oil-Based
Gloss
Sikagard62[a]
Quikrete® Sand/Topping Mix
Industrial Polysulfide Joint
Sealant [b]
Field caulk 1
Field caulk 2
Short Name
Oil-based primer
Alkyd paint
Epoxy
Laboratory-prepared
concrete
Laboratory-prepared
caulk
N/A
N/A
Binder or Base
Material
Oil-based primer
Oil-based enamel
Solvent-free epoxy
Cement and sand
Polysulfide resin
Unknown
Unknown
Recommended Use
Wood, metal, drywall,
interior/exterior
Wood, metal, drywall,
interior/exterior
Concrete, steel
Flooring, walls, driveways
Concrete expansion joints
N/A
N/A
Recommended
Application Method
Brush, roller, airless sprayer
Brush, roller, airless sprayer
Brush, roller, airless sprayer
Trowel and level
Bulk caulk gun
N/A
N/A
^Two-part coating system.
[b] Two-part caulk.
-------
Table 2.3. Summary of AMTS evaluation tests
Test ID
AMTS-E1 (El)
AMTS-E2 (E2)
AMTS-E3 (E3)
AMTS-E4 (E4)
AMTS-E5 (E5)
AMTS-E6 (E6)
AMTS-E7 (E7)
AMTS-E8 (E8)
AMTS-E9 (E9)
AMTS-E10 (E10)
AMTS-Ell(Ell)
AMTS-E12 (E12)
AMTS-E13 (E13) [a]
Material category
Coating
Coating
Coating
Coating
Coating
Coating
Caulk
Caulk
Caulk
Caulk
Concrete
Concrete
Coating
Matrix
Oil-based primer
Oil-based primer
Alkyd paint
Alkyd paint
Epoxy
Epoxy
Field caulk 1
Field caulk 2
Lab caulk 1
Lab caulk 2
Laboratory-prepared
concrete 1
Laboratory-prepared
concrete 2
Oil-based primer
Aroclor 1254
(w/w)
0.72%
0.36%
0.72%
0.31%
0.79%
0.32%
9.62%
1.52%
0.37%
1.01%
0.30%
0.08%
0.30% [b]
Number of samples
Active paste
2
2
2
2
2
2
5
3
2
4
2
2
4
Inactive paste
2
0
1
0
1
0
1
0
0
2
2
0
2
Application
duration (days)
7
7
7
7
7
7
5,10
5
5
5,10
7
7
1,3
^Test for degradation mechanism study.
[b] Concentration of PCB-209.
-------
2.27 Tests for Coating Materials
The AMTS paste was tested on three different types of coatings, i.e., oil-based primer ("primer"), oil-based
alkyd paint ("alkyd"), and solvent-free epoxy coating ("epoxy"). Each coating was spiked with two different
concentrations of Aroclor 1254, approximately 0.7% and 0.3%, respectively. To add PCBs to the primer and
alkyd paint, the stock paint was shaken for 15 minutes in a paint shaker (Red Devil, Model #54100H). The
weight of an empty 60-mL amber jar was recorded. A calculated amount of Aroclor 1254 was added to the
jar, and the jar was re-weighed. Then, a calculated amount of paint was added to the jar. The epoxy
consisted of two components, i.e., Part A, the epoxy resin and Part B, the activator. The same process was
used to add PCBs to the epoxy, except that Aroclor 1254 was added only to Part B. The final epoxy product
was created by mixing Part A with Part B in a 1:1 ratio according to the manufacturer's instructions. Table
2.4 shows the concentrations of Aroclor 1254 in formulation of each coating material and component.
Table 2.4. Laboratory preparation of Aroclor 1254-spiked coating
Source ID
SI 5900
S16000
S16100
S16200
S16300
S16400
Test ID
AMTS-E1
AMTS-E2
AMTS-E3
AMTS-E4
AMTS-E5
AMTS-E6
Matrix
Oil-based primer
Oil-based primer
Alkyd paint
Alkyd paint
Epoxy, Part B
Epoxy, Part B
Aroclor 1254
Source (g)[a]
0.289
12.842 [b]
0.283
10.998 [c]
0.354
0.142
Coating (g)
40.038
12.814
39.137
14.853
24.865 [d]
23.635 [e]
Concentration of
Aroclor 1254 (w/w)ra
0.72%
0.36%
0.72%
0.31%
0.79%
0.32%
source was pure Aroclor 1254 unless indicated otherwise.
[b] To make 0.3% Aroclor 1254 primer, 12.842 g of S15900 was diluted with the stock primer.
[c] To make 0.3% Aroclor 1254 alkyd, 10.998 g of S16100 was diluted with stock alkyd paint.
[d] Later, Part B was mixed with 19.5117 g of Part A of the epoxy system to obtain the final concentration.
[e] Later, Part B was mixed with 19.6743 g of Part A of the epoxy stem to obtain the final concentration.
[f| Values are based on wet paint.
A 21.9 x 28.6-cm2 piece of free-film release paper (Paul N. Gardner Co., Inc., Item #PC-RP-1K) was
painted using each of the paints listed in Table 2.4. The release paper, with one side coated with silicone,
was designed for easy stripping of the dried paint film. Paint was applied using an artist's brush that was
0.75 in (1.9 cm) wide. Two coats of paint were applied, except forthe epoxy. Since the epoxy is thicker than
the alkyd and primer, only one coat was necessary to cover the piece of release paper uniformly. There was
a two-hour time interval between the two coats of the primer and at least seven hours for the alkyd. These
curing times met the minimum curing time requirements recommended by the manufacturers. Each piece of
painted release paper was cured for three months prior to use.
Test coupons (i.e., paint film on release paper) were created from the release paper using a 48-mm diameter
arch punch (C.S. Osborne & Co., Arch Punch No. 149). To determine the initial PCB concentrations in the
coupons, duplicate paint films were peeled off from the release paper and extracted for GC/MS analysis.
For treatment tests, a small piece of tape was placed on the back of each coupon, and the coupons were
attached to a 15.24 * 7.62-cm2 aluminum panel (Figure 2.2).
-------
Active paste (i.e., with magnesium powder) and inactive paste (i.e., without magnesium powder) were
applied to the appropriate coupons (Figure 2.3). The Dupli-Color® Undercoat was then sprayed over the
paste to prevent the evaporation of the ethanol (Figure 2.4). Touch-ups were made whenever necessary to
ensure that there were no cracks in the coating layer. Since significant PCB degradation was reported in the
AMTS paste even after only three days of treatment (Krug et al., 2010), we selected 5 to 10 days as
treatment time for the method evaluation for different test materials. All paints were treated for seven days.
After treatment, the Dupli-Color® Undercoat and the paste were removed from the coupons. The paint was
then peeled off the release paper and placed in a scintillation vial for extraction.
Figure 2.2. Coupons for paint tests (left: alkyd; center- epoxy; right: primer)
Figure 2.3. Application of inactive AMTS paste to a paint coupon
-------
Figure 2.4. Paint coupon covered with the Dupli-Color® Undercoat
2.2. 2 Tests for Field Caulk
Two types of interior caulk, collected from an unoccupied building scheduled for demolition, were used for
this test. The concentrations of Aroclor 1254 in these two types of caulk differed by a factor of 10.
A cube of caulk was inserted into a vise-like holder (Figure 2.5). The apparatus held the caulk firmly in
place and exposed only the top surface of the cube. Active and inactive pastes were applied to the
designated caulk pieces (Figures 2.6), and the pastes were then covered with the Dupli-Color® Undercoat
(2.7). After either five or ten days, the paste was removed from the caulk. Two slices of caulk, each with an
approximate thicknesse of 2 mm, were cut from the treated caulk sample using the apparatus shown in
Figure 2.8. These slices were cut perpendicular to the paste-treated side. Each slice was then diced (Figure
2.9) and extracted via sonication (Section 2.3).
10
-------
Figure 2.5. Field caulk in the holder before treatment
Figure 2.6. Field caulk covered with AMTS active paste
11
-------
Figure 2.7. AMTS paste covered with the Dupli-Color® Undercoat
Figure 2.8. Caulk being cut into slices for extraction
12
-------
Figure 2.9. Caulk slices for extraction
2.2.3 Tests for Laboratory-made Caulk
The caulking material used was Industrial Polysulfide Joint Sealant (THIOKOL 223 5M, PolySpec,
Houston, TX). The caulking material consisted of a resin (Part A) and a hardener (Part B).
An aliquot of Aroclor 1254 was added to Part B of the polysulfide caulk system since Part B was less
viscous and more suitable for mixing than Part A. Then, Part B was added to Part A, and they were mixed
until they were homogenized (Table 2.5). The caulk was then transferred to a 114 x 12.8 x 6.6-mm3 (L x W
x D) polyurethane foam mold (LAST-A-FOAM® FR-7100, General Plastics Manufacturing Co., Tacoma,
WA) with a spatula (Figure 2.10).
After curing for five days, the caulk was removed from the mold and placed into a vise-like apparatus
(Figure 2.11). The apparatus held the caulk firmly in place and exposed only the top surface of the piece.
Active or inactive pastes were applied to the designated caulk pieces, and the paste was then covered with
Dupli-Color® Undercoat. After seven days, the paste was removed from the caulk. Two slices of caulk, each
with an approximate thickness of 2 mm, were cut from the treated caulk, using the apparatus shown in
Figure 2.8. These slices were cut perpendicular to the paste application side. Then, each slice was diced and
extracted via sonication.
13
-------
Table 2.5. Preparation of Aroclor 1254-spiked laboratory-mixed caulk
Source ID
S 16700
S 16600
Test ID
AMTS-E9
AMTS-E10
Caulk
Part B (g)
2.211
4.405
Part A (g)
27.847
55.65
Aroclor 1254
Weight (g)
0.113
0.613
Concentration (w/w)
0.37% (3700 ppm)
1.01%(10100ppm)
Figure 2.10. Laboratory-made caulk (left: caulk curing in the mold; right: caulk used in tests)
Figure 2.11. Laboratory-made caulk after AMTS treatment
14
-------
2.2.4 Tests for Laboratory-made Concrete
Quikrete® Sand/Topping Mix (Product No. 1103) was used to create the concrete substrate that contained
Aroclor 1254 (Table 2. 6). This ready-to-use concrete mixture consisted of Portland cement and
commercial-grade sands. Since PCBs are not soluble in water, Aroclor 1254 was not added directly to the
wet concrete. Instead, the desired amount of Aroclor 1254 was measured into a 60-mL wide-mouth jar.
Concrete was weighed in an aluminum weighing dish and transferred to the jar containing the PCBs. Ten
milliliters of hexane were mixed with the components to distribute the PCBs evenly throughout the
concrete. The mixture was left standing in the fume hood for a minimum of one hour to ensure that all of the
hexane had evaporated from the concrete. Once the concrete was completely dry again, 10 mL of water was
mixed into the concrete. The concrete was poured into a cylindrical (12.5-mm diameter, 6-mm deep)
stainless steel mold to make test coupons (Figure 2.12).
Table 2.6. Preparation of Aroclor 1254-spiked laboratory-made concrete
Source ID
S 16800
S 16900
Test ID
AMTS-E11
AMTS-E12
Concrete (g)
25.006
25.006
Aroclor 1254
Weight (g)
0.074
0.021
Concentration (w/w)
0.30% (3000 ppm)
0.08% (800 ppm)
Figure 2.12. Concrete pieces (left: in the stainless steel mold; right: individual coupons)
The concrete coupons were cured for seven days. With the coupons still in the mold, either active or inactive
paste was applied to the top of the concrete and sprayed with Dupli-Color® Undercoat (Figures 2.13 and
2.14). After seven days, the paste and Dupli-Color® Undercoat were removed from the coupons. The
concrete pieces were then removed from the mold and placed in scintillation vials. The vials were shaken
vigorously in order to completely break the concrete coupon apart (Figure 2.15) for extraction.
15
-------
Figure 2.13. Application of inactive AMTS paste on the laboratory-made concrete
Figure 2.14. Laboratory-made concrete covered with AMTS paste and coating
16
-------
Figure 2.15. Crushed concrete ready for extraction
2.2 5 Reaction Mechanism Study
The proposed reaction mechanisms for the AMTS method reduce PCBs to biphenyl (Equations 1.1 and 1.2
in Section 1). The possibility that the active paste would form partially-dechlorinated homologues was
examined by applying the active paste to a primer coating that contained 0.30% PCB-209. Using PCB-209,
which contains 10 chlorine atoms, allows detection of PCB congeners with lower chlorine numbers, if they
are produced. A 21.9 x 28.6-cm2 piece of free film release paper was painted with the primer (Table 2.7)
using a 0.75 in (1.9 cm) wide artist's brush. A second coat was applied two hours later. The painted release
paper was cured for two days prior to use. Coupons of the substrate were created by using a 48-mm
diameter arch punch. A small piece of tape was placed on the back of each coupon, and the coupons were
attached to a 15.2 x 7.62-cm2 aluminum panel for treatment.
Table 2.7. Preparation of PCB-209-spiked paint for the reaction mechanism study
Source ID
S16500
Test ID
AMTS-E13
Primer (g)
44.937
Congener #209
Weight (g)
0.135
Concentration (w/w)
0.30% (3000 ppm)
Active and inactive pastes were applied to the appropriate coupons. The Dupli-Color® Undercoat was then
sprayed over the paste. After either 24 or 72 hours, the Dupli-Color®Undercoat and the paste were removed
from the coupons. The paint was peeled off the release paper and placed in a scintillation vial for extraction.
In addition, approximately 500 mg of paste from each sample coupon were transferred to a vial for
extraction.
17
-------
2.3 Sample Extraction and Analysis
2.3.1 Sample Extraction
In our previous studies (Guo et al., 2011; 2012a), we concluded that the sonication method for extraction of
caulk samples is comparable with the Soxhlet extraction method. Thus, the sonication method was selected
for all samples in the AMTS evaluation. Ten milliliters of hexane was pipetted into the 20-mL amber
scintillation vial that contained the sample. The recovery check standard (RCS, 100 (iL) was added to the
vial using a gas-tight syringe. The vial was sonicated for 30 minutes. Once the solution was cooled to room
temperature, the hexane was decanted from the substrate and placed into a new scintillation vial, which
contained approximately 500 mg of sodium sulfate (anhydrous grade or equivalent, Fisher, Pittsburgh, PA).
An aliquot (~1.5 mL) of the solution was then transferred from the amber scintillation vial into a screw-top
vial that contained 3 mL of sulfuric acid. The sample was shaken for 30 seconds and then remained still
until separation occurred. An aliquot of the top layer (hexane) was transferred into a 2-mL vial, and the
sulfuric acid layer was discarded. This sulfuric acid wash was not used on the samples for determining the
reaction products in Section 2.2.5.
Finally, 900 \\L of the acid-washed extraction solution was added to a 1-mL volumetric flask. Using a gas-
tight syringe, 100 (iL of internal standard were inserted into the flask. The solution was brought to volume
using the extract and mixed thoroughly with a vortex mixer (Fisher Scientific). A portion of the final
solution was transfered to a GC vial for analysis.
2. 3.2 Analytical Instrument and Reagents
The analytical instrument used for the quantitative analysis of PCB congeners was the Agilent 6980/5973N
GC/MS (Agilent, Santa Clara, CA) with CTC PAL Auto Sampler (LEAP Technology, Carrboro, NC).
Recovery check standards (RCSs) and internal standards (ISs) were spiked before extraction and GC
analysis (Table 2.8). The internal standard solution for spiking contained 10 ng/mL of each IS. The RCS
solution for spiking contained 5 ng/mL of each RCS. The certified PCB homologue standard (in isooctane)
was purchased from AccuStandard, Inc. (New Haven, CT). The certified 13C-labeled internal standards and
recovery check standards (in nonane) were purchased from Wellington Laboratories, Inc. (Shawnee
Mission, KS). The certified TMX standard (in acetone) was purchased from ULTRA Scientific (North
Kingstown, RI).
18
-------
Table 2.8. Chemical names and CAS Registration Numbers for the internal standards and recovery
check standards
Purpose
Internal
Standard
Recovery
Check
Standard
Short Name
13C-PCB-4
13C-PCB-52
13C-PCB-194
TMX
13C-PCB-77
13C-PCB-206
IUPAC Name
2,2'-Dichloro[13C12]biphenyl
2,2',5,5'-Tetrachloro[13C12]biphenyl
2,2',3,3',4,4',5,5',-Octachloro[13C12]biphenyl
1 ,2,3 ,5-Tetrachloro-4,6-dimethylbenzene
3,3',4,4'-Tetrachloro[13C12]biphenyl
2,2',3,3',4,4',5,5',6-Nonachloro[13C12]biphenyl
CAS#
234432-86-1
208263-80-3
208263-74-5
877-09-8
105600-23-5
208263-75-6
2.3.3 Sample Analysis for AMTS Efficiency Tests
The target compounds for analysis are listed in Table 2.9.
Table 2.9. Chemical names and CAS Registration Numbers for the PCB congeners analyzed
Congener #
17
52
66
77
101
105
110
118
154
187
Short Name
PCB-17
PCB-52
PCB-66
PCB-77
PCB-101
PCB-105
PCB-110
PCB-118
PCB-154
PCB-187
IUPAC Name
2,2',4-Trichlorobiphenyl
2,2',5,5'-Tetrachlorobiphenyl
2,3 ',4,4'-Tetrachlorobiphenyl
3,3',4,4'-Tetrachlorobiphenyl
2,2',4,5,5'-Pentachlorobiphenyl
2,3,3',4,4'-Pentachlorobiphenyl
2,3,3',4',6-Pentachlorobiphenyl
2,3',4,4',5-Pentachlorobiphenyl
2,2',4,4',5,6'-Hexachlorobiphenyl
2,2',3,4',5,5',6-Heptachlorobiphenyl
CAS#
37680-66-3
35693-99-3
32598-10-0
32598-13-3
37680-73-2
32598-14-4
38380-03-9
31508-00-6
60145-22-4
52663-68-0
The analytical method used for this project was a modification of EPA Method 8082A (U.S. EPA, 2007)
and EPA Method 1668B (U.S. EPA, 2008b). The GC/MS was calibrated with PCB congeners in the range
of 5 to 200 ng/mL. The GC/MS calibration and quantitation were performed using the relative response
factor (RRF) method. The conditions of the analytical instrument were detailed in Part 1 of this report series
(Guoetal.,2011).
2.3.4 Sample Analysis for the Reaction Mechanism Study
The analytical method used for the reaction mechanism study was a modification of EPA Method 680
(EPA, 1985). The GC/MS was calibrated using the relative response factor (RRF) method with PCB
congeners from each homologue group, except for nonachlorobiphenyl, in the range of 5 to 600 ng/mL
(Table 2.10). The RRF for decachlorobiphenyl was used as the calibration congener for both the
19
-------
nonachlorobiphenyl and decachlorobiphenyl homologue groups. The operating conditions of the instrument
are presented in Table 2.11. The mass selective detector (MSB) selected ion monitoring (SIM) parameters
were changed over time during analysis to achieve the best sensitivity, and these parameters are presented in
Table 2.12.
Table 2.10. Chemical names and CAS Registration Numbers for EPA Method 680
Congener #
1
5
29
50
87
154
188
200
209
Short Name
PCB-2
PCB-5
PCB-29
PCB-50
PCB-87
PCB-154
PCB-188
PCB-200
PCB-209
IUPAC Name
2-Chlorobiphenyl
2,3-Dichlorobiphenyl
2,4,5-Trichlorobiphenyl
2,2',4,6-Tetrachlorobiphenyl
2,2',3,4,5'-Pentachlorobiphenyl
2,2',4,4',5,6'-Hexachlorobiphenyl
2,2',3,4',5,6,6'-Heptachlorobiphenyl
2,2',3,3',4,5',6,6'-Octachlorobiphenyl
2,2',3,3'4,4',5,5',6,6'-Decachlorobiphenyl
CAS#
2051-60-7
16605-91-7
15862-07-4
62796-65-0
38380-02-8
60145-22-4
74487-85-7
52663-73-7
2052-24-3
Table 2.11. Operating conditions for the Agilent 6890/5973N GC/MS for EPA Method 680
Parameters
Injector
Injection volume
Inlet temperature
Inlet mode
Inlet Flow
Carrier gas
GC column
Oven temperature
program
Transfer line temperature
Acquisition Mode
Solvent delay
Settings
CTCPAL
luL
250 °C
Splitless
1.0 mL/min measured at 100 °C
Helium
SGE BPX5 30 m with 0.25-mm ID and 0.25-jHn film thickness
100 °C for 2 min, to 150 °C at 15 °C/mm, to 200 °C at 3 °C/mm, to 300 °C at
8 °C/min, hold for 4 min, total time = 38.50 min
280 °C
SIM
6 min
20
-------
Table 2.12. SIM acquisition parameters for the Agilent 6890/5973N GC/MS for the analysis of PCB
homologues
Homologues
Monochlorobiphenyls
Dichlorobiphenyls
Trichlorobiphenyls
Tetrachlorobiphenyls
Pentachlorobiphenyls
Hexachlorobiphenyls
Heptachlorobiphenyls
Octachlorobiphenyls
Nonachlorobiphenyls
Decachlorobiphenyl
TMX (RCS) [a]
13C-PCB-77 (RCS)
13C-PCB-206 (RCS)
13C-PCB-4 (IS) [b]
13C-PCB-52 (IS)
13C-PCB-194 (IS)
Internal
Standard
13C-PCB-4
13C-PCB-4
13C-PCB-4
13C-PCB-4
13C-PCB-52
13C-PCB-52
13C-PCB-52
13C-PCB-52
13C-PCB-194
13C-PCB-194
13C-PCB-4
13C-PCB-52
13C-PCB-194
—
—
—
Retention Time
Windows (min)
6.00-20.60
6.00 - 20.60
6.00-26.30
6.00-26.30
20.60-28.25
20.60-33.05
20.60-33.05
28.25-38.50
28.25-38.50
33.05-38.50
6.00-20.60
23.7
31.0
10.2
17.8
30.2
Primary Ion
(m/z)
188
222
256
292
326
360
394
430
464
498
244
304
476
234
304
442
Confirmation Ions
(m/z)
190
224
258
290
324
358
392
428
462
500
246
306
478
236
306
444
[a] TMX is tetrachloro-/w-xylene; RCS is recovery check standard.
[b] IS is internal standard.
2.4 Environmental Parameters
The tests were performed under laboratory conditions, which were about 23 °C and 30-40% relative
humidity
2.5 Measurement of the Dimensions of Test Coupons
Calculating the effective penetration depths for treated caulk and concrete coupons (described in Section
5.2) requires measurements of the sample dimensions. The measurements were made with AutoCAD LT
2005 (Autodesk, Inc., San Rafael, CA). The objects were photographed on a sheet of quad-rule paper with
square grids that measured 0.5 cm, and a NIST-calibrated caliper (Mitutoyo Corp. Kawasaki, Japan) was
used to obtain the measurements. The photographs were edited using Microsoft Office Picture Manager and
then copied into AutoCAD LT 2005 to determine the area. In AutoCAD, the objects were scaled to the grid
measurement of the paper, and a series of polylines was then connected to outline the object and closed to
form the shape of the object. The area and perimeter of the shape were calculated based on the initial scale
factor using the "list" command. To check the accuracy of the scale factor, a polyline drawing was made
around four squares of the paper grid to determine the area. The area measurement was deemed acceptable
if the grid area measurement, made using the NIST caliper, was within 5% of the physical measurement of
21
-------
the four square grids. This method of measuring area made it possible to determine the surface areas of the
small caulk slices and the concrete coupons.
22
-------
3. Quality Assurance and Quality Control
Quality assurance (QA) and quality control (QC) procedures were implemented in this project by following
guidelines and procedures detailed in the approved Category II Quality Assurance Project Plan (QAPP),
Poly chlorinated Biphenyls (PCBs) in Caulk: Evaluation of coatings for encapsulating building materials
contaminated'by poly'chlorinated'biphenyls (PCBs) and a NASA method for PCS destruction. Quality
control samples consisted of method blank, extraction method blank, solvent blank, and duplicates. Daily
calibration check samples were analyzed on each instrument on each day of analysis. Results of QA/QC
activities are described in the following subsections.
3.1 Data Quality Indicator Goals for Critical Measurements
Data quality indicator (DQI) goals for the measurement parameters and validation methods are listed in Part
1 of this report series (Guo et al., 2011).
3.2 GC/MS Instrument Calibration
The GC/MS calibration and quantitation of PCBs were performed by using the relative response factor
(RRF) method based on peak areas of extracted ion profiles for target analytes relative to those of the
internal standard. For PCB congeners in Aroclor 1254, the calibration standards (in hexane) were prepared
at six levels in the range from approximately 5 to 200 ng/mL. For EPA Method 680, PCB congener
concentration varied because the concentrations of the congeners in the solution that was purchased
(Accustandard, catalog # M-680A) were different. Three internal standards were added in each standard
solution for different PCB congeners. The calibration curve was obtained by injecting 1 |oL of the prepared
standards in triplicate at each concentration level. Table 3.1 summarizes all GC/MS calibrations conducted
for the project. The percent relative standard deviation (RSD) of the average RRFs meets the DQI goal of
25%.
The Internal Audit Program (IAP) was implemented to minimize any systematic errors. The IAP standards
contained three calibrated PCB congeners, and IAP standards were analyzed after the calibration was
completed. The certified IAP standards for PCB congeners from Aroclor 1254 were purchased from
ChemService (West Chester, PA) and those of the PCB homologue were purchased from Ultra Scientific
(North Kingstown, RI). The certified IAP standards were different from the supplier of the standards used
for calibration. Table 3.2 presents the results of the analysis of the IAP standards for each calibration. The
analytical results indicated that the recoveries of the lAPs ranged from 78% to 115% and that the percent
RSDs ranged from 0.01% to 2.86%. All of these results meet the criteria for IAP analysis, which are 100 ±
25% recovery with percentage RSD of triplicate analyses within 25%.
23
-------
Table 3.1. GC/MS calibration for PCB congeners [a]
For Aroclor 1254
Analytes
PCB-17
PCB-52
PCB-101
PCB-154
PCB-110
PCB-77
PCB-66
PCB-118
PCB-105
PCB-187
TMX (RCS)
13C-PCB-77 (RCS)
13C-PCB-206 (RCS)
9/29/2011
RRF
0.76
1.07
0.85
0.77
1.04
1.09
1.13
0.89
0.75
0.48
0.43
0.99
0.97
%RSD
4.71
3.32
5.47
6.23
11.4
16.5
5.01
9.04
14.6
7.88
2.22
14.9
6.25
Range (ng/mL)
PQL[bl
5.00
5.01
5.01
4.98
5.01
5.01
5.03
5.05
5.00
4.98
5.01
5.00
5.00
Hi Cal [cl
200
200
200
199
200
200
201
202
200
199
201
200
200
For PCB Homologues (Congener Groups)
Analytes
PCB-1
PCB-5
PCB-29
PCB-50
PCB-87
PCB-154
PCB-188
PCB-200
PCB-209
TMX (RCS)
13C-PCB-77 (RCS)
13C-PCB-206 (RCS)
—
11/11/2011
RRF
2.51
1.71
1.27
1.11
1.09
1.46
1.51
1.13
2.90
0.42
0.95
1.00
—
%RSD
6.76
9.90
14.4
15.6
19.8
23.8
22.7
24.9
22.8
5.15
22.8
16.1
—
11/18/2011
RRF
2.58
1.79
1.33
1.11
1.10
1.38
1.42
1.08
2.17
0.43
0.93
0.88
—
%RSD
5.12
6.88
9.90
13.0
16.3
20.9
20.9
24.2
21.2
5.68
18.8
15.9
--
Range (ng/mL)
PQL
5.00
4.97
5.02
10.0
9.98
9.95
14.9
14.9
25
5.01
5.00
5.00
—
Hi Cal [bl
200
199
201
400
399
398
595
596
500
201
200
200
—
LaJ The DQI goal for %RSD was 25%; RRF data are for n = 3.
[b] PQL is practical quantification limi. It is the lowest calibration concentration.
[c] Hi Cal is the highest calibration concentration.
24
-------
Table 3.2. IAP results for each calibration of PCB congeners w
Calibration
Date
9/29/2011
11/11/2011
11/18/2011
Analyte
PCB-52
PCB-101
PCB-77
PCB-1
PCB-5
PCB-29
PCB-50
PCB-87
PCB-154
PCB-188
PCB-200
PCB-209
PCB-1
PCB-5
PCB-29
PCB-50
PCB-87
PCB-154
PCB-188
PCB-200
PCB-209
IAP Concentration
(ng/mL)
36.0
36.0
36.0
62.5
62.5
62.5
125
125
125
188
188
313
62.5
62.5
62.5
125
125
125
188
188
313
Recovery (n = 3)
Average %
115
101
80.9
89.2
84.3
77.8
89.8
83.8
89.3
89.7
88.7
90.6
88.3
86.2
82.1
89.7
86.0
86.7
87.1
89.0
90.4
%RSD
0.01
0.78
0.95
0.34
0.68
0.72
0.76
1.84
1.60
2.86
1.61
1.40
0.56
0.84
0.86
1.31
0.88
1.66
1.28
0.35
1.52
1 The DQI goal for %RSD was 25%.
3.3 Detection Limits
After each calibration, the instrument detection limit (IDL) was determined by analyzing the lowest
calibration standard seven times and then calculating three standard deviations from the measured
concentrations of the standard. IDLs are listed in Table 3.3 for all calibrated PCB congeners.
The detection limits for the sonication method were reported in the report entitled Laboratory Study of
Poly chlorinated Biphenyl (PCB) Contamination and Mitigation in Buildings, Part II. Transport from
Primary Sources to Building Materials and Settled Dust (Guo et al., 2012a), Section 5.2.
25
-------
Table 3.3. Instrument detection limits (IDLs) for PCB congeners on GC/MS (ng/mL)
For Aroclor 1254
Analytes
PCB-17
PCB-52
PCB-101
PCB-154
PCB-110
PCB-77
PCB-66
PCB-118
PCB-105
PCB-187
TMX (RCS)
13C-PCB-77 (RCS)
13C-PCB-206 (RCS)
9/29/2011
0.46
0.32
0.80
0.51
0.74
0.55
0.71
0.67
0.83
0.78
0.34
0.57
1.40
For PCB Homologues (Congener Groups)
Analytes
PCB-1
PCB-5
PCB-29
PCB-50
PCB-87
PCB-154
PCB-188
PCB-200
PCB-209
TMX (RCS)
13C-PCB-77 (RCS)
13C-PCB-206 (RCS)
—
11/11/2011
0.18
0.33
0.19
0.35
0.42
0.26
0.42
0.47
1.29
0.77
0.15
0.57
—
11/18/2011
0.11
0.19
0.26
0.34
0.38
0.45
0.63
1.00
0.79
0.53
0.31
1.01
—
3.4 Quality Control Samples
Data quality control samples discussed here include method blank, extraction method blank, solvent blank
and duplicates.
The method blank was carried through the complete sample preparation and analytical procedure, from
AMTS paste application to GC/MS analysis, but there were no PCBs in the materials. A typical method
blank sample showed the contribution of the contamination in the whole test process. The results are
summarized in Table 3.4. The method blank samples were shared within each type of tests. No method
blank samples were collected for mechanism study test E13. Most of the method blank samples were below
PQL with the exception of tests El 1 and E12, in which PCB-101, PCB-110, PCB-118, and PCB-105 had
higher amounts per sample, most likely due to high carryover from the GC/MS analysis.
26
-------
Table 3.4. Summary of method blank for tests (ng/sample) w
Analytes
PCB-17
PCB-52
PCB-101
PCB-154
PCB-110
PCB-77
PCB-66
PCB-118
PCB-105
PCB-187
Test ID
E1&E2
Q QQ
Q ()Q
M5
4r74
2 gg
4^7-
4^65
447
4r^>
344
E3&E4
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
2^89
Q QQ
Q QQ
Q QQ
Q QQ
E5&E6
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
I gQ
Q QQ
E7&E8
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
E9
Q QQ
^*
7:62-
0.74
44r2-
Q QQ
±36
m^
4^7
Q QQ
E10
Q QQ
9r74
224
2 gQ
334
Q QQ
6^
4*6
m^
zm
E11&E12
Q QQ
34^
85.5
*§0
128
0^
44:3
151
92.8
444
1 Values in strikethrough are below the PQL; each sample was extracted in 10 mL of hexane.
The extraction method blank was generated using hexane solvent following exactly the same extraction and
GC/MS analysis procedures as the samples. The extraction method blank documented the contamination
during solvent extraction and GC/MS analysis. When the samples were extracted at the same time, only one
extraction blank was prepared. The results are presented in Table 3.5. Concentrations of PCBs in the
extraction method blank that was shared by tests El through E6 were lower than the PQL, but the RCS test
failed, so the data were not reported.
27
-------
Table 3.5. Summary of extraction method blanks for tests (ng/sample) w
Analytes
PCB-17
PCB-52
PCB-101
PCB-154
PCB-110
PCB-77
PCB-66
PCB-118
PCB-105
PCB-187
Test ID
E7&E8
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
E9 & E10
Q QQ
4r40
3r43
Q QQ
433
Q QQ
Q QQ
5r94
Q QQ
Q QQ
E11&E12
Q QQ
0^9
Q 2g
Q QQ
«6
0^
&43
I gg
«6
Q QQ
Analytes
PCB-1C1
PCB-2C1
PCB-3C1
PCB-4C1
PCB-5C1
PCB-6C1
PCB-7C1
PCB-8C1
PCB-9C1
PCB-10C1
E13
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
LaJ Values with strikethrough are below the PQL; each sample was extracted using 10 mL of hexane.
The solvent blank was used to determine whether there was any contamination in the solvent used for
sample extraction. The blank was prepared by adding 10 mL of extraction solvent to a 20-mL vial while
conducting solvent extraction. The solvent (1 mL) was then spiked with internal standards and analyzed by
GC/MS with the other samples. The results are shown in Table 3.6. No solvent blank samples were prepared
for Test Ell, Test E12, or Test E13. All data shown in Table 3.6 are below the PQL.
Table 3.6. Summary of solvent blank for tests (ng/sample) w
Analytes
PCB-17
PCB-52
PCB-101
PCB-154
PCB-110
PCB-77
PCB-66
PCB-118
PCB-105
PCB-187
Test ID
El, E2, E3, E4, E5, E6
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
E7,E8
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
Q QQ
E9, E10
Q QQ
Q QQ
4^42-
Q QQ
^m
Q 2g
4^6*
4r&7-
4r54
Q QQ
LaJ Values in strikethrough are below the PQL; each sample was extracted using 10 mL of hexane.
Duplicate samples were used to estimate the precision of the sampling and analysis methods. The DQI was
set to be RSD < 25%. Duplicate samples were prepared for each AMTS evaluation test condition. The
28
-------
results are summarized in the Result Section of this report. Table 3.7 summarizes the number of duplicate
PCB congeners analyzed in each test and shows the number of duplicate PCB congeners that failed. The
data quality was not necessarily unacceptable if the data did not meet the DQI for the tests because failing to
meet the DQI might imply that the concentrations of the PCB congeners in the test materials were uneven.
Table 3.7. Summary of duplicate samples for PCB congeners
Test ID
AMTS-E1
AMTS-E2
AMTS-E3
AMTS-E4
AMTS-E5
AMTS-E6
AMTS-E7
AMTS-E8
AMTS-E9
AMTS-E10
AMTS-E11
AMTS-E12
AMTS-E13
Matrix
Oil-based primer
Oil-based primer
Alkyd paint
Alkyd paint
Epoxy
Epoxy
Field caulk 1
Field caulk 2
Lab caulk 1
Lab caulk 2
Laboratory-prepared
concrete 1
Laboratory-prepared
concrete 2
Oil-based primer
Number of duplicate
samples
3
2
2
2
2
2
4
3
2
4
3
2
3
Number of duplicate
congeners
26
17
19
18
14
12
70
40
27
70
30
20
19
Number of duplicate
congeners failed
8
1
9
0
1
5
0
0
9
1
0
0
11
The depth of each piece of caulk and concrete was measured multiple times. The precision of the multiple
measurements is reported in the Results section of this report.
3.5 Daily Calibration Check
On each day that analyses were conducted, at least one daily calibration check (DCC) sample was analyzed
to document the performance of the instrument. DCC samples were analyzed at the beginning of and during
the analysis sequence on each day. Table 3.8 summarizes the average recovery of DCCs for the tests. The
recoveries meet the laboratory criterion of 75 to 125% recovery for acceptable GC/MS instrument
performance.
29
-------
Table 3.8. Average recoveries of DCCs for AMTS method evaluation tests
Test Type
For Aroclor
1254 w
For Method
680 [c]
DCC
Analytes
PCB-17
PCB-52
PCB-101
PCB-154
PCB-110
PCB-77
PCB-66
PCB-118
PCB-105
PCB-187
TMX (RCS)
13C-PCB-77 (RCS)
13C-PCB-206 (RCS)
PCB-1
PCB-5
PCB-29
PCB-50
PCB-87
PCB-154
PCB-188
PCB-200
PCB-209
TMX (RCS)
13C-PCB-77 (RCS)
13C-PCB-206 (RCS)
Average
%Recovery
99.7
100
96.0
96.5
102
103
96.4
98.8
103
104
98.8
104
101
104
109
109
104
104
101
100
105
94.8
103
107
98.5
SD
0.053
0.024
0.028
0.028
0.045
0.063
0.055
0.046
0.056
0.052
0.044
0.065
0.039
0.023
0.060
0.082
0.016
0.053
0.045
0.055
0.071
0.105
0.036
0.093
0.035
%RSD
5.34
2.37
2.93
2.86
4.40
6.10
5.67
4.62
5.48
5.02
4.43
6.25
3.84
2.24
5.53
7.52
1.54
5.14
4.51
5.50
6.81
11.1
3.52
8.62
3.60
NW
37
37
37
37
37
37
37
37
37
37
37
37
37
8
8
8
8
8
8
8
8
8
8
8
8
LaJ N is the number of DCCs analyzed.
[b] Modified EPA Method 8082A and EPA Method 1668B were used for analysis of congeners in Aroclor 1254
[c] Modified EPA Method 680 was used for the reaction mechanism study
3.6 Recovery Check Standards
Three recovery check standards (RCSs), i.e., TMX, 13C-PCB-77, and 13C-PCB-206, were spiked in each of
the samples before extraction to serve as the laboratory controls. When the measured concentrations of
PCBs in the sample were above the highest calibration level, which happened mostly during bulk analysis,
the extract was diluted to re-analyze the sample. In that case, recoveries of RCSs were not reported. The
30
-------
analytical results are considered acceptable if the percent recovery of laboratory controls was in the range of
60-140% for at least two of the three recovery check standards. The results of the recovery check standards
are listed in Table 3.9.
Table 3.9. Results summary of the recovery check standards
Test ID
AMTS-E1
AMTS-E2
AMTS-E3
AMTS-E4
AMTS-E5
AMTS-E6
AMTS-E7
AMTS-E8
AMTS-E9
AMTS-E10
AMTS-E11
AMTS-E12
AMTS-E13
Matrix
Oil-based primer
Oil-based primer
Alkyd paint
Alkyd paint
Epoxy
Epoxy
Field caulk l[a]
Field caulk 2[a]
Lab caulk l[a]
Lab caulk 2[a]
Laboratory-prepared
concrete 1
Laboratory-prepared
concrete 2
Oil-based primer
Number of samples
8
6
6
6
7
6
15
9
8
16
7
6
9
Number of RCS failed
0
0
0
0
o
5
i
0
0
0
0
0
0
o
5
1 two slices were extracted and analyzed for each caulk piece.
31
-------
4. Results
4.1 Terminology and Definitions
In this study, the effectiveness of the AMTS method for removing PCBs from test materials was expressed
as percent removal efficiency (or % efficiency), as defined by Equation 4.1:
I C }
"/^Efficiency =1 M x 100,
(4.1)
where C0 is the concentration of PCB congeners or Aroclor 1254 in the original, untreated test materials,
and Ci is the concentration of PCB congeners or Aroclor 1254 in the materials after the treatment by
AMTS. Both C0 and C were measured by GC/MS.
The "average efficiency for PCB congeners" is the average efficiency for the ten quantified PCB congeners
in each sample.
The concentration of Aroclor 1254 in the samples was calculated following the adaptation of Method
8082A, which calculates the Aroclor concentration in three steps. The details were described in the report
entitled Laboratory Study ofPolychlorinatedBiphenyl (PCB) Contamination and Mitigation in Buildings,
Parti. Emissions from PCB-Containing Caulking Materials and Light Ballasts (Guo et al., 2011), Section
4.1.10.
The weight fractions of selected PCB congeners, Fi; in the Aroclor 1254 calculation are summarized in
Table 4.1.
Table 4.1. Weight fractions (Fj) of selected PCB congeners used to calculate Aroclor 1254
concentration
PCB Congeners
PCB-52
PCB-101
PCB-154, PCB-1 10, PCB- 77 [a]
PCB-118
PCB-105
Wsi (ng) [bl
43.43
66.32
74.66
53.31
22.12
Ws (ng) [bl
1002
1002
1002
1002
1002
F.[b]
0.043
0.066
0.075
0.053
0.022
[a] PCB-154, PCB-110, and PCB-77 co-eluted.[bl WS1 is the content of congener i in the Aroclor standard that was
injected, Ws is the amount of the Aroclor standard that was injected, and F; is the weight fraction of congener i in the
Aroclor standard injected, given by the ratio WS1AVS.
4.2 Tests for Coating Materials
Three types of coating materials prepared in the laboratory with different concentrations of PCBs were
tested to evaluate the AMTS method. The samples were treated with the AMTS paste for seven days. The
32
-------
concentrations of PCB congeners in the samples were measured before and after the treatment. The initial
concentrations in the materials before the treatment are summarized in Table 4.2.
Table 4.2. Summary of the concentrations of PCB congener (jig/g) and Aroclor 1254 (w/w)
measured by GC/MS in the untreated test materials w
Analytes
Aroclor 1254 (% w/w) w
Aroclor 1254 (ng/g)[b]
PCB-17 (ng/g)
PCB-52(ng/g)
PCB-101 (ng/g)
PCB-154 Gig/g)
PCB-110(ng/g)
PCB-77 Gig/g)
PCB-66 (ng/g)
PCB-118(ng/g)
PCB-105(ng/g)
PCB-187 (ng/g)
Aroclor 1254 (% w/w)
Aroclor 1254 (ng/g )
Test ID and Matrix
El
Primer
0.717
7170
3.18
268
535
53.4
567
0.96
55.2
389
260
17.5
0.84
8400
E2
Primer
0.359
3590
1.41
142
221
27.1
251
0.53
24.1
198
96.9
6.66
0.37
3700
E3
Alkyd
0.718
7180
4.90
462
595
91.6
652
1.47
80.1
482
354
23.5
1.10
11000
E4
Alkyd
0.306
3060
1.85
193
316
40.6
354
0.68
29.4
262
121
10.3
0.50
5000
E5
Epoxy
0.792
7920
0.01
0.77
1.20
0.10
1.34
0.00
0.22
1.62
0.77
0.03
0.002
200
E6
Epoxy
0.324
3210
Q Q^
NR
NR
Q Q3
NR
Q QQ
0.07
NR
NR
0 pi
0.001
100
LaJ Average of duplicate samples; values in strikethrough are below the PQL; NR means "not reported" due to DQI
failure; measured concentrations are for cured coatings.
[b] Spiked concentration, calculated by wet weight.
The efficiency of the AMTS treatment was calculated using Equation (4.1), and the results are
summarized in Tables 4.3 (average percent efficiency for the target congeners) and 4.4 (percent efficiency
for Aroclor 1254). The difference between the two sets of results is rather minor.
Tests AMTS-E5 and AMTS-E6 were conducted to determine the efficiency of the AMTS on an epoxy
coating, which consisted of two components, Part A and Part B. Aroclor 1254 was added to Part B, then
Part B was mixed with Part A to form solid epoxy (after curing). When the epoxy was extracted with hexane
solvent, the extraction efficiency was very low, and the precision of the data was very poor. Most of the data
did not meet the DQI goals and, therefore, the results are not reported here.
33
-------
Table 4.3. Percent removal efficiencies for the sum of the target PCB congeners for coating materials
after the AMTS treatment[a]
Test ID
El
E2
E3
E4
Coating
Primer
Primer
Alkyd
Alkyd
Aroclor 1254
(%w/w)[bl
0.717
0.359
0.718
0.306
Sample ID [cl
AP-A
AP-B
IP-A
IP-B
AP-A
AP-B
AP-A
AP-B
IP
AP-A
AP-B
%Efficiency [dl
96.8
96.9
92.6
96.3
96.8
95.5
82.4
89.4
89.7
80.9
80.3
Average
96.9
94.5
96.2
85.9
~
80.6
SD
0.01
2.59
0.96
4.89
~
0.48
%RSD
0.01
2.74
1.00
5.70
-
0.60
[a] Treated for seven days.
[b] Spiked concentration, calculated based on the wet weight of the coating materials.
[c] AP (active paste); IP (inactive paste).
[d] Average efficiency for ten PCB congeners.
34
-------
Table 4.4 Percent removal efficiencies for Aroclor 1254 after AMTS treatment of coatings w
Test ID
El
E2
E3
E4
Coating
Primer
Primer
Alkyd
Alkyd
%Aroclor 1254 (w/w) [bl
0.717
0.359
0.718
0.306
Sample ID [cl
Untreated
AP-A
AP-B
IP-A
IP-B
Untreated
AP-A
AP-B
Untreated
AP-A
AP-B
IP
Untreated
AP-A
AP-B
Concentration (jig/g)
8345
194
202
510
267
3692
112
151
10952
1991
1159
1138
4998
903
952
%Efficiency
-
97.7
97.6
93.9
96.8
-
97.0
95.9
-
81.8
89.4
89.6
-
81.9
80.9
Average
-
97.6
95.3
~
96.4
~
85.6
~
~
81.4
SD
-
0.06
2.05
-
0.75
-
5.37
-
-
0.70
%RSD
-
0.06
2.15
~
0.77
~
6.27
~
~
0.86
aj Treated for seven days (concentrations of the untreated coating materials are the averages of duplicate samples.).
[b] Spiked concentrations, calculated by wet weight.
[c] AP (active paste); IP (inactive paste).
35
-------
4.3 Tests for Caulk Materials
The two samples of field caulk used in this study were collected from an unoccupied building scheduled for
demolition. Based on the previous measurements (Guo et al., 2011), the concentrations of Aroclor 1254 in
these samples differed by an order of magnitude. Preparation of the laboratory-made caulk is described in
Section 2.3.2. After treatment, two slices of caulk were cut from each sample (Figure 4.1). Then, they were
extracted and analyzed for PCBs separately.
Slice 1
Slice 2
Paste Application
Side
Depth
Width
Length
Figure 4.1. Caulk slices for extraction and analysis
36
-------
Table 4.5. Summary of the concentrations of PCB congeners (ug/g) and Aroclor 1254 (w/w)
measured by GC/MS in the original caulk w
Analytes
PCB-17 (ug/g)
PCB-52 (ug/g)
PCB-101 (ug/g)
PCB-154 (ug/g)
PCB-110(ug/g)
PCB-77 (ug/g)
PCB-66 (ug/g)
PCB-118(ug/g)
PCB-105 (ug/g)
PCB-187 (ug/g)
Aroclor 1254 (% w/w)
Aroclor 1254 (ug/g)
Test ID and Matrix
E7
Field Caulk 1
25.4
3224
6233
724
6692
19.0
710
5433
2440
239
9.62
96200
E8
Field Caulk 2
2.38
463
925
124
1031
3.16
121
893
424
38.1
1.52
15200
E9[b]
Lab-prepared
Caulk 1
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
E10 [bl
Lab-prepared
Caulk 2
2.02
161
237
26.1
241
0.63
27.9
206
92.3
10.1
0.38
3800
[a] Average of four slices from duplicate samples; NR means "not reported" due to DQI failure.
Ibl Data for cured laboratory-made caulk.
All pieces of caulk samples were treated with AMTS paste for five to ten days (See Section 2.2.1). The
efficiency of the AMTS treatment was calculated using Equation (4.1), and the results for the field-caulk
tests are summarized in Tables 4.6 and 4.7.
37
-------
Table 4.6. Percent removal efficiencies for PCB congeners for field caulk after AMTS treatment
Test ID
E7
E8
%Aroclor 1254
(w/w) M
9.62
1.52
Sample ID [cl
AP 5-day A [d]
AP 5-day B
AP 5-day C
AP 10-day A
AP 10-day B
IP 5-day
AP 5-day A
AP 5-day B
AP 5-day C
%Efficiency [el
16.2
23.0
20.7
6.45
12.5
19.3
18.8
18.3
35.6
Average
20.0
9.45
-
24.2
SD
0.03
0.04
-
0.10
%RSD
17.2
9.45
-
40.6
[b] Concentration in the untreated samples.
[c] AP (active paste application); IP (inactive paste application).
[d] 5-day means that that treatment time was 5 days.
[e] Average efficiency of ten target congeners.
38
-------
Table 4.7. Percent removal efficiencies for Aroclor 1254 for field caulk after AMTS treatment w
Test ID
E7
E8
%Aroclor 1254 (w/w)
9.62
1.52
Sample ID [bl
Untreated
AP 5-day A [c]
AP 5-day B
AP 5-day C
AP 10-day A
AP 10-day B
IP5-day
Untreated
AP 5-day A
AP 5-day B
AP 5-day C
Concentration (jig/g)
96200
83086
74715
74914
92817
84615
76150
15240
12376
13025
9778
%Efficiency
-
13.6
22.3
22.1
20.8
3.5
12.0
-
18.8
14.5
35.8
Average
-
19.4
7.8
~
~
23.1
SD
-
0.05
0.06
-
-
0.11
%RSD
-
25.6
77.5
-
-
48.9
LaJ Treated for five to ten days (Two slices were analyzed for each sample.). Concentrations of untreated field caulk are the averages of duplicate samples.
[b] AP (active paste application); IP (inactive paste application).
[c] 5-day means that the treatment time was 5 days.
39
-------
For each slice of the field caulk, four measurements were made to obtain the average depth of the slice. Two
slices were cut from one piece of field caulk and analyzed for PCBs. Thus, the average depth of two slices
was used to represent the depth of each piece of the field caulk. The AutoCAD measurements of the depths
of the caulk samples are summarized in Table 4.8. The difference in the depths for each piece of field caulk
was less than 8% of RSD with the exception of one case in which the difference was 21% of RSD.
However, the difference between different pieces of the caulk samples was larger because the pieces of field
caulk were not as uniform as the laboratory-made caulk samples.
Most of the data collected from tests of the laboratory-made caulk did not meet the DQI, so they were not
reported. The failed data led to a negative value of %Efficiency for the AMTS method when the measured
concentration after the treatment was higher than the measured concentration before the treatment.
Table 4.8. Summary of depth (height) measurements of pieces of field caulk
Test ID
E7
E8
Sample ID [al
AP 5-day A Slice 1 [c]
AP 5-day A Slice2
AP 5-day B Slice 1
AP 5-day B Slice 2
AP 5-day C Slice 1
AP 5-day C Slice 2
AP 10-day A Slice 1
AP 10-day A Slice 2
AP 10-day B Slice 1
AP 10-day B Slice 2
IP 5-day Slice 1
IP 5-day Slice 2
AP 5-day A Slice 1
AP 5-day A Slice 2
AP 5-day B Slice 1
AP 5-day B Slice 2
AP 5-day C Slice 1
AP 5-day C Slice 2
Depth (mm) [bl
4.56
4.74
5.27
5.03
5.41
5.65
4.80
3.54
7.29
6.55
5.67
6.31
5.11
5.51
3.28
3.38
6.26
5.69
Average (mm)
4.65
5.15
5.53
4.17
6.92
5.99
5.31
3.33
5.97
SD
0.13
0.18
0.17
0.89
0.52
0.46
0.28
0.07
0.40
%RSD
2.74
3.40
3.10
21.3
7.56
7.61
5.33
2.10
6.75
AP (active paste application); IP (inactive paste application). [ ] Average of four measurements from one slice
[c] 5-day means that the treatment time was 5 days.
40
-------
4.4 Tests for Concrete
Tests AMTS-E11 and AMTS-E12 were conducted to determine the efficiency of the AMTS method for
laboratory-made concrete. The procedure for preparing the concrete samples is described in Section 2.2.4.
The initial concentrations of the target congeners and Aroclor 1254 are summarized in Table 4.9.
Table 4.9. Summary of the concentrations of the PCB congeners and Aroclor 1254 in the untreated,
laboratory-made concrete measured by GC/MS w
Analytes
PCB-17
PCB-52
PCB-101
PCB-154
PCB-110
PCB-77
PCB-66
PCB-118
PCB-105
PCB-187
Aroclor 1254
Aroclor 1254
Units
Hg/g
Hg/g
Hg/g
Hg/g
Hg/g
Hg/g
Hg/g
Hg/g
Hg/g
Hg/g
% (w/w)
Hg/g
Test ID
Ell
1.25
106
184
21.4
196
0.46
18.1
159
72.0
0.44
0.29
2900
E12
0.25
28.7
53.2
5.87
56.3
0.12
4.90
46.1
20.3
1.68
0.082
820
1 Average of four slices from duplicate samples.
All laboratory-made concrete coupons were treated with AMTS paste for seven days. The efficiency of the
AMTS treatment was calculated using Equation (4.1), and the results are summarized in Tables 4.10 and
4.11.
41
-------
Table 4.10. Average percent removal efficiencies of the AMTS treatment for PCB
laboratory-made concrete w
congeners in
Test ID
Ell
E12
%Aroclor 1254
(w/w)M
0.30
0.08
Sample ID [cl
AP 7-day A[e]
AP 7-day B
IP 7-day A
IP 7-day B
AP 7-day A
AP 7-day B
%Efficiency [dl
55.8
53.0
41.3
52.6
31.1
47.0
Average
54.4
46.9
39.1
SD
0.02
0.08
0.11
%RSD
3.67
17.1
28.8
^ Trp^tpH fnr CPA/PTI Have ^ ^ Pnnrpntratinn ralmlatpH hv wpt wpioht LCJ AP ^rtn/p nuctp^ TP finnr'tn/p nuctp^
[d] Average efficiency often PCB congeners.[e] 7-day means that the treatment time was 7 days.
The dimensions of each of the concrete coupons were about the same because they were prepared by using
the stainless steel mold (Figure 2.12). The depths of the concrete coupons were determined by measuring
seven blank coupons using the AutoCAD software. The average of the seven measurements was 6.13 mm,
and this value was used as the depth of all of the concrete coupons. The standard deviation of these seven
measurements was 0.20 mm.
42
-------
Table 4.11. Removal efficiencies of Aroclor 1254 after AMTS treatment of laboratory-made concrete w
Test ID
Ell
E12
Aroclor 1254
%(w/w)M
0.30
0.08
Sample ID [cl
Untreated
AP 7-day A [d]
AP 7-day B
IP 7-day A
IP 7-day B
Untreated
AP 7-day A
AP 7-day B
Concentration
2879
1380
1460
1839
1472
818
594
445
Removal Efficiency
%Efficiency
~
52.1
49.3
36.1
48.9
~
27.4
45.6
Average
-
50.7
42.5
-
36.5
SD
-
0.02
0.09
-
0.13
%RSD
-
3.89
21.2
-
35.3
aj Treated for seven days (Concentrations of the original laboratory-made concrete are the averages of duplicate samples.).
[b] Concentration calculated by wet weight.
[c] AP (active paste application), IP (inactive paste application).
[d] 7-day means that the treatment time was 7 days.
43
-------
4.5 Partial Dechlorination Investigation
The potential formation of partially-dechlorinated PCB congeners after treatment by the active paste was
examined in Test AMTS-E13. The test was designed to determine whether there were any PCB-209
degradation products in the primer after the AMTS treatment. PCB congeners were quantified as
homologues. The original concentrations of the PCB homologues in the primer are listed in Table 4.12
Table 4.12. Concentrations of PCB homologues in the untreated primer measured by GC/MS w
Analytes
PCB-209 spiked
PCB-CM
PCB-C1#2
PCB-C1#3
PCB-C1#4
PCB-C1#5
PCB-C1#6
PCB-C1#7
PCB-C1#8
PCB-C1#9
PCB-C1#10 (PCB-209)
Concentration (ng/g)
2980 [b]
Q QQ
Q Qg
Q 02
0 op
0 op
OrM
5.63
0 op
29.56
1924
LaJ Test El 3; average of duplicate samples; values in strikethrough are below the PQL.
[b] Equivalent to 0.298% (w/w).
After treatment with active and inactive pastes, there was a significant reduction of PCB-209 in all samples.
The PCB-209 concentration reduced by more than 90% after 1 day of treatment with active paste and 95%
after 3 days of inactive paste treatment. However, the duplicate samples did not meet the DQI (i.e., < 25%),
so no data were reported here. The compounds detected during the test, using the primer as an example, are
presented in the chromatograms in Figure 4.2. Neither biphenyl nor major PCB homologues were detected
in the primer and the paste after the AMTS treatment. Biphenyl's absence could not be explained by the
results, be it due to biphenyl's further degradation or due to biphenyl not being a major reaction product.
More study is needed to better understand the reaction mechanism.
44
-------
$
i
5
OJ
s
ro
^
4-»
O.
">
£
s
'5
>.
c
Ol
£
Q.
m
10
a
. —
-------
5. Discussion
5.1 General Performance of the AMTS Method
The results of laboratory testing for the AMTS method, presented in Section 4, demonstrate that the method
removed PCBs from coating materials, i.e. primer and alkyd, efficiently (Figure 5.1). The removal
efficiency was greater than 80% for all samples after seven days of treatment. This result is consistent with
the results in a previous report on this issue (Krug etal., 2010). Our test results indicated that the AMTS
method is more efficient for removing PCBs from the primer than from the alkyd paint.
The percent removal efficiencies were lower for thicker materials including caulk and concrete. For the field
caulk that was tested, the AMTS method removed 12 to 36% of the PCB congeners (4 to 36% in terms of
Aroclor 1254). For the laboratory-made concrete, the removal efficiency ranged from 39 to 54% for the
PCB congeners (27 to 52% in terms of Aroclor 1254).
For each substrate, the AMTS method was evaluated at two different PCB concentration levels. The results
showed that the removal efficiencies for paint, field caulk, and laboratory-mixed concrete were not affected
significantly by the concentrations of the PCBs.
Due to a problem that occurred during the tests for epoxy paint and the laboratory-mixed caulk, we were
unable to assess the performance of the AMTS method for these materials.
100%
Figure 5.1. Efficiency of the AMTS method used on different materials
(AP-active paste, IP-inactive paste, H-high PCB concentrations, L-low PCB concentrations, FC-field caulk,
LC-laboratory-prepared concrete)
46
-------
5.2 Effective Penetration Depth
The percent removal efficiency is an appropriate indicator for the performance of the AMTS method for
coating materials, but it may be misleading for thick sources such as caulk and concrete, because the depths
to which the solvents in the paste can penetrate are limited. As a result, the percent removal efficiency is
affected by the thickness of the source. To resolve this problem, the concept of "effective penetration depth"
is introduced. The effective penetration depth is the thickness of a layer of the source material near the
treated surface in which all PCBs are removed and beyond which the PCBs remain intact (Figure 5.2).
Before treatment After treatment
Equivalent
Figure 5.2. Schematic diagram of effective penetration depth (Dp)
As an idealized case, we assume that the PCB concentration in the caulk or concrete is uniform initially and
that the AMTS treatment is 100% effective within the effective penetration depth and it has no effect
beyond the effective penetration depth. If only one side of the source is treated by the AMTS, the effective
penetration depth is defined by Equation 5.1:
(5.1)
where: Dp = effective penetration depth (mm)
L = thickness of the source (mm)
C0 = uniform concentration of PCBs in the source before treatment ((^g/g)
Ci = average concentration of PCBs in the source after treatment ((ig/g)
47
-------
Although calculating the effective penetration depth (Dp) requires the thickness of the material (L), the
effective penetration depth itself is not a function of the thickness. In fact, a major advantage of employing
this parameter as a measure of removal efficiency over other parameters, such as percent reduction, is that
the effective penetration depth is independent of the thickness of the caulk. As illustrated in Table 5.1, if two
pieces of caulk are of the same type, same length, and same width but have different thicknesses, the percent
reduction after treatment will be different (see the second row from bottom in the table), but the effective
penetration depth will be the same (the bottom row). In this report, the concept of the effective penetration
depth serves two purposes: (1) to evaluate the performance of the AMTS method for thick sources, and (2)
To use this parameter in the barrier model as described in Section 5.3. This parameter is not designed for use
in the field.
Table 5.1. Comparison of removal efficiencies expressed as percent efficiency and effective
penetration depth using hypothetical values w
Parameters
Sample thickness , L (cm)
Sample area for treatment, A (cm2)
Density of material, p (g/cm3)
Concentration of PCBs before treatment, C0 (ng/g)
Amount of PCBs before treatment, W0 (ng) [b]
Amount of PCBs removed (ng) after treatment
Amount of PCBs remaining after treatment, W (ng)
Average concentration of PCBs after treatment, Ci (ng/g) [c]
Percent removal efficiency, [(C0 - CO / C0] X100
Effective penetration depth, Dp = L (C0 - Ci ) / C0, (mm)
Sample 1
0.5
1
1
100
50
10
40
80
20
1.0
Sample 2
1.0
1
1
100
100
10
90
90
10
1.0
Two pieces of caulk of the same type and area but with different thicknesses.
[b] W0 = C0 x p x L x A.
[c]d=W/p/(LxA).
The effective penetration depths for field caulk and the laboratory-made concrete were calculated using
Equation (5.1), and the results are summarized in Tables 5.2 and 5.3.
-------
Table 5.2. Effectiveness of the AMTS method expressed as effective penetration depth for the field
caulk
Test ID
E7
E8
%Aroclor 1254
(w/w) [al
9.62
1.52
Sample ID [bl
AP 5-day A
AP 5-day B
AP 5-day C
AP 10-day A
AP 10-day B
IP 5-day
AP 5-day A
AP 5-day B
AP 5-day C
Sample
Thickness (mm)
4.65
5.15
5.53
6.92
5.99
4.17
5.31
3.33
5.97
Effective penetration depth (Dp)
Dp (mm)
0.63
1.15
1.22
0.24
0.72
0.87
1.00
0.48
2.14
Average
1.00
0.48
-
1.21
SD
0.32
0.34
-
0.85
%RSD
32.0
70.1
~
70.2
lj Concentration determined by GC/MS.LbJ AP (active paste), IP (inactive paste).
Table 5.3. The effectiveness of the AMTS method expressed as effective penetration depth for the
laboratory-made concrete
Test ID
Ell
E12
%Aroclor 1254
(w/w) [al
0.30
0.08
Sample ID[bl
AP 7-day A
AP 7-day B
IP 7-day A
IP 7-day B
AP 7-day A
AP 7-day B
Sample
Thickness (mm)
6.13
6.13
6.13
6.13
6.13
6.13
Effective penetration depth (Dp)
Dp (mm)
3.19
3.02
2.22
3.00
1.68
2.80
Average
3.11
2.61
2.24
SD
0.12
0.55
0.79
%RSD
3.89
21.2
35.3
aj AP (active paste), IP (inactive paste).
5.3 Predicting the "Bleed-back" of PCBs from Treated Sources
5.3.1 Model Description
After treatment by the AMTS method, PCBs are essentially removed from the top layer of the source, and
this layer acts as a barrier to the migration of PCBs from the untreated portion of the source, thereby
reducing the concentrations of PCBs at the surface of the source and in the room air. As a practical matter, it
is important to know whether the "bleed-back" of PCBs is a concern for the AMTS method. A fugacity-
based barrier model, developed by Yuan et al. (2007), was used to predict the bleed-back of PCBs from
treated sources. The details of the model are described in Section 5 of Part 3 of this report series, entitled
Laboratory Study of Poly chlorinated Biphenyl (PCB) Contamination and Mitigation in Buildings, Part 3.
Evaluation of the Encapsulation Method (Guo et al., 2012b). When the AMTS treatment process is applied,
49
-------
only the treated side of the source is exposed to room air, and the distance from inside the PCB source to the
surface of the top layer is defined as x (Figure 5.3).
A classroom scenario was assumed for the simulation, and the parameters are listed in Table 5.4 and Figure
5.3.
Table 5.4. Classroom scenario for the simulation using the fugacity-based barrier model
Parameter
Classroom volume
Exposed surface area of the PCB caulk
Thickness of the original PCB caulk
Exposed surface area of the PCB concrete
Thickness of the original PCB concrete
Concentration of PCB s in the inlet air
Air change flow rate
Symbol
V
A!
LTI
A2
LT2
Ym
Q
Value
100
0.5
0.02
10
0.01
0
2.78xlO"2
Units
m3
m2
m
m2
m
(ig/m3
m3/s
Notes
20mm
10 mm
= 100 m3/h
1 2
x = 0
Yin- Q y, Q
Room volume (V)
Source surface area (A)
Layer 2: Barrier Layer,C,(x,t)
Layer 1: PCB Source^fot)
A
X
Figure 5.3. Schematic classroom scenario for the fugacity-based barrier model (Yuan et al., 2007)
50
-------
5.3.2 Simulations for Treated Caulk
In addition to the average effective penetration depth obtained from the experiments (Section 5.2) for five
days of active AMTS paste treatment, other parameters used for the simulation by the model are listed in
Table 5.4 and 5.5. The values are for congener #110, the most abundant congener in Aroclor 1254. The
effective penetration depths presented in Section 5.2 were used as the thickness of the barrier layer.
Table 5.5. Input parameters for the fugacity-based barrier model for PCB caulk (PCB-110)
Parameter
Thickness of the source layer (LT1 -Dp)
Thickness of the barrier layer (Dp)
Molecular weight of the contaminant (congener #1 10)
Initial concentration in the source layer
Initial concentration in the barrier layer (AMTS treated)
Diffusion coefficient
Material/air partition coefficients
Symbol
Li
L2
MW
Coi
C02
Di,D2
K!,K2
Value
0.019
0.001
3.27xl08
8.03 xlO9
0
2.98xlO"15
3.64xl07
Units
m
m
(ig/mol
(ig/m3
(ig/m3
(m2/s)
dimensionless
Notes
19mm[a]
lmm[b]
= 327 g/mol
= 6692tig/g[c]
[d]
[d]
LaJ LT1 = Lj + L2 is the original thickness of the AMTS-treated caulk sample.LbJ Based on experimental data.
[c] Assuming the density of the caulk is 1.2 g/cm3; concentration based on experimental data for field caulk 1.
M Floto frnm d^n at ol t">M IV in tliic ^oco F> = Q
1 Data from Guo et al. (2011); in this case,
= K2.
Figure 5.4 shows the concentration profile of congener #110 in the PCB source caulk as a function of time.
The x-axis is the direction of the PCB concentration from the bottom of the PCB source caulk to the
interface of the source and the layer from which the PCBs have been removed.
6000 -
^ 4000 -
0 -
C
^sSJ
1
\
10 days
100 days
1000 days
^^5000 days
) 5 10 15 20
Distance from Bottom, x (mm)
Figure 5.4. Concentration profiles [Ci(x)j for congener #110 in the bottom layer (i.e., below the
effective penetration depth) of the PCB caulk after treatment by the active AMTS paste
(The effective penetration depth is between x = 19 mm and x = 20 mm)
51
-------
The profile of the congener #110 in the top layer of the caulk from which the PCBs have been removed is
shown in Figure 5.5. The x-axis is the direction from the source-barrier interface to the surface of the barrier.
The concentration of the PCB at x = 0 mm (i.e., the interface of the source and barrier) decreases over time.
The calculations performed by the model showed that the concentration of the PCB at x = 1 mm (i.e., at the
surface of the barrier that is exposed to air) increases at first and then decreases over time.
The average concentration of PCBs as a function of time in the top layer from which the PCBs had been
removed is shown in Figure 5.6. The contaminant accumulates in the barrier in the early days, after which
there is a slow decrease. The decrease is caused mainly by the concentration gradient formed at the interface
of the source and the barrier.
Figure 5.7 shows the concentration of PCBs at the exposed surface of the caulk from which PCBs had been
removed by the AMTS treatment. This concentration at the surface is directly linked to the contribution of
the treated source to the PCB concentration in the air. Note that, in this case, the concentration at the
exposed surface is much less than the average concentration in the barrier layer (Figure 5.6).
Figure 5.8 shows the contribution of the encapsulated source to the PCB concentration in the room air,
following the same pattern as the PCB concentration on the surface and in the barrier layer, i.e., the PCB
concentration increases first and then slowly decreases.
6000
10 days
• 100 days
•500 days
1000 days
•5000 days
0.2
0.4 0.6
y (mm)
0.8
Figure 5.5. Concentration profiles for congener #110 as a function of depth in the top layer of the
caulk from which the congener had been removed by the AMTS treatment
(The effective penetration depth is between y = 0 mm and y = 1 mm. The caulk/air interface is at y = 1 mm.)
52
-------
2000
1000
2000 3000 4000
Elapsed Time (days)
5000
6000
Figure 5.6. Average concentration of congener #110 as a function of time (C2) in the top layer of the
caulk from which the congener had been removed by treatment with AMTS
1000 2000 3000 4000
Elapsed Time (days)
5000
6000
Figure 5.7. Concentrations of congener #110 at the air exposed surface of the caulk [C2 (x=L2)] as a
function of time
53
-------
250
£ 200
M
C
jo
4-1
2
4-1
C
Ol
u
C
o
u
150
100
0 1000 2000 3000 4000 5000 6000
Elapsed Time (days)
Figure 5.8. Concentrations of congener #110 in bulk room air as a function of time due to emissions
from the treated PCB caulk assuming that the air is well mixed
5.3.3 Simulations for Treated Concrete
The parameters used for simulation of the results of treating PCB-containing concrete with the AMTS
method are listed in Tables 5.4 and 5.6.
Table 5.6. Input parameters for the fugacity-based simulation model for PCB-containing concrete
(PCB-110)
Parameter
Thickness of the source layer (LT2-Dp)
Thickness of the barrier layer (Dp)
Molecular weight of the contaminant (congener #1 10)
Initial concentration in the source layer (C0)
Initial concentration in the barrier layer (AMTS treated)
Diffusion coefficient
Partition coefficient between the source layer and air
Symbol
Li
L2
MW
CQI
C02
Dj,D2
Kj,K2
Value
0.007
0.003
3.27xl08
3.91xl08
0
4.00xlO'15
6.95 xlO7
Units
m
m
ug/mol
ug/m3
ug/m3
(m2/s)
dimensionless
Notes
7mm[a]
Snim031
= 327 g/mol
=196tig/g[c]
[d]
[d]
L»J T . =
L2 = Dp; LT2 = LI + L2 is the total thickness of the AMTS-treated concrete sample.L J Based on experimental data.
Assuming the density of the concrete is 2 g/cm3' concentration is based on experimental data.[d] Data from Guo et al.
(2012a); in this case, Dj = D2, Kj = K2
Figure 5.9 shows the PCB-110 concentration in the concrete as a function of time and depth. The blue
dotted line in the figure separates the top layer and the bottom layer.
54
-------
Profiles of the contaminant in the concrete layer from which the PCBs had been removed by treatment with
AMTS are shown in Figure 5.9. Figure 5.10 shows the average concentration of PCBs as function of time in
the concrete layer from which the PCBs had been removed by treatment with AMTS. The PCB
concentrations in the exposed surface of the concrete from which the PCBs had been removed are shown in
Figure 5.11. The PCB-110 concentration in room air from emissions from the concrete surface as a function
of time is shown in Figure 5.12. For comparison, Figures 5.10 through 5.12 also show the concentrations for
untreated concrete. The figures show that the PCB concentrations in the concrete layer from which PCBs
had been removed, on the surface of that layer, and in the room air were still increasing after 5000 days. The
difference that exists between caulk and concrete may be due to the combined effects of the partition and
diffusion coefficients and the effective penetration depth.
250
200
Bottom layer
(C01 = 196 pig/g)
Top layer
10 days
• 100 days
• 500 days
1000 days
• 5000 days
2468
Distance from Bottom, x (mm)
10
Figure 5.9. Concentration profiles for congener #110 in the PCB-containing concrete as a function
of time (t) and depth (x), [C(x,t)], after AMTS treatment
(The concrete/air interface is at y = 10 mm. C0i and C02 are the initial concentrations after treatment in the
bottom and top layers, respectively. Effective penetration depth is between x = 7 mm and x = 10 mm)
55
-------
200
1000
2000 3000 4000
Elapsed Time (days)
5000
6000
Figure 5.10. Average concentration of congener #110 as a function of time (C2) in the top layer of
the concrete from which the congener had been removed by treatment with AMTS
1000
100
•AMIS Treated
•Untreated
0.001 4
0 1000 2000 3000 4000 5000 6000
Elapsed Time (days)
Figure 5.11. Concentrations of congener #110 at the surface of the concrete that is exposed to air
[C2(x = L2)] as a function of time
56
-------
,£
c
c
o
4->
ra
4-1
01
u
c
o
u
10000
1000
100
AMIS treated
Untreated
0 1000 2000 3000 4000 5000 6000
Elapsed Time (days)
Figure 5.12. Concentration of congener #110 in bulk room air as a function of time due to emissions
from the treated concrete assuming that the air is well mixed
5.3.4 Summary of Mathematical Modeling
After treatment by the AMTS method, the top layer of a thick source becomes almost free of PCBs. This
layer acts as a barrier that impedes the migration of PCBs from the deep layers to the surface, thereby
reducing the concentration at the surface of the source and emissions to the room air. Whether this barrier
can provide adequate protection to the occupants is determined by several factors, including the initial
concentration of PCBs in the source, the environmental quality criteria that are selected, the effective
penetration depth, and the partition and diffusion coefficients for the substrate. For PCB concentrations in
room air, the area of the source and the ventilation rate are also important parameters. Because the effective
penetration depths for AMTS (See Section 5.2) is greater than the thickness of most encapsulants (Guo et
al., 2012b), the barrier created by the AMTS may perform better and lasts longer than the encapsulants for
the same source. However, the "bleed-back" of PCBs remains of concern for AMTS treated sources when
the initial PCB concentration in the source is high. Therefore, post-treatment environmental monitoring,
such as periodical air and wipe sampling, is necessary. The dependence of PCB bleed-back on the initial
concentration and the thickness of the encapsulant is described in Guo et al. (2012b).
5.3.5 Limitations of Mathematical Modeling
The simulation results presented above should be considered semi-quantitative because the partition and
diffusion coefficients used were rough estimates and because the predictions of the model have not been
verified. In addition, it was assumed that the PCB concentration in the source was uniform initially, which is
an oversimplification of the real-world situation.
57
-------
5.4 Reaction Mechanisms
According to the literature (DeVor et al, 2009 and Krug et al, 2010), the mechanisms for PCB destruction
by the AMTS method include the generation of hydrogen and dechlorination (Figure 5.13). A key step of
these reactions is the replacement of a chlorine atom in the PCB molecule by a hydrogen atom.
Generation of Hydrogen
Magnesium (Reducing agent)
Mg + 2 ROM ^ (RO)2Mg + H2
\
Proton Donor (Alcohol, acid, or water)
Dechlorination
+ n H •* 2 CH + n HCI
1210
PCB
Biphenyl
Figure 5.13. PCB destruction mechanisms proposed in the literature
If the reactions in Figure 5.13 represent the major mechanisms for PCB destruction, then there is a chance
for incomplete chlorine stripping (i.e., generation of PCB congeners with fewer chlorine atoms). According
to the reaction mechanism in Figure 5.13, the disappearance of one mole of PCBs yields one mole of
biphenyl if the PCBs are completely dechlorinated. Thus, ascertaining the quantity of biphenyl that is
formed and determining whether less-chlorinated PCB congeners are present can shed light on the reaction
mechanisms associated with the AMTS process.
In this study, PCB-209 was spiked into a primer and treated with the AMTS paste to test the above
hypotheses. Significant reduction of PCB-209 was observed (> 90%) after 24 hours of treatment. However,
neither biphenyl nor any other major, less-chlorinated PCB congeners were detected. Further investigation is
needed to determine whether other reaction byproducts are formed in the process and whether any of the
byproducts are of environmental concerns.
58
-------
5.5 Organic Solvents
The AMTS pastes contain ethanol, and the Dupli-Color® Undercoat (used as an encapsulant for the paste)
contains toluene, mineral spirits, and asphalt. The potential effects of these chemicals on indoor air quality
and their potential for posing a fire hazard should be evaluated when the pastes are applied inside buildings
in large quantities.
5.6 Study Limitations
This study was conducted in a relatively short period of time, about 3 months, and only 13 tests were
conducted with four types of substrates, i.e., coating materials, field caulk, laboratory-made caulk, and
laboratory-made concrete.
Among the laboratory-made PCB sources that we tested, the data for the epoxy coating and caulk did not
meet the data quality goals and, thus, were not reported.
The PCB-containing concrete samples were made in the laboratory with the PCBs mixed in the concrete
during concrete preparation. Mixing PCBs into concrete during initial preparation of the concrete is not
representative of the conditions under which PCBs are found in the concrete or masonry of actual buildings.
The parameters used for the barrier models, including solid/air partition coefficients and solid-phase
diffusion coefficients, are rough estimates based on previous experiments (Guo et al., 2011; 2012a). The
effective penetration depths used in the simulations were the average of experimental data that contained
considerable uncertainty (See Tables 5.2 and 5.3). The fact that the precision results did not meet the DQI
for the tests does not necessarily mean that the data quality was unacceptable. Rather, it might imply that the
concentrations of PCB congeners in the test materials were uneven.
The study of the reaction mechanism for the AMTS was exploratory, and did not provide evidence for the
formation of biphenyl or less chlorinated PCBs. Further study is needed to understand the reaction
mechanism better.
There are several important areas that this study did not investigate. They include: (1) the effectiveness of
multiple treatments; (2) the properties and conditions of the source materials after being treated with AMTS;
(3) the PCB residual concentration in the active paste after treatment.
This study evaluated the AMTS technology that was available in early 2011. Since then, the developer of
this method has conducted research aimed to improve the performance of the method for contaminated
masonry materials by modifying the formulation and application procedure (Quinn, 2012). This study did
not evaluate these new methods.
59
-------
6. Conclusions
Active and inactive AMTS pastes were tested in the laboratory with four types of PCB-containing
substrates, i.e., coating materials, field caulk, laboratory-made caulk, and laboratory-made concrete. The
performance of the AMTS method was evaluated based on the PCB removal efficiency and the effective
penetration depth [Sections 2 and 4].
The test results showed that the AMTS method is highly efficient (> 80 %) for removing PCBs from the
coating materials we tested (i.e., the primer and alkyd paint) even for initial PCB concentration levels of
several thousand parts per million. The AMTS method was more efficient in removing PCBs from the
primer than from the alkyd paint. Our findings were consistent with the results reported from an earlier
study (Krug et al., 2010) [Section 4.2].
The high removal efficiencies obtained with the primer and the alkyd paint were associated with relatively
thin sources. For thick sources, i.e., the field caulk and the laboratory-made concrete, the AMTS method
was less effective in its ability to remove PCBs because of its limited effective penetration depth. The
treatment removed 12 to 36% of the target congeners (4 to 36% in term of Aroclor 1254) in the field caulk
that contained 15200 and 96200 ppm Aroclor 1254, and it removed 39 to 54% of the target congeners (27 to
52% in term of Aroclor 1254) in the laboratory-made concrete that contained 820 and 2900 ppm Aroclor
1254 [Sections 4.3 and 4.4].
The AMTS method was evaluated for two different PCB concentration levels in the four substrates. The
results showed that the PCB removal efficiencies for the two paints, the field caulk, and the laboratory-made
concrete were affected very little by the initial PCB concentrations, i.e., whether the concentrations were
high or low [Sections 4.2, 4.3, and 4.4].
The effective penetration depth, which is independent of the thicknesses of the AMTS-treated materials, was
determined for the field caulk and the laboratory-made concrete. The average effective penetration depth for
the field caulk was about 1 mm with 50% RSD, and the average effective penetration depth for the
laboratory-made concrete was about 3 mm with 16% RSD [Section 5.2].
After the AMTS treatment, the top layer of a thick source from which the PCBs had been removed acted as
a temporary barrier to the migration of the PCBs from the untreated portion of the source substrate to the
treated surface. The fugacity-based barrier model is a useful tool for predicting the "bleed-back" of PCBs,
which can be a limiting factor for treating thick sources that contain high concentrations of PCBs. Therefore,
post-treatment environmental monitoring is necessary [Section 5.3].
While PCB reduction via the substitution of hydrogen for one of the chlorine atoms in the PCB molecule
has been depicted as the major mechanism for the AMTS method, the possibility exists that other reactions,
as yet unidentified, may also occur. Further work will be required to confirm that no hazardous byproducts
(e.g., partially-dechlorinated PCBs) are generated during application of the AMTS method [Sections 4.5 and
5.4].
Overall, the AMTS method has the potential to become a viable method for mitigating PCB contamination
in buildings. This method is promising for treating contaminated masonry materials near the expansion
joints after the caulking material is removed because the AMTS can treat sources that contain several
60
-------
thousand ppm PCBs, for which the encapsulation method may not be effective. However, more research is
needed because the current method has limited effective penetration depths for thick sources, including
masonry materials. In addition, the reaction mechanism needs to be verified. The potential effect of the
method on indoor air quality during the treatment also needs further evaluation.
61
-------
7. Recommendations
The authors recommend the following topics for future research.
Evaluate the effectiveness of the latest AMTS method for treating masonry materials. One of the major tasks
in mitigating PCB contamination in buildings is to deal with the contaminated masonry materials adjacent to
PCB caulking. This report showed that the AMTS method had limited effective penetration depth for
concrete after a single treatment. Efforts have been made by the developer of the AMTS method to improve
its performance for PCB-contaminated masonry materials. This improved method should be evaluated
because, if the effective penetration depth can be significantly increased, the AMTS method may play a role
in treating contaminated masonry materials in buildings.
Verify the reaction mechanisms. The reaction mechanisms proposed by the developer of the AMTS method
(Figure 5.13) produce one mole of biphenyl for each mole of PCBs destroyed. In this study, exploratory
experiments were conducted to determine the formation of biphenyl. However, the results were negative
(Section 4.5). The authors recommend verifying the reaction mechanism by a combination of GC/MS
analyses and mass balance for chlorine. The GC/MS analyses help identify specific by-products formed.
According to Figure 5.13, complete dechlorination of one mole of a PCB congener that contains n chlorine
atoms will yield n moles of HC1. Thus, a chlorine mass balance may provide a definitive answer to the
question.
Evaluate the advantages and disadvantages of using active and inactive pastes. As described in Sections 1.1
and 2.1, the active AMTS paste extracts and degrades PCBs in a single step while the inactive paste in two
separate steps. It has been reported that, unlike the inactive paste in which the extracted PCBs react with the
reducing agent in a container after the treatment, the active paste cannot degrade the extracted PCBs to
below 50 mg/kg if the PCB concentration in the source is very high (>20,000 mg/kg) (Krug et. al., 2010).
The advantages and disadvantages of using the two types of pastes should be further evaluated under
realistic application conditions.
Select the cover paint that is suited for indoor uses. The current AMTS method uses a rubberized coating
material to encapsulate the paste applied to the substrate to prevent the solvents from evaporation. This
coating material is not intended for indoor applications and should be replaced by one that has minimum
effects on indoor air quality.
Conduct field measurements. Future evaluation in the laboratory should be complemented by field
measurements in PCB-contaminated buildings. Effects on the physical properties of the substrates after
multiple treatments should be examined. The evaluation should include analysis of PCB residual
concentrations in the active paste after treatment because it is a limiting factor for treating sources with high
PCB content.
Evaluate other chemical methods. Chemical methods other than the AMTS for on-site treatment of PCB-
contaminated materials need to be identified and evaluated.
62
-------
Acknowledgments
The authors thank Dr. Jacqueline Quinn and her group at the National Aeronautics and Space
Administration (NASA) for providing the formulation and technical advice on the AMTS method; Drs. John
Little and Zhe Liu of Virginia Polytechnic Institute and State University for providing the MATLAB code
for the fugacity-based barrier model; Russell Logan and Robert H. Pope of ARCADIS for laboratory
support; Robert Wright of EPA's National Risk Management Research Laboratory and Joan Bursey of
EPA's National Homeland Security Research Center for QA support.
63
-------
References
(Website accessibilities were last verified on February 28, 2012)
Agarwal, S., Al-abed, S. R., and Dionysiou, D. D. (2007). Enhanced corrosion-based Pd/Mg bimetallic
system for dechlorination of PCBs, Environmental Science & Technology, 41, 3722-3727.
ATSDR (2000), Toxicological profile for polychlorinated biphenyls (PCBs), Agency for Toxic Substances
and Disease Registry. http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=142&tid=26
Brown, J. F., Jr., Lawton, R. W., and Morgan, C. B. (1994). PCB metabolism, persistence, and health effects
after occupational exposure: implications for risk assessment, Chemosphere, 29, 2287-2294.
DeVor, R., Carvalho-Knighton, K., Aitken, B., Maloney, P., Holland, E., Talalaj, L., Fidler, R., Elsheimer,
S., Clausen, C. A, and Geiger, C. L. (2008). Dechlorination comparison of mono-substituted PCBs with
Mg/Pd in different solvent systems, Chemosphere, 73, 896-900.
DeVor, R., Carvalho-Knighton, K., Aitken, B., Maloney, P., Holland, E., Talalaj, L., Elsheimer, S., Clausen,
C. A, and Geiger, C. L. (2009). Mechanism of the degradation of individual PCB congeners using
mechanically alloyed Mg/Pd in methanol, Chemosphere, 76, 761-766.
EH&E (2012). Literature review of mitigation methods for PCBs in buildings, EPA/600/R-12/034, prepared
for U.S. EPA by Environmental Health & Engineering, Inc.
Erickson, M. D. and Kaley II, R. G. (2011). Applications of polychlorinated biphenyls, Environmental
Science and Pollution Research, 18(2): 135-151.
Guo, Z., Liu, X., Krebs, A. K., Stinson, R. A., Nardin, J. A., Pope, H. R, and Roache, N. F. (2011).
Laboratory study of polychlorinated biphenyl (PCB) contamination and mitigation in buildings, Part 1.
Characterization of primary sources, EPA/600/R-11/156, U.S. EPA, Office of Research and Development,
National Risk Management Research Laboratory, 127 pp.
Guo, Z., Liu, X., Krebs, A. K., Greenwell, D. J., Roache, N. F., Stinson, R. A., Nardin, J. A., and Pope, H.
R. (2012a). Laboratory study of polychlorinated biphenyl (PCB) contamination and mitigation in buildings,
Part 2. Transport from primary sources to building materials and settled dust, EPA/600/R-11/156A, U.S.
EPA, Office of Research and Development, National Risk Management Research Laboratory, 166 pp.
Guo, Z., Liu, X., Krebs, A. K., Roache, N. F., Stinson, R. A., Nardin, J. A., Pope, H. R, and Logan, R. D.
(2012b). Laboratory study of polychlorinated biphenyl (PCB) contamination and mitigation in buildings,
Part 3. Evaluation of the encapsulation method, EPA/600/R-11/156B, U.S. EPA, Office of Research and
Development, National Risk Management Research Laboratory, 108 pp.
Herrick, R. F., McClean, M. D., Meeker, J. D., Baxter, L. K., and Weymouth, G. A. (2004). An
unrecognized source of PCB contamination in schools and other buildings, Environmental Health
Perspectives, 112(10), 1051-1053.
64
-------
Krug, T., O'Hara, S., Quinn, J., Clausen, C., Geiger, C., Captain, J., and Ruiz, N. (2010). Application of a
bimetallic treatment system (BTS) for PCB removal from older structures on DoD facilities, Naval Facilities
Engineering Command Engineering Service Center (NAFAC ESC), Technical Report, TR-2352-ENV.
https://portal.navfac.navy.mil/portal/page/portal/docs/doc_store jub/tr-2352-env.pdf
Kume, A., Monguchi, Y., Hattori, K., Nagase, H., and Sajiki, H. (2008). Pd/C-catalyzed practical
degradation of PCBs at room temperature, Applied Catalysis B: Environmental, 81, 274-282.
Maloney, P., DeVor, R., Novaes-Card, S., Saitta, E., Quinn, J., Caulsen, C., and Geiger, C. (2011).
Dechlorination of polychlorinated biphenyls using magnesium and acidified alcohols, Journal of Hazardous
Materials, 187, 235-240.
Quinn, J. W. (2012). Personal Communications (April 24).
Robson, M., Melymuk, L., Csiszar, S. A., Giang, A., Diamond, M. L., and Helm, P. A. (2010). Continuing
sources of PCBs: the significance of building sealants, Environmental International, 36, 506-513.
U.S. EPA (1985). EPA Method 680, Determination of pesticides And PCBs in water and soil sediment by
gas chromatography mass spectrometry.
http://www.caslab.com/EPA-Methods/PDF/EPA-Method-680.pdf
U.S. EPA (2005). Code of Federal Regulations, Title 40, Part 761, Polychorinated biphenyls (PCBs)
manufacturing processing, distribution in commerce, and use prohibitions (2005 Edition).
http://www.epa.gov/epawaste/hazard/tsd/pcbs/pubs/2005-761 .pdf
U.S. EPA (2007). EPA Method 8082A, Polychlorinated biphenyls (PCBs) by gas chromatography.
http://www.epa.gov/waste/hazard/testmethods/sw846/pdfs/8082a.pdf
U.S. EPA (2008a). Health Effects of PCBs. http://www.epa.gov/epawaste/hazard/tsd/pcbs/pubs/effects.htm
U.S. EPA (2008b). EPA Method 1668B, Chlorinated biphenyl congeners in water, soil, sediment, biosolids
and tissue by high resolution gas chromatography/high resolution mass spectrometry (HRGC/HRMS).
http: //epa.gov/waterscience/methods/method/files/1668 .pdf
U.S. EPA (2009). Current best practices for PCBs in caulk fact sheet — removal and clean-up of PCBs in
caulk and PCB-contaminated soil and building material, http://www.epa.gov/pcbsincaulk/caulkremoval.htm
Yuan, H., Little, J. C., Marand, E., and Liu, Z. (2007). Using fugacity to predict volatile emissions from
layered materials with a clay/polymer diffusion barrier, Atmospheric Environment, 41, 9300-9308.
65
------- |