-------
qualitatively and quantitatively reproducible patterns of Aroclors 1242
and 1254, as depicted in Figure 6. The estimated average concentration
of total PCBs in the replicate samples (Table 10) was 200 jig/g (ppm).
Quantitation of total PCBs was referenced to total area counts of the
chromatogram between retention times of 11 and 22 min. Total peak area
was converted to extract concentration using a response factor derived
from a five point calibration curve. Calibration standards consisted of
equal parts by mass of Aroclors 1242 and 1254 serially diluted in
hexane. Extract concentration was translated into sample concentration
using the following equation:
where Cs is the sample concentration, ng/g, Cex is the extract
concentration, ng/mL, F1 is a correction factor equalling 10, Vex is the
final extract volume, ml, and Ms is the mass of the sample, g.
The PCB level reported in Table 10 exceeds the value reported for this
site one year after emergency response remediation (<1 ng/g) by two
orders of magnitude. The level even exceeds the maximum value reported
(157 ng/g) in the initial TSCA investigation prior to any stabilization
and compositing of the lagoon sediments. What can account for these
discrepancies?
Analysis of PCBs in contaminated soils, sludges and sediments is not a
straightforward procedure. Extreme care must be taken to ensure that
the PCBs are exhaustively extracted from the sample matrix and that,
once extracted, levels of interferences such as oils, sulfur and organic
contaminants are sufficiently reduced by cleanup steps to allow for an
F 1s a unitless, composite correction factor accounting for fraction of the Initial extract which was cleaned
up and analyzed (100 out of 500 ml), as well as for the loss of extract (5 out of 10 mL) that results from
loading of the GPC loop Injector.
35
-------
SAMPLE, REP 1
SAMPLE, REP 2
SAMPLE REP 3
SAMPLE REP 4
STD - AROCLORS 1242 & 1254
10 12 14 16 18 20 22 24 26
TIME, min
Figure 6. GC/ECD chromatograms of replicate sample extracts of
stabilized sludge from the Westville site and a standard
of combined Aroclors 1242 and 1254.
36
-------
unambiguous identification and accurate quantitation of characteristic
Aroclor chromatographic patterns. Even with relatively clean samples,
established chromatographic methods for PCBs have historically evidenced
poor reproducibility in multilaboratory applications (20).
TABLE 10. ANALYSIS OF PCBs IN STABILIZED LAGOON SLUDGE
FROM THE WESTVILLE EMERGENCY RESPONSE SITE
Replicate
Samples
1
2
3
4
Concentration
as Aroclors
mean
std. dev.
RSD (%)
of PCBs (ug/g)
1242 and 1254
202
197
190
210
199.8
8.4
4.2
In light of these circumstances the discrepancies in reported PCB levels
noted above are not surprising, particularly when one considers that the
samples in question are oil lagoon sludges replete with several types of
interferences (oil, sulfur, other semivolatile organics). To illustrate
the magnitude of the analytical problem, consider a GC/MS chromatogram
of the identical lagoon sludge sample shown in Figure 8. The sample was
processed through both florisil and GPC cleanups. Gravimetric residue
analysis of the sample extract evidenced a 94% cleanup efficiency.
Nonetheless, the hopelessly intractable GC/MS chromatogram in Figure 8
was obtained. An attempt to measure PCBs in this sample using approved
GC/MS methods (21) would more than likely have resulted in a reported
zero or "not detected" concentration, because aliphatic interferences
have completely obliterated the characteristic Aroclor peak patterns and
PCB mass spectra.
37
-------
o;
i i i ' i * i i iiiiii ' i i i i i i
16 20 24 28
Figure 7. Total ion chromatogram of stabilized sludge extract.
These interference problems are not entirely obviated by application of
electron capture detection. Characteristically high levels of sulfur in
these oil-bearing sludges can impair Aroclor pattern recognition just as
effectively as aliphatic interferences in GC/MS. GPC cleanup, designed
to remove both organic constituents and sulfur, is not 100% effective,
particularly when levels of these interferences exceed the capacity of
the gel. This laboratory, for example, has encountered municipal
sludges and marine sediments where the method-recommended sample size
had to be adjusted downward to accommodate GPC column capacity, or which
otherwise required an additional extract cleanup with activated copper.
Problems related to instrument selection, chromatographic interpretation
and extract cleanup notwithstanding, the difficulty of simply extracting
PCBs from porous soil and sediments has been well documented (22, 23).
The most plausible mechanism accounting for the poor extractability of
PCBs in soils by conventional techniques (as well as for the persistence
38
-------
of PCBs in the environment) is encapsulation by water (24). Several
analytical strategies have been routinely employed to recover PCBs from
moist porous soil matrices. One approach is to use a hydrophilic
partitioning solvent, such as methanol, to mobilize the PCBs. This
process can be enhanced by sonication. The PCBs are thus made more
accessible to the extracting solvent, in this case, methylene chloride.
A second approach involves mineral acid digestion, where the focus is
more on breaking down the porous fabric of the matrix itself. This
digestion can employ HC1, H2S04 or HF. The latter, while quite
effective, is somewhat cumbersome and precludes the use of conventional
glassware. Sulfuric acid digestion is contraindicated for stabilized
soils since this process results in the formation of insoluble calcium
sulfate (gypsum) which could also encapsulate the PCBs.
To determine whether the Westville lagoon sludge presented PCB
extraction problems which demanded use of the above procedures, the
sample was reanalyzed using Soxhlet extraction without acid digestion.
All other method conditions, including cleanup, were identical to those
employed in the prior analysis (Table 10). Soxhlet extraction was
performed in accordance with Method 3540 (25). Each of two 20-g
portions of the sample homogenate was treated with sufficient anhydrous
sodium sulfate to dewater the matrix, placed in glass thimbles, and
extracted for 24 h with methylene chloride. The resultant extracts were
diluted to 500 ml. A 10 ml portion of each extract was then processed
through florisil and GPC cleanups, solvent exchanged to hexane, and
analyzed by GC/ECD.
Results of this analysis demonstrated total PCB levels of 222 and 218
ng/g for the duplicates, as Aroclors 1242 and 1254. This agrees within
10% with results reported in Table 10. Thus, the lagoon sludge in
question did not impose any significant extraction problems because a
conventional, routinely applied extraction procedure afforded a recovery
comparable to that of a more stringent technique.
39
-------
The foregoing discussion offers several analytical scenarios to account
for the failure of laboratories participating in the Westville site
evaluation to adequately detect PCBs in stabilized oil lagoon sludge.
One or more of these circumstances could explain the questionably low or
null findings for PCBs reported after the material was solidified.
Pinpointing the exact cause(s) of these discrepancies, however, is
problematical since documentation of sampling and analysis procedures
employed in these site evaluations has not been available. In fact, it
is possible that either the archived sample available for our analysis
or samples furnished to laboratories who performed the original analyses
were not representative of the bulk material. In spite of these open
questions, it appears likely that the incipient evidence supporting
claims of PCB dechlorination in the field by in-situ lime treatment is
based on erroneous data.
40
-------
CONCLUSIONS
Treatment of PCB-fortified synthetic soil with quicklime in open vessels
resulted in large losses of all three PCB congeners. The bulk of these
losses, 60 to 80% of starting concentration occurred in the first five
hours of treatment and immediately following lime slaking and sample
heating steps. Subsequent losses of PCBs were less pronounced; about
10% to 30% of the original spiking levels over the balance of the 72 h
treatment period. The copious excursion of steam and matrix
particulates during the slaking process, the evaporative losses of PCB
congeners over time evidenced in untreated samples, and the absence of
significant levels of PCB decomposition products all support the
hypothesis that volatilization, rather than decomposition, accounts for
the preponderance of PCB losses observed. Furthermore, the
concentration versus time-of-treatment curves of the congeners agreed
reasonably well with Thibodeaux's model for PCB volatilization in soil,
when numerical constants, variables and assumptions consistent with
these experiments were used.
Minimal evidence of PCB dechlorination was observed. Monochlorobi-
phenyl, trichlorobiphenyl, and hydroxy and methoxy-substituted
chlorobiphenyls, were found sporadically and in relatively small
abundance. The presence of decomposition products did not appear to be
a function of maximum slaking temperature or treatment time. No
products of phenyl-phenyl bond cleavage were observed. Most of the
products were consistent with mono-substitution.
An archived field sample from the Westville, IN site analyzed during
this study did not support previous claims of PCB decomposition by in-
situ lime treatment. The Westville sludge, which was reported to
contain post-remediation, PCB levels less than 1 ppm and which
41
-------
accordingly provided a catalyst for this and other research into
lime-promoted destruction of PCBs, was found in this study to contain an
aggregate Aroclor 1242 and 1254 level of 200 ppm.
The destruction of PCBs by application of quicklime to contaminated
soil, sediment or sludge has thus not been demonstrated, either by
controlled benchtop experiments or by retrospective analysis of a sample
from a remediation site where the process was applied. Evidence of PCB
volatilization suggests that use of reactive quicklime as an in-situ
treatment may even be contraindicated due to the potential for migration
of PCBs as vapor or airborne particulates.
The presence of small amounts of partially dechlorinated PCBs after
quicklime treatment may warrant further investigation to obtain a better
understanding of PCB reactivity. However, the low product yields
observed upon simple addition of quicklime and water suggest that any
process based on CaO will require other reagents, catalysts, or more
extreme reaction conditions. In-situ treatment processes would be
constrained by PCB volatilization and the possible formation of toxic
reaction products.
Further work is needed to determine the exact effects of bulking
processes employed to temporarily stabilize PCB-containing wastes in the
field. CaO-containing materials are often used to improve the handling
characteristics of such wastes. We are currently constructing a pilot-
scale apparatus that will allow measurement of vapor and particulate
phase losses of PCBs under conditions likely to be encountered in field
applications. The results of these studies will determine the direction
of our future quicklime research.
42
-------
REFERENCES
1. U.S.EPA, Superfund Records of Decision Update. Office of Solid
Waste and Emergency Response, Intermittent Bulletin, vol. 6, no.l,
Publication Number 9200.5-2161, 1990.
2. Safe, S., Polychlorinated Biphenyls (PCBs): Mutagenicity and
Carcinogenicity. Mutation Research, v. 220, pp. 31-47, 1989.
3. U.S. EPA, Guidance on Remedial Actions for Superfund Sites with
PCB Contamination. EPA/540/G-90/007, Office of Emergency and
Remedial Response, p. 144,1990.
4. Brunelle, D. J., and Singleton, D. A., Chemical Reaction of
Polychlorinated Biphenyls on Soils with Poly(Ethylene Glycol)/KOH.
Chemosphere, 14, 173-181, 1985.
5. Jordan, 0. D., System and Apparatus for the Continuous Destruction
and Removal of Polychlorinated Biphenyls from Fluids. U.S. Patent
No. 4,4340,401, 1982.
6. Brown, J. F., and Lynch, M. E., Method for Removing
Polychlorinated Biphenyls from Transformer Oil. U.S. Patent No.
4,377,471, 1983.
7. Norman, 0. L., and Handler, L. H., Method of Destruction of
Polychlorinated Biphenyls. U.S. Patent No. 4,379,746, 1983.
8. Kitchens, J. F., Jones, W. E., Anspach, G. L., and Schubert, D.
C., Light-Activated Reduction of Chemicals for Destruction of PCBs
in Oil and Soil. In Detoxification of Hazardous Waste, ed. by
Exner, J. H., Ann Arbor Science, Ann Arbor, Michigan, 1981.
9. Kalmaz, E. V., Craig, R. B., and Zimmerman, G. W., Kinetics Model
and Simulation of Concentration Variations of Species of PCBs
Involved in Photochemical Transformation. In Detoxification of
Hazardous Waste, ed. by Exner, J. H., Ann Arbor Science, Ann
Arbor, Michigan, 1981.
10. Kornel, A., Rogers, C. J., and Sparks, H. L., Method for the
Destruction of Halogenated Organic Compounds in a Contaminated
Medium. U.S. Patent Application No. 350,425, May 11, 1989.
43
-------
11. Payne, J., Boelsing, F., Habekost, A., Hirschfeld, G., and Birke,
V., Complete Ambient-Temperature Dehalogenation of PCB in
Comtaminated Soil Using Hydrophobic Lime and Other Reagents, EPRI
1991 PCB Seminar, October 8-11, 1991, Baltimore, MA (in press).
12. Boyd, S. A., Mortland, M. M., and Chiou, C. T., Sorption
Characteristics of Organic Compounds on Hexadecyltrimethyl ammonium
- Smectite, SoilSci, Soc. Am. J., vol.52, pp. 652-657, 1988.
13. Boyd, S. A., Lee, J. F., and Mortland, M. M., Attenuating Organic
Comtaminant Mobility by Soil Modification, Nature, vol. 333, pp.
345-347, 1988.
14. U.S. EPA, Technology Evaluation Report: SITE Program
Demonstration Test, HAZCON Solidification, Douglassville,
Pennsylvania, Volume I. Report No. EPA/540/5-89/001 a, EPA Center
for Environmental Research Information, Cincinnati, Ohio, 45268,
1989.
15. U.S. EPA, Technology Evaluation Report: SITE Program
Demonstration Test International Waste Technologies In Site
Stabilization/Solidification, Hialeah, Florida. EPA Report No.
EPA/540/5-89/004a, EPA Center for Environmental Research
Information, Cincinnati, Ohio, 45268, 1989.
16. Thyagarajan, B. S., Process for Treatment of Fluids Contaminated
with Polychlorinated Biphenyls. U.S. Patent No. 4,612,404, 1986.
17. Wilwerding, C. M., Degradation of Polychlorinated Biphenyls. U.S.
Patent No. 4,931,167, 1990.
18. Manchak, F., Methods of Transforming Sludge into Ecologically
Acceptable Solid Material. U.S. Patent No. 4,079,003, 1978.
19. Hutzinger, 0., Safe, S., and Zitko, V., The Chemistry of PCBs.
CRC Press, Cleveland, Ohio, 269 p., 1974.
20. Alford-Stevens, A. L., Eichelberger, J. W., and Budde, W. L.,
"Multilaboratory Study of Automated Determinations of
Polychlorinated Biphenyls and Chlorinated Pesticides in Water,
Soil and Sediment By Gas Chromatography/Mass Spectrometry,"
Analytical Chemistry. 22, 304, 1988.
21. U.S. EPA, Methods for Organic Chemical Analysis of Municipal and
Industrial Wastewater. EPA 600/4-82-057, U.S. Environmental
Protection Agency, Office of Research and Development,
Environmental Monitoring and Support Laboratory, Cincinnati, Ohio,
45268, 1982.
44
-------
22. Bellar, T. A., and Lichtenberg, J. J. Some Factors Affecting the
Recovery of Polychlorinated Biphenyls (PCBs) from Water and Bottom
Samples. In Water Quality Parameters; American Society for
Testing and Materials: Philadelphia, PA, 1975; Special Technical
Publications 573, pp. 206-219.
23. Bellar, T. A., Lichtenberg, J. J., and Lonneman, S. Recovery of
Organic Compounds From Environmentally Contaminated Bottom
Materials. In Contaminants and Sediments; Ann Arbor Science: Ann
Arbor, MI, 1980; vol. 2, pp. 57-70.
24. Weitzman, L., Pluhar, D., and Barth, R., "Solvent Washing of PCB-
Contaminated Soil," Final Report on Research Pro.iect 1263-15,
Electric Power Research Institute, April, 1988.
25. U.S. EPA, SW-846. Test Methods for Evaluating Solid Waste. Vol.
IB, Third Edition, U.S. Environmental Protection Agency, Office of
Solid Water and Emergency Response, November 1986.
45
-------
APPENDIX A: DRAFT FINAL REPORT SUBMITTED TO EPA BY RMC
FINAL REPORT ON THE "DISAPPEARING PCBs" PROJECT1
Dr. R. Soundararajan
RMC Environmental & Analytical Laboratories
214 West Main Plaza
West Plains, MO 65775
February 4, 1991
The draft report presented in this appendix has been editted by EPA
staff for clarity. Technical corrections are presented as footnotes so that
the content of the original draft report is preserved.
46
-------
INTRODUCTION
In response to a U.S. EPA Region V report from Mr. Robert J. Bowden,
Chief, Emergency and Enforcement Response Branch, to Mr. Timothy Oppelt,
Director, RREL, the following study was conducted to identify the processes
(chemical and physical) which may be involved in the apparent PCB
concentration changes reported at the General Refining Site and other
similar site locations.
BACKGROUND
Oily soils at CERLA sites frequently contain PCBs with levels
typically between 200-300 ppm. In an effort to stop the spreading or
migration of oily contamination and PCBs at those sites, lime and/or fly
ash is often added in an attempt to minimize this spreading or migration.
In several instances, it has been found subsequent to treatment, that PCB
concentration levels in the treated soils have been materially reduced.
The apparent reduction exceeds that explained by simple dilution. A
possible explanation for these discrepancies could be poor analytical
testing or poor sampling at the site.
Samples for this study were provided by Region V of Chicago. All
pertinent information was provided by Region V, Chicago. Information
provided includes: site description, sampling areas, site treatments, if
any, and chain of custody forms, etc. RMC pursued a course of diligent
sample management and preparation as well as accurate analysis of all
samples. Further, three known PCB individual isomers were spiked on a
synthetic soil matrix and were subjected to quicklime treatment under
controlled conditions in this lab. The results and the conclusions drawn
from them are presented in this final report.
EXPERIMENTAL DESIGN
The experimental part of this project consisted of two major
sections. The first section was the extraction of site samples provided by
U.S. EPA Region V for the identification of any PCB residues. Fourier
Transform Infra-Red spectrometry (FTIR) and Differential Scanning
47
-------
Calorimetry (DSC) were used for the characterization of the extract.
However, the GC/MS and FTIR studies revealed that there were no PCBs
present in the residue. Extractions were carried out on 100 gram samples
with hexane and acetone and the final volume of the extract was reduced to
1 ml and was analyzed by GC/MS. FTIR of the extracts was run as KBr
pellets.
The second part of the experimental section consisted of preparing a
spiked matrix made up of sand:silicon dioxiderdiatomaceous earth in a 1:1:1
ratio. Three individual PCS isomers obtained from Ultra Scientific were
dissolved in methanol:methylene chloride solvent and spiked to yield a
concentration of 1333 ppm each. Fifty grams of this mixture was thoroughly
mixed with pre-calcined commercial quicklime in the ratio of 1:1 at first.
However, the rate of the reaction, the reproducibility of results, and the
variation in the intermediate products warranted minor modification. To
achieve concordant results, numerous experiments were conducted to
reproduce the site conditions. It must be remembered that tons of high
calcium fly ash (CaO) were added at the site involving millions of
kilocalories of heat. The heat would be sustained for an extended period
of time since both soil and quicklime are insulators. Hence, the mixing
was done during the final six sets of experiments as follows:
1. The spiked soil was mixed with quicklime in the ratio of 1:2.
2. Water was added slowly with vigorous stirring until the
temperature rose to a maximum.
3. The reaction vessel was set aside for an hour. More water was
then added and the temperature was maintained around 80-90
degrees C for at least three hours on a hot plate.
4. Another set of experiments (Steps 1 and 2) were done inside a
glove bag. The effluent gases were purged into a tenax column
of an LSC II device and desorbed into a Finnigan 5100 GC/MS for
identification and quantitation of the effluents.
5. Samples were taken at 24, 48, 72, 96, 120, and 720 hours from
the six sets of reaction vessels, dried in a desiccator over
P2°5 extracted with acetone and/or other suitable solvents,
reduced to 1 ml, and analyzed using a modification of EPA
Method 680. The PCB standards were used to create a five point
calibration curve with auto quan methods using proper
3,5-Dichlorobiphenyl,
3,3',5,5'-Tetrachlorobiphenyl,
2,2',4,4',5,5'-Hexachlorobiphenyl
48
-------
quantitation ions. The analysis was done on a full scan basis.
Whenever necessary, proper dilutions were carried out.
RESULTS
Results of six different experiments are given in Tables 1-9; the
graphic representation is shown in Fig. 1. These six sets of experiments
were conducted at different times. There were other experiments where some
slight modifications were used such as keeping the reaction temperature
elevated for 48 hours, etc. These experiments, of course, are not within
the scope of these investigations. These experiments were used only to
confirm certain kinetic factors in the chemical reactions. The overall
results may be presented as follows:
1. During the exothermic CaO + H20 reaction, no PCBs were found to
volatilize. No fragments of PCBs (chlorobenzenes) were seen,
either.
2. The biphenyl structure was not preserved at the end of the
reaction. The CC bond between the benzene rings was
completely destroyed.
3. Only one substituted phenol was identified as one of the
intermediate products.
4. Both alkyl- and chlorine-substituted cyclohexanes were found as
intermediate products.
5. Saturation of the benzene ring, cleavage of the aromatic ring,
and subsequent oxidation of the terminal carbon atoms are
strongly indicated.
6. Presence of inorganic chloride in the post-treated waste was
confirmed. There were no Cl ions in any of the reagents
(except covalent Cl in PCBs) during this reaction. This
confirms the fact that the chlorine in the PCBs was removed.
7. The concentrations of all PCBs dropped after 24 hours, but
after 48 hours the tetrachlorobiphenyl completely disappeared.
The other two compounds (di and hexa) were also reduced
substantially. After 72 hours, all of them disappeared. Only
traces (<5.0 ppm) of the hexachlorobiphenyl were seen.
8. In a related experiment (data not included in this report),
pure CaO did not bring about this reaction.
9. When the reaction medium was kept at elevated temperatures
("95-100 degrees C) the reaction was much faster. The entire
destruction was completed within 36 hours.
49
-------
DISCUSSION
Although the fact that PCBs are destroyed when treated with fly ash
in the presence of water is confirmed, we are left with a number of
perplexing, unanswered questions that need to be answered. A methodical
investigation into the inner workings of this complex reaction is
warranted. The end products and the following postulates can only be
considered as the "tip of the iceberg." In the following segments let us
consider the possible chemical reactions that could yield the observed
intermediates and end products. These considerations are based on well
established concepts of both organic and inorganic chemistry. The most
conspicuous reaction is the reaction between calcium oxide and water
forming calcium hydroxide and heat :
CaO + H2O > Ca(OH)2 + A H
AH = 235.68 k.cal/mole (1)
This also results in several secondary reactions such as
CxHyCl2 + CaO > CaC2 + CO2 + CaCl2 + H2O (2)
It is worth noting that one mole (56 grams of CaO) releases 235.68
cals of heat. In a field mixing situation, one ton of CaO can liberate
/
3.82 million kilocalories of heat , which can help to sustain the reaction
for several days. The Ca(OH)2 formed in this reaction raises the pH to 13.
Assuming that the AH (heat of formation of Ca(OH)2) brings about a simple
thermolysis (split by heat energy) we can see intermediates ranging from
chlorobenzenes to hydrogen chloride, which of course will be neutralized
immediately after formation:
The heat of formation cited for calcium hydroxide is correct for its
formation from constituent elements in standard state, but not correct for its
formation as shown in the equation. The heat of formation of calcium
hydroxide from calcium oxide and water is -15.6 kcal/mol. Data for these
calculations were obtained from: R. C. Weast, M. J. Astle, and W. H. Beyer,
CRC Handbook of Chemistry and Physics, CRC Press, Boca Raton, Florida, pp.
D50-D93, 1986. tech. ed.
4
Using -15.6 kcal/mol for the heat of formation leads to evolution of
0.25 million kcal heat per ton of Ca(OH)2 formed from CaO and HpO. tech. ed.
50
-------
A
> CXHV + C6H5C1
Thermolysis rn J. rri ,
CaO C0 + C°2 4
(3)
Cl Cl
We do have indirect evidence for this reaction. One of the
intermediate products is a phenol. The presence of the phenol can be
explained by a simple 3^2 reaction between the chlorobenzene and the
Ca(OH)2.
Cl
'H
(4)
chlorobenzene
However, the most intriguing aspect of the entire treatment is the presence
of cyclohexane derivatives, which are ring saturation products. It appears
that after the initial thermolysis and SN2 substitution, the phenolic
compounds seem to undergo reduction. During the simulated reactions in the
laboratory there was no source of hydrogen to bring about such reductions.
The possibility that water could have been split into H2 and O2 is quite
slim unless there is a strong catalysis hitherto unknown involved. It is
imperative to point out that in the commercial quicklime there are numerous
redox systems that could bring about every conceivable organic reaction.
An examination of the E values of these redox systems (in commercial
quicklime) confirms this view. Hence, the formation of cyclohexane
derivatives may be visualized as follows:
Mn+l/Mn
+ HC1
(5)
Addition of certain additives such as slag powder would enhance this type
of chemical reaction.
Yet another chemical factor to be considered here is steric
hindrance. In heavily chlorinated PCBs, the bulk of chlorines would
prevent the approach of OH for substitution. It appears that partial
breaking of CCl bonds is involved during the exothermic step (1) which
leaves the aromatic ring with only a few chlorines.
51
-------
One of the most significant end products is the ester of hexane dioic
acid. The formation of this product very strongly suggests that following
the ring saturation there is ring cleavage and subsequent oxidation of the
terminal carbons to carboxylic acid functional groups. It is also
interesting to note that no other dicarboxylic acid derivative was found.
The six member carbon chain is another indication that its precursor was a
six membered ring. This oxidation phenomenon can be attributed to both
quicklime and dissolved oxygen in the water that is added during this
treatment process.
It has been established that the organic chlorine in the PCBs has
become inorganic CaClj. The Cl was measured with the aid of ion selective
electrodes. The original reaction medium (CaO, sand, etc.) did not have
any chloride in it before the reaction began. This evidence again supports
the idea that the chlorines were either removed by thermolysis or by a
simple nucleophilic substitution process. It has also been established
that the reaction rates of this process are directly proportional to the
reaction temperature. At elevated temperatures, the disappearance of PCBs
was faster. During the investigations, we have found that the reaction
rates are directly proportional to the concentration of quicklime. This
observation is in agreement with the law of mass action. The site samples
were subjected to massive extraction procedures, but none of them had even
traces of PCBs. This is not due to any stabilization, encapsulation, or
masking, but due to the fact that the PCBs have been destroyed completely.
Further, the total GC/MS analysis of the site samples showed
considerable amounts of long chain saturated hydrocarbons. These compounds
during excessive heat release could have saturated the benzene rings in the
PCBs as the whole process resembles a closed system. The entire phenomenon
can be speculated on, in light of experimental data, as follows:
52
-------
FLYASH-PCB REACTIONS SCENARIO I
REACTION 1
CaO + H20
Ca(OH)2 + 235.68 kcal/mol
thermolysis
hydrolysis
OH
OH
substituted phenol
(i.e., 235.68 kcal per 56 grams;
therefore, 3.82 million kcal per
ton of CaO)
REACTION 2
thermo-
lysis
A
IHJ
thermolysis, reduction,
\hydrolysis
OH \ A
+ HC1
CH3 4-methyl-
cyclohexanol
thermolysis, reduction, ring
cleavage, oxidation of
terminal carbons
ROOC (CH2)4~COOR
hexane dioic acid ester
53
-------
FLYASH-PCB REACTIONS SCENARIO II
HC1
A
CaCO,
CaCl-
CaC,
C02 + CO
FLYASH-PCB REACTIONS SCENARIO III
AH
CnH2n+2
2 [H]
[H]
Cl
[H]
OH-
OH
54
-------
CONCLUSIONS
1. The interaction between high calcium fly ash (CaO) and
polychlorinated biphenyls (PCBs) results in the total
destruction of PCBs.
2. The reaction mechanism is still unclear. This would warrant
thorough investigation where the reaction would be frozen at
different time intervals and the intermediates would be
analyzed and identified, possibly by GC/MS/MS.
3. The phenomenon of catalysis is very strongly indicated, but
which catalyst brings about this destruction is yet to be
determined.
4. The stoichiometry as well as the upper organic threshold are
yet to be determined to avoid fire/explosion and volatiles
release into the atmosphere during site remediation.
5. The prospects for the application of this process for the
destruction of other organic wastes appears to be bright, but
systematic and thorough investigations are needed.
In essence, our investigations with limited scope, resources, and
time, indicate that this process needs to be evaluated properly since its
effectiveness and cost efficiency are phenomenal before full scale use in
the field.
REFERENCES
1. Cotton, F. Albert and Wilkinson, Geoffrey, Advanced Inorganic
Chemistry, 4th Edition, John Wiley & Sons, 1980.
2. Morrison, R.T. and Boyd, R.N., Organic Chemistry. 2nd Edition, Allyn
and Bacon, Inc. , Boston, 1958.
3. Gibbons, J.J. and Soundararajan, R. , "The Nature of Chemical
Bonding between Organic Wastes and Organophilic Binders, part 2",
American Laboratory 21, (7) 70-79, 1989.
55
-------
TABLE 1
DICHLOROBIPHENYL CONCENTRATION (ppm)
Zero Hours - (Baseline)
Sample #
Run 1
Run 2
Run 3
Average
1
1335
1329
1375
1346
2
1300
1346
1292
1313
3
1285
1340
1301
1309
4
1400
1395
1362
1386
5
1320
1304
1278
1301
6
1362
1286
1390
1346
Statistical Calculations: mean 1333
standard deviation 32.08
relative standard deviation 2.41%
TABLE 2
DICHLOROBIPHENYL CONCENTRATION (ppm)
Sample #
Run 1
Run 2
Run 3
Average
1
734
780
747
754
2
700
762
774
745
24 hours
3
689
645
670
668
4
710
690
704
701
5
781
761
772
771
6
660
620
678
653
Statistical Calculations: mean 1333
standard deviation 32.08
relative standard deviation 2.41%
TABLE 3
DICHLOROBIPHENYL CONCENTRATRION (PPItO
Sample #
Run 1
Run 2
Run 3
Average
1
138
140
140
139
2
135
140
132
125
48 hours
3
121
125
130
125
4
135
140
142
139
5
155
161
189
165
6
121
134
137
131
Statistical Calculations: mean 1333
standard deviation 32.08
relative standard deviation 2.41%
56
-------
TABLE 4
TETRACHLOROBIPHENYL CONCENTRATION (ppnO
Sample #
Run 1
Run 2
Run 3
Average
Statistical
1
1216
1230
1290
1242
Calculations:
0 hours (baseline)
234
1210 1179 1221
1193 1201 1243
1225 1192 1300
1209 1191 1255
mean 1333
standard deviation 32.08
relative standard deviation 2.
5
1315
1302
1317
1311
41%
TABLE 5
TETRACHLOROBIPHENYL CONCENTRATION fppm)
Sample #
Run 1
Run 2
Run 3
Average
Statistical
1
837
880
855
857
Calculations :
24 hours
234
828 775 835
868 787 868
867 776 876
854 779 860
mean 1333
standard deviation 32.08
relative standard deviation 2.
5
925
936
920
927
41%
TABLE 6
TETRACHLOROBIPHENYL CONCENTRATION (ppm)
Sample #
Run 1
Run 2
Run 3
1
11.2
10.9
11.7
48 hours
234
10.1 9.3 11.1
11.4 10.1 10.6
9.8 8.9 11.4
5
12.3
11.8
12.0
6
1208
1197
1182
1196
6
793
810
800
801
6
9.6
8.8
9.5
Statistical Calculations:
mean 1333
standard deviation 32.08
relative standard deviation 2.41%
57
-------
TABLE 7
HEXACHLOROBIPHENYL CONCENTRATION (ppml
Sample #1 2 3 4 5 6
Run 1 1315 1310 1301 1316 1382 1300
Run 2 1345 1352 1329 1329 1391 1298
Run 3 1325 1328 1313 1350 1385 1302
Average 1328 1330 1314 1332 1386 1300
Statistical Calculations: mean 1333
standard deviation 32.08
relative standard deviation 2.41%
TABLE 8
HEXACHLOROBIPHENYL CONCENTRATION (ppm)
24 hours
Sample #
Run 1
Run 2
Run 3
Average
1
612
632
596
613
2
614
645
582
614
3
547
535
529
537
4
630
608
602
613
5
691
701
689
694
6
552
525
601
559
Statistical Calculations: mean 1333
standard deviation 32.08
relative standard deviation 2.41%
TABLE 9
HEXACHLOROBIPHENYL CONCENTRATION (ppm)
48 hours
Sample #
Run 1
Run 2
Run 3
Average
1
334
341
327
334
2
327
357
302
329
3
302
307
300
303
4
329
333
341
334
5
408
393
389
397
6
321
317
302
313
Statistical Calculations: mean 1333
standard deviation 32.08
relative standard deviation 2.41%
58
-------
1400
CONGENER REMAINING, mg/kg
200
400
TIME, h
600
800
DCBP
TCBP
HCBP
Figure 1. Average loss of PCB congeners over time.
59
-------
RIG
11- 13-"3>3 14:45:08
SAMPLE: FIRST BASELINE CHECK
COHGS.: PR
RANGE: Q 1,35-jS
O^TH: =131r5Li =J81
CftLI: ;:i:i36 42
590 TO 300
im. a-
ffv
O
RIC
730
509 ''2=^2;:">-'.
3158010. 130
13741409.
I
j
i
i
(
527tfy.
i
j
:"522=:c^
! j
|
i
i i
j i
}
1
' ' ' I ' ' ' ' i ' ' ' ' i ' ' ' ' I ' ' ' ' 1 ' ' ' ' i-
3 650 700 750 809 350 900 SCflH
90 10:50 1.1:40 12:30 13:20 14:10 15:00 TIME
-------
5IC D^Tii: PQ1113C #1
11-13 '--£1 13:41:£-£' CALI: PQ1113B #2
3ftNPL£: PC&/QL SECOND CHE:X i5HOLiRS OESC/Pft PS SOP
SCANS 588 TO 386
RANGE: -:
LAEEL: :-'
QUnri: H 2, 2.8 J 8 BASc: U 28.,
cr>
190243S.
RIG !
535
122524.
A
531 '; 634 714 J.8,4.
2356, I \ 342. 794. 8i'2w-
. : : ,11:84. , : .! \. , , ., , ,,732. 4175. 42275- i
352
52£.
1827.
. j - , 1 , ! ; : , . j , , . : ,..,.,.,,,,.. !
38 5 -"8 S0f. S58 788 758 398 358 398 SCAN
:2'j 3. If. J0:9y 18:58 11:48 12:3? 13:29 14:18 15:86 TIME
-------
RIG DATA: F-31113C 410
11/13/39 13:41:99 CALI: PQ1113B *2
SAMPLE: PCS/QL SECOND CHECK 4SHOURS DESC/PS: PS SCF
CQNOS.: PR
RANGE: S 1..313? LABEL: N 2. 3.9 QUAN: A 2, 2.9
SCANS 599 TO
i
_
i
:
]
803
1082430.
-.cr-j A .~i-?i~
i-Jf fc
535
122624. :,
72f3Q- 704.
Hi' nil ' i i 1 1\_^ . 1 1 1 in i 4175. '
,
852
526.
1827. ' '
I
1
i
i
i n i i t »
h ' I ' I ' i, ' i ' i
00 600 703 300 300 1008
:20 ; (0:00 '.1:40 i3:28 15:00 16:49
SCAN
TIME
-------
11-' 13' 39 13:41:38
SntlPLS: PCS/QL SECOND HH
CONDS. : PR
D-"R: PQ1119C 41
C-Li: PQ1113B *2
4::Hn:;R'" ':£;' 'FR P'' '"OF
,-Tpf-ji .;- ; j ;.'-:'-'5 ±£?< '] ". " :-. ~-;'^;i' i ~- "> .'. »i Ph'jt' '' '^> :
w _^__ .. ... ... ..... _. ... _ _
213:?:;.
25.7^
222 "
,,.,.,
;i
l\
| '', 54; ?S3
i -, i7.; 387.
i1 V ' > 146:', 4382.
' : . | . . . ! j : , 1 1 -..,,..,,, ; i . ,
21333.
222. 8bb
i 8.588
1 "" ;' l i ----- j~ -j ~, ;
383
::;0:-
3"i " ' ";
292
_
|
7K
1538. J
19298. ,!
59.
4l::< 385S3.
232.987
± 8.588
i * 1 ^ ^ i ' j i 1 i ' ; : : i 1 j i 1 i 1 j ; T : ~~ ; . > ' , ' ' *
388
32948.
1 89 . 8-
359 _
J
|
1
j
1 ' 1 I | 1 i 1 1 | i i i i | . 1 i 1 j 1 i 1 i | i I C 1 |
lk):-il 32843.
368. 188
± 0. 588
, . . . i . i . . l
889
i . 1982498.
-i
RIC _
-
257^
595 ; i
551 ^irJC-M 654 714 784 1
295S. '7^6%' 343- ?04'- 8320' i
... . .11184. . ' '"A!-' .' ..'792. 4175. 42975. j
'. ' ' ' i ? ' ' ' i ' ' ' r : ! ' ' ' i ' ' ' i ' ' ' ' i
88 559 688 558 789 758 808
^370- 1802438.
852
52S.
1327.
... 1 , ... i
350 909 SCAN
:;20 3:18 18:08 18:58 11:40 12:38 13:28 14:18 15:98 TIME
-------
'hH-LE j
LIBRARY SEARCH
11/19/39 11:32:99 + 7:90
SAMPLE: FCB/GL TREATED 48HOURS AFTER
COHOS.: ?R
ENHANCED (S 15B 2H 9D
DATA: PQ1113A * 429
CALI: PQ1113B # 2
BASE M/2: 205
RIC: 29631.
1
!
il__
ll,
|
I.. II ."I 1 i .'-
1
1 ,
'HENOL, 2'b-BISa . l-QIMETHYLETHYD-4-METHYL-
U UU
C17.H27.02.H
3 °S 265
?AHi< 2
£ O|Tp|/J
F!JR~"S49 i
[ t i t'T*» *->*-* ' r»'T"1 r*' ' "^ '^ ' T~
PHENOL, 2,6-81S(1,1-DIMETHYLETHYL >-4-METHYL-- nETHYLCARBAMATE
H HT'S!
B PK 205
*" 23253
PUR 4?8
11,14-OCTADECADIYNOIC ACID, METHYL ESTER
-------
LIBRARY SEARCH DATA: PQ1113A * 3S4 SuSE M/Z: 79
11/13/39 11:32:98 + it:2* CAL!: PQ1113B it 2 RIC: 2373536.
SAMPLE: PC3/QL TREATED 4SHOUP5 AFTER
CONDS.: PR
1317 r ' r
iAflPLE
t ,
i
!
i f
r
.. l.i..ii. ii 1 .! , f
Cl-t.H2S.G4 HEXANEDIOIC ACID, f'lCNGC-ETHYLHEXYL; ESTER
. , ylx^7
B PK "55
RANK 1
» 24633
o> PUR 753
Ul
r
:
]
i
''')!!. 1 . -...
C22.H42.G4 HEXAHEDIOIC ACID, DIQCTYL ESTER
H HT1^ |
B PK. 123
RANK 2
* 34723
PUR 641
r
1 |
1
j| .
II 1 . , .
C22.H42.04 HEXftNEDIQIC ACID, BIS(2-ETHYLHEXYL> ESTER
H HT^fS ]
B PK 41
RANK 3
# 34722
PUR S32
r
\
I
i
1
""2 39 1.90 1.50 799 '^58 '»? "59
-------
LIBRARY SEARCH
11/19/30 12:14:80 + 3:43
SAMPLE: PCS /QL FIRST CHECK 24 .-.WI
CONGS.: PR
ENHANCED '.: 1..E 2N 07^
DAT*: -Q11.1
CnLI: .-"Hii
DEIC. 'PR PS £GP
en
CS.H13;CL2
M wr?& ]
B FK 31
RANK 1
f* 7334
PUP 502 '
CS.H19.CL2
M UT1!-^ "
6 PK 81
RANK 2
* 7333
PUR 451
C5.H7.H
M WT M.
B PK 81
RANK 3
# : 413
PUR 443
CV'CLQKE/.AriE, 1..2-DICHL'jRO-. CIS-
CVCLOHEXAHE, 1,3-DICHLORO-, CIS-
1H-PYRROLE, 1-METHVL-
-UlUL
,P-
Cil
-------
APPENDIX B - SELECTED CHROMATOGRAMS, MASS SPECTRA
AND COMPOUND IDENTIFICATION
MATERIALS
Stock solutions and blank extracts were analyzed by GC/MS to identify
contaminants that could affect experimental results. Figures B-l through
B-3 show chromatograms of the individual PCB congener solutions prior to
mixing. Figure B-4 is a chromatogram of the three internal standards used
for calculating relative retention times and response factors. In each
case, the flat baseline indicates the absence of contaminants that could
interfere with PCB and reaction product analyses. The chromatogram of the
actual spiking solution fortified with internal standards is shown in
Figure B-5. Mass spectra obtained in this study are compared to reference
spectra for the three PCB congeners in Figures B-6 through B-8.
Blank samples of synthetic soil were processed through lime
treatment, extraction and analysis to check for contaminants in solvents
and glassware used in the procedures. These blanks were treated exactly
the same as experimental samples except that the spiking solution did not
contain the PCB congeners. Six chromatographic peaks were detected (Fig.
B-9) at low concentrations (compare ion counts in Fig. B-9 with Fig. B-5).
The four peaks that could be matched to reference spectra were identified
as common laboratory contaminants (Figs. B-10 through B-15).
TENTATIVE PRODUCT IDENTIFICATION
Aliquot samples of extracts produced in the open-vessel experiments
were sent to Battelle Columbus for identification of potential
decomposition products. The values shown in Table 8 of the main body of
this report are from analyses performed by Battelle. Selected data are
presented here to support the identification and semi-quantitation of
products.
67
-------
Stock Solution and Solvent
Chromatograms of methylene chloride solvent and the stock PCB
solution used to spike the synthetic soil are shown in Figures B-16 and B-
17, respectively. The flat baseline of the solvent chromatogram indicates
that the compounds tentatively identified as reaction products are not
solvent contaminants. The stock solution, at 50-fold dilution, shows one
peak in addition to the spiked congeners and internal standard at a
retention time of 32.56 minutes. Mass spectra and peak identifications are
shown in Figs. B-18 through B-22. The unknown compound was not
sufficiently resolved or concentrated for identification.
Open-Vessel Reaction, 72 hour sample-replicate 2
The 72-hour, replicate-2 sample was selected to show compound
identifications since this sample contained all the products observed in
the open vessel reactions. Figures B-23 through B-43 illustrate the
chromatogram, peak mass spectra, and tentative compound identifications for
the 72-hour sample. Peaks numbered 2, 8, and 14 on the chromatogram (Fig
B-23) are the spiked PCB congeners, identified by relative retention time
and mass spectrum compared to pure PCB standards. Peak 20 (Fig. B-23) is
the internal standard, chrysene-d^* used to compute relative retention time
and to semi-quantify unknown peaks. Unless otherwise indicated,
identification of unknown compounds was based on searches of the NIST data
base and manual interpretation. The probability of each of several product
identifications being correct is shown under each mass spectrum, except for
peak 18 which yielded no interpretable spectrum.
High-probability identifications were made for peaks 1, 3, 5, 6, 9
and 12 (Figs. B-24, B-26, B-28, B-29, B-32 and B-35) as mono- through
pentachloro biphenyls. Isomer identifications could not be made from mass
spectral data. Further identification would require GC/MS analysis of pure
samples of all the candidate congeners. Peak 19 yielded a strong MS match
for a methoxypentachlorobiphenyl.
Moderate-probability matches were found for peaks 4, 7, 10, 11 and 17
(Figs. B-27, B-30, B-33, B-34 and B-40) yielding tentative identifications
of hydroxymonobiphenyl, tetrachlorobiphenyl (not the starting TCBP
congener), pentachlorobiphenyl, methoxytrichlorobiphenyl, and
tetrachlorodibenzofuran, respectively. The remaining peaks did not yield
good matches with library spectra; tentative identifications were
determined by manual interpretation. Peak 15, eluting at a relative
68
-------
retention time of 0.9495, may be a contaminant since the diluted stock
solution yielded an unidentified peak at a relative retention time of
0.9498 (Fig. B-17).
The tetrachlorodibenzofuran (TCDF) compound warranted further
examination because of possible toxicity of TCDFs. A sample of 2,3,7,8-
TCDF available at Battelle was used to spike a d!2-chrysene-fortified PCB
calibration standard; the spiked sample was then analyzed by GC/MS.
2,3,7,8-TCDF eluted at a relative retention time of 0.9578 and exhibited a
response factor of 0.359 relative to d!2-chrysene. Product compounds in
open-vessel extracts that were tentatively identified as TCDFs eluted at
relative retention times of 0.9570 to 0.9573 (see peak 17, Fig. B-23 for
example). The relative retention time along with spectral matching support
the identification of TCDF. Isomer identification is less certain, since
other TCDFs may have nearly identical retention times. The total ion
current response factor measured for 2,3,7,8-TCDF was used to estimate TCDF
concentrations in all open-vessel extracts shown in Table 8. Figure B-43
shows the chromatogram of the 72-h, untreated control sample, indicating no
contamination that could be interpretted here as reaction products.
69
-------
Fi le >BIPCB
4998889-
3699889-
329889CV-
2888898-
2498888-
2898889-
1688889-
1299880-
389990-
499998-
48.8-458.9 amu. 3>s DICHLOROBIPHEHYL
-108
-?8
k*e
-70
-r.O
-50
-40
-38
^0
-10
4 6 8 19 12 14 16 18 28 22 24 26 28 39 32
Figure B-l.
Total ion chromatogram of 3,5-dichlorobiphenyl stock
solution.
File >TEPCB
1R88988-
1698999-
1499098-
1290999-
1808088-
889889-
689889-
488880-
280800-
nJ
4 '5 ' 8 ' 18 ' 12 ' 14 ' 16 ' 18 ' 28
22 24 26 28 38 3<
-188
-90
-SO
-7ft
-..9
-10
CO
-in
ki
Figure B-2,
Total ion chromatogram of 3,3',5,5'-tetrachlorobiphenyl
stock solution.
70
-------
File >HXPCB
2209099-
2000000-
1899099-
1690998-
1499999-
1208809-
1999999-
-
888899-
608099-
499800-
299089-
IIC
p ' i ' T r i i i i r~-r- 7- i i i i -|-' i i i
4 6 8 10 12 14 16 18 29 22 24
1 I i
26 28 30 3^
-190
-*0
-89
.,
-1H
-30
-2"
18
Figure B-3.
Total ion chromatogram of 2,2',4,4',5,5'-
hexachlorobiphenyl stock solution.
Fil- >5UPLC
798000-
6S000B-
600098-
580080-
450980-
400009-
359880-
309909-
259880-
280000-
159009-
100000-
58809-
0-
49.8-4S0.0 »mu. SUPELCO STD
A
L.
c
4 f, 8 10 12 14 16 18 20 22 24 26 28 30 t-
-100
r?0
';7Q
^.0
!^«
^«
-Jro
-20
-19
^)
Figure B-4.
Total ion chromatogram of internal standard spiking
solution containing: acenaphthene-d^g (A), phenanthrene-
d10 (B), and chrysene-d12 (C).
71
-------
File '.5TDOI 48.9-4S9.9 a*u . BI ,TETRR ,HEXfl- CHLOROB1PHEMYL
TIC P
2400000-
220000O-
20»0990-
1800000-
1600800-
1499909-
1 200000-
1009009-
399009-
699000-
400000-
290000-
0-
A
D
E
F
l i r~ r r-* r^ r-1 i l ^T "~~\ ^~r^ r^ r~ r T^T ^T ^T ^T r~r^ r-* r-1 i
4 6 8 10 12 14 16 18 20 23 24 26
' i i i i i
23 30 S.
-100
^r,
-7'.'
,..
H
-10
0
Figure B-5.
Total ion chromatogram of primary dilution standard with
internal standards: acenaphthene-d1Q (A), 3,5-
dichlorobiphenyl (B), phenanthrene-d1Q (C), 3,3',5,5'-
tetrachlorobiphenyl (D), 2,2',4,4',5,5'-
hexachlorobiphenyl (E), and chrysene-d^
Fil» >STD88
Bpk flb 9999
18888-
8808-
6888-
488B-
2999-
9=
75
88
^ "
File >BIGDB
Spk Ob 9999
10009-
8999-
6009-
4009-
2009-
0-
88
80
STD STB 88 US/ML Scan 682
128 168 288JL,0, 2<18 288 32a 36e 488 M'24 "ln'
-^
152
126
/
-, ....J. ,r
186
|
2E5 231 329
, \ v 307 S 3S' 38S *z*
r^~. .. ^ . ? s f r . i
129 168 288 249 288 328 368 488
-198
-89
-69
-48
-29
-9
3,S-D1CHLOROBIPHEMYU Scan S2963
FIT 9.98 min
120 160 298 240 289 329 369 499
--
152
84 {
75
~~^
/
80
111
,S 186
f '
1 1, .
226
188
-88
<0
-48
-28
128 ' 168 ' 288 ' 249 ' 288 ' 328 ' 368 ' 480 '
Figure B-6.
Mass spectrum of 3,5-dichlorobiphenyl compared to NIST
library reference spectrum.
72
-------
Fil* >STD88
Bpk Ob 9999
18089-
8809-
6009-
4808-
2980-
8-
Fil* >BIGDB
Bpk Rb 9999
8899-
6909-
4899-
2099-
9-
STD
88 128 168 298
2
74 "9 icn
\ 92 X" ' '84
k ' 1 1 '
..i.r *_,.!...«. ij..,.lL « i.l....... Hi... .
80 129 168 289
STD 80 UC/ML Scan 932
20.13 din.
290 C~
H
?
L. x. 1. . 7 r 3r T
240 280 320 360 400 448
1 , 1 ' -Blph«nyl , tetrachloro- Scan
FLT 8.0(
80 120 168 289 248 280 320 368 400 446
i ... i i ... i oaf.\
298 1
1
lie
92 S ise
/ r /
i L t /
I'M 1 i,
80 120 160 208
" 1 ^
248 280 328 368 408 448
-108
-80
69
-(0
-20
-9
6327S
) nin.
-tee
-80
-68
-28
-9
Figure B-7.
Mass spectrum of 3,3 ', 5, 5 ' -tetrachlorobiphenyl compared
to closest matching spectrum in NIST library. Isomeric
structure of reference spectrum not identified in
library.
File >STD88
Bpk flb 9999
18900-
8808-
6889-
4898-
2989-
0-
File >BIGDB
Bpk Pb 9999
10098-
8808-
6900-
4080-
0-
STB STD 88 UC/ML
88 129 160 288 240 288 320 3^
298
(
\\
.,..\,..Ij^, i 22.....i..A..!\l. i.
80 128 160 208 240 288 328 3<
1 ,1 --Biph*nyl , 2. 2', 4, 4', S. 5' -hexachloro-
FLT
298
/" / 3/5
88 ' 120 ' 160 ' 208 ' 249 ' 289 ' 329 ' 3<
Scar
23.7;
&.,,,.!?!, ,,.,,
1
L T /2 .
9 489
Sc an
9.9t
'A , 4?B
366
>9 408
1133
min.
-199
89
-60
-«9
-29
8
69792
) nin.
-100
-89
-r,*
-40
-29
Figure B-8.
Mass spectrum of 2,2',4,4',5,5'-hexachlorobiphenyl
compared to NIST library reference spectrum.
73
-------
Fil* > 1BLPN
LI
1990O-
.699*
14999-
12909-
19990-
a«90
"~
2000-
4S.0-4Se.0 ««u. SXIRftCT
A
\
Vs ^
4 A 8 10
tl! 3THOT5T
B
C
1
1
' IftMhuA. t/n-> JMAfcJW^-j
^n*Wf\*'yWri**lMM"^Tv
12 14 1
PC8.S
D
E
.A^r*^ M^
> ie
Hi^^*awLM>'wWi'aw«^wVk'''*lv
^r Tiji/ifurMyvpifyv* ^ iTf B^P ! ^
29 22 24 26
F
^
^31
-too
-90
-«.o
-^'.'
-.0
'!"
-2..
A - a,a-dimethyl benzyl alcohol
B - no suitable match
C - 2,6-bis(l,l-dimethylethyl)-4-methyl-phenol(BHT)
D - l-(l-cyclohexen-l-yl)-4-methoxybenzene
E - dibutylphthalate
F - no suitable match
2«»99-
;;,,.,
,,>
l-.9>:'i>
12000-
l-.OOfVH
,0-v
29i]>0-
45.9-459.8 ...j. CXtRaCT
IIC
E
A
'
VL_,
4 * a 19
t, UIMOUt
i
C
1
' WwV*«illrVlkt«Wl'
12 14 I
PCB.S f.
D
E
VllWfrW
> 18
tP 2
^i^Ws*yVwVfflWMOfl*rm T
29 ' 22 ' Z-t ' 26
F
^
2f
-?0
.9
-70
"
-10
-10
Figure B-9.
Total ion chromatogram of quicklime-treated soil blank
duplicates. Treated blanks consisted of synthetic soil
spiked with solvent only and processed through quicklime
treatment, heating steps and extraction.
74
-------
Fi le > tPLBH
Bpk Ob 999?
10000-
8000-
6000-
4000-
2000-
77
"\
[
llllll. ll
JOS
1.1 :
ee
Fi 1» >BIGBB
Bpk flb 3741
4P00-
3000-
2000-
1000-
0-
77
.,
h
80
EXTRflCTED WIHOUT PCB.S tear, S4
4.46 IT. 1 n .
121 r
1S2 178 , s \ 308 33S "x 387 ,'
129 16B Zee 240 280 320 360 400
-100
-8'S
-60
-40
-28
-1BLOII EXTRRCTEB MIHOUT PCB.S Scan 402
Bpk- Pb ?99? 10.6> mil,.
1000O-
3009-
6000-
4000-
2000-
8-
164
S0 1
6f /'' 132 I 192 28s 23, 283 ^ 33 412
Jl. ,lL,Jlll 1 , , , \.,.lll!ll . \ ^ .-^ N . V f ! ! . f'''
80 ' ,20 ' 160 ' 200 ' 240 ' 280 ' 320 ' 360 ' 400
-1 no
-SO
-60
-40
-20
-0
File >BICDB Phenol, 2 ,5-dichloro- Scan 40071
Bpk Pb 9999 FLT 0.00 n, i n .
19009-
8000-
6?00-
200O-
0-
--
*? *> ,33
. 1 , i, \
,66
\
80 12B 166 299 240 289 329 360 400
-1 00
-80
-60
-10
Figure B-ll.
Best NIST library spectral match for Peak B (Fig. B-9)
Identification questionable, with probability = 0.31.
75
-------
Fi 1* > 1BLB1I
Bpk Bb 9999
10008-
8888-
6000-
4000J
2000-
0-
EXTRBCTED HIHOUT PCB.S Scan 419
10.93 Kin.
285 r
S7
' 91 .., US 177
\ ' f f
,.il.i .1.1 .L j .in riv.ii.L, . , .1 i.
88 128
i 256 37-*
243 / 296 322 347 , 413
.1 ,' r . f , ' ' . .' !
168 280 248 288 328 368 400
-100
Uo
48
-10
File >BIGFB Phenol, 2 ,6-bi 5< I , 1 -dime thylethyl ) -4-inethyl - tear. 4?7'i7
Bpk fib 9999 FLT 8.00 »tir,.
10000-
8000-
6000-
4000-
2000-
105 i
\ 119
1 1 '1 '<
88 128
285 r
*S 177
\
168 280 248 288 320 368 400
-1 80
-80
*«
-*
Figure B-12.
NIST library matching spectrum for Peak C (Fig. B-9).
Excellent match, with probability = 0.96. Compound is
commonly known as BHT (butylated hydroxytoluene), a
common antioxidant.
Fil- >IBLfiH EXTRBCTED MIHOUT PCB.S Scan ril
Bpk Pb 9999 16.11 n.i n .
18800-
800O-
6000-
4000-
2000-
0-
188
8\ 94 '.60
i ' 139 1 1
i. .J ,il..al . '. . . ill J
207 259 286 41*
s N. X 312 338 389
30 ' 129 ' 168 ' 288 ' 248 ' 288 ' 320 ' 360 ' 488 ' 44
-180
-90
-60
-10
-*<
8
Fiie >BIGDB Benzene, 1 - ( I -eye 1 ohexen- 1 -y 1) -4-m* thoxy- S<- 3n 46070
Bpl- Ob 9999 FLT 0.00 mir. .
10000-
388O-
6000-
4080-
2000-
0-
1
160
129
1
111 1 J ,.
88
f
189
80 128 168 288 248 288 328 360 400 44
100
-80
*:o
40
-iO
-0
?
Figure B-13.
Best NIST library spectral match for Peak D (Fig. B-9)
Match somewhat uncertain, with probability = 0.52
76
-------
File >1BLRN
Epk Ob 999?
10990-
8809-
6890-
4990-
2090-
8-
Fi le >E1GBB
Bpk flb 99?'?
18088-
8800-
608O-
4909-
2080-
9-
1
76 194
\ \ 121
I..I b I. .1 ft . .1
80 128
1 .2-
1
76 104
X N
.1, ..),,.. ,»., , , ,,,,
T f T ''-r--f-*-* ^^
80 120
EXTROCTED W1HOUT PC8.S ?c
13. S
49
,,, 287 326 37P
183 "J 248 x 388 / < 391 43ft
(..{.,{.? .*.:' / . "\ '. .
168 ' 200 ' 248 ' 239 328 ' 360 ' 408 ' 44'
Benzenedic arboxy 1 i c ac i d , dibutyl ester Scan
FUT 9.8'
49
223
, ,'
i i - i - i - - i - - i i t - - - i - - i ' 1 ' ' ' I ' ' ' I ' ' ' I i
160 200 240 280 320 368 400 44r
in SC.S
1 nil..
-100
-90
-60
-10
-^
J
36477
1 mm.
-100
-fty
60
-40
T,
Figure B-14.
Best NIST library spectral match for Peak E (Fig. B-9)
F 1 1- 1 PLUM
Ppk Ob 99??
10000-
8000-
6999-
2990-
9-
186 129
... / i. .01. .. .
89 128
Fil* >B1GDB 2-Propen-
Bpk Ob 9999
iceetv
398O-
6990-
4900-
2990-
0-
107
1
128
/'
88 128
EXTRPCIED WIHOUT PCB.S t- ,: an 1 I _">
248
297 .
1S6 194 /'
{.. .' .U . Jl
~~~
447
281 317 343 369 3 ?S
'..'./' f
169 ' 288 ' 248 ' 288 ' 328 ' 369 ' 499 440
-inn
-99
-6.0
-«
1 -on« , 1 - (2-hydroxyph*ny 1 ) -3- ( 4-hydroxyph»ny It - Sc »n SS919
FLT n.OO «ln.
248
147
S 178 !
( I <( ' i
242
; - "
160 288 248 288 328 368 490 449
100
-en
-60
-JO
-jft
-o
Figure B-15.
Best NIST library spectral match for Peak F (Fig. B-10)
Questionable match, with probability = 0.42.
77
-------
TOTflL ION CHROMfiTOGROM
Fil« >B7010 45.0-450.0 Amu. BLBNK MECL2
TIC
400 800 1300
400000-
360000:
320000-
280000
£40000^
£00000-
l6000'>
1£OOOO
;30000-
4000O
v-1-.
Figure B-16.
OMOLYST LHK D12-CHR
1600 2000
1 ' 1
!£:
Ifa
1 1
£0
I ' 1 ' 1
24
£6
1 '
3£
1 1
36
40
Total ion chromatogram of solvent spiked with chrysene-d12
as internal standard.
78
-------
4*. 0-450 .0 ami*, DI 1,^'TEDS-TOCK 1 :50
(tOC TIC
*vu 13CO ilC'u lev
T LMK Dli-CHRYSEr
liC'O 1-tOC'
4000^
iOOOC^j
1
1
.'A'.".-!
£'_ ,'i c'i . w cc' . u c:^ .0 C!M . 0 c'5 . 0 ft- . u
£.'£: . 0 £':* .0 ?« . -
|i -e7'<':' 4? .0-4^.0.0 SMU, DiLUIED ETOOI'- 1:50 flNHLviT LHK Dlc-C'HRY; £N
^Di: TIC
Figure B-17.
iO . v 31 .0 3£ . 0 ij: .C i*t .C 35.0 3t". .0 37 . 0 36 .0 3* .0 40 , u
Total ion chromatogram of PCB solution used to spike
open-vessel samples spiked with internal standard.
79
-------
File >B7330 DILUTED STOCK 1:50 RNfiLYST LHK D12-CHRYSENE $ £0 Scan 11331
Epk fib 9999.
-j
IS
40
Bpl< Pb 9999.
,
-i
J
' f ' r-
40
File >B!6DB
Bpk Hb 9999.
SUE'.
1
63 '"'5 93 in l£6 135
/ '' '' """-V '"--
' "'- , - "- ' -" -- 1 7" r. r.l I ll'
, -f .,.[...,.. 1 | 1 . 1 ,..
8.0 120
25.06 triin.
£££
F. £ _--*
1 6 £ 1 S 6 1 9 6 £ o o
, .,-- /: / . --
t
, L
160 £00
1,1 ' - B i p h * n y ! , £ , 6 - d i c h 1 o r o - S c an 3 0 £ 0 6
0 .p.0_ min .
1
, | , i . , . . . | . . , , . , . | . .
80 120
ii^'c:
.r'1" *"* v-
186
r ,-i
i i ....... | . . u
1 fc 0 £ 0 0
1 ,1 '-Bipheny 1 , 4 , 4 ' -dichloro- Scan 80196
0.00 rn i n .
2££
15E .--
] 43
.- 1 -*'
1 rt
T \J'
File >EI6BB
B p k H b 9 999 .
-|
1 51
63 7* «3 m 1£6 136
..;' .,j'i. .. i. ... "">.. "'";. >' .,1
SO 120
i ,1 '-Biphstvvl , 4,4'-d
1
63 ?5 93 HI ^ 136
16£ 136 Pn7
L -"~~" ,' -"~~~
E
1 r r-i
160 £00
ichloro- Scan 80198
0 . 00 mi n .
£'££
52 r"" i-
16 £ 186
~^[r"1-)i' T ;r i "i "Y
Figure B-18.
Mass spectrum of first peak in spiking sample
chromatogram (Fig. B-17). Matches ranged in probability
from 0.88-0.96; library did not contain the DCBP congener
used in this study.
80
-------
File >B7330 DILUTED STOCK 1:50
Bpk fib 9999.
j 50 74 92 110 12?
40 SO 120
1
le >BIGDB 1 ,l'-BiDhenvl
Bpk fib 9999.
T
j 74 92 11° 128
40 80 ISO
File >BI6DB 1 ,1 '-Bipheny 1 ,
Bpk fib 9999.
-j
50 ?4 92 110 123
ni { -X ,'.,,.' ,<,
u i i ' i ' i i i
40 SO 120
File >BI6DB 1 ,1 '-Biphenvl ,
Bpk fib 9999.
1
40 3 w 12 0
flNfiLYST LHK D12-CHRYSENE 9
SUB
220
150 170 ^j34 ^ 2£5' 254
160 200 £40
, 2 ,3 ,4 ' , 6-tetrachloro-
220
150 M
160 £00 £40
2,2' ,6,6 '-tetrachloro-
220
150 ±70 ^184 ^ 225 255
it,|.i.|rjS|...[i..|..l^,
160 200 £40
£,£ ' ,3, 3 '-tetrachlc.ro-
159 1S9 £20 257
i * i i i i i i i i i i i i i \ i i j i i 1 j i
160 £00 £40
20 Scan 1393
29.67 rnin.
292
I, L
2SO
Scan 100720
0 .00 min .
£92
.--' ,_
1
280
Scan 100684
0 .00 min .
292
,--' ,
£S3
L_L
£30
Scan 100705
0.00 min.
292
")_
1 | 1 1 1 | 1 -U
£SO
Figure B-19.
Mass spectrum of second peak in spiking sample
chromatogram (Fig. B-17). Matches ranged in probability
from 0.59-0.79; library did not contain the TCBP congener
used in this study.
81
-------
File
Bpk
\_
U
Bpk
File
Bpk
File
Bpk
>B7330
fib 9999
DILUTED STOCK 1
:50
fiNflLYST
LHK D12-CHRYSENE II 20 Scan
SUB
32
.42
1548
min
m
360
-\
1 «
J n>
40
>BI3DB
fib 9999
74 109127
7 / (._'_. r
80 120
3,4,5,3'
145
160
,4',
132
'
£18
~--s
200
22£ £55
V"~~ _/
240
£90
|^ ^
280 320
5'-Hexachlorobip'henvl
3*7
~~
111
360
l_
C
I n
Scan 11515
0.00
min
9
,
360
H
n ' i
40
>BIGDB
fib 9999
1 . ^ --
j ./. .». ,. .1 4 - r - C- ^
r ' 1 ' i T I ' l
80 120
145
s*
160
l.l'-Biphenyl, 2
.
130
/
T" '
218
£00
,£',4,4'
£22 254
£40
£90
280 3£0
,6,6'-hexachloro-
388
1
I
360
Scan 115160
0
.00
mm
,
360
b
40
>BI6DB
fib 9999
^>3 109 ]_27
T / ..... 4 riy: i'
r * i i 'i 'i
80 130
0 0 S -3
"- , - , " >
145
**
160
4,4 '
130
/
't ' '
218
200
222 £53
£40
290
if 325
|L ^,-
i * i i
£80 320
,5-Hexachlorobiphenyl
327
-^
i
I
1 ' 1 ""
360
Scan 115163
0 .00 min .
360
j
U 'l i
40
1 l I ' 1 l 1 ' I
80 120
r i
160
200
i \ ' i '
£40
i ' i ' i
£30 3£0
i '
I
V '
I
360
Figure B-20.
Mass spectrum of third peak in spiking sample
chromatogram (Fig. B-17). Matches ranged in probability
from 0.91-0.94; library did not contain the HCBP congener
used in this study.
82
-------
File
Bpk
-.
le
Bpk
File
Bpk
>B7330 DILUTED STOCK 1:50 RNRLYST
fib 9999.
1 «L 7 108 1
n 1 ,1 1 1 ^~~" 1
- V 1 " ' ' f""' 1 ' 1 '' 1 '
80 120
SUB flDD NRM
/5° 168 ^i 227
160 200 240
LHK D12-CHRYSENE BIGDB fi-CONIDEHDRIN DIMETHYLETHER
Qb 9999.
] 55 91 128 1
01..i<.TT./__. /..r.
30 120
51 139225 233
i n^ T i r " i
160 300 340
269
I 299 3£5
1 * T 'H ' 1 ' \
280 320
>BIGDB B-CONIDEMDRIM DIMETHYL ETHER
fib 9999.
_i
| 55 91 115 151 189225 238
File
Bpk
v ' i i ' i i i
30 120
>BIGDB
fib 9999.
1 ' 1 1 I ' t
160 200 240
269
1 299 325
i \ i "i "i i i i \
230 320
T
1 i ' i
360
32.56 min.
382
406C
400
Scan 117916
0.00 min.
384
353
' 1 ' 1
360
337 E
400
Scan 117915
0.00 min.
334
353
i i
360
Excel sinine
' i_
387 [
1 ~ rn
400
Scan 117342
0.00 min.
384
j 153 186 200
J
U i | i | i | i | i |
80 120
160 £00 £40
324
£80 320
353
360
335 F
i p i i i 0
400
Figure B-21.
Mass spectrum of fourth peak in spiking sample
chromatogram (Fig. B-17). Matches insufficient for
identification (probability 0.11-0.12).
83
-------
File
Bpk
>B7330
fib 9999
DILUTED STOCK 1
] 52 66 90 1'?6
"N
. le
Bpk
File
Bpk
File
Bpk
l-l 1 , , / ,
PR I POL
flb 9999
j 44
pii
>BIGDB
flb 9999
.1?
>BIGDB
fib 9999
1 t?
j-~
.-,il., . ..
/ "J"!
, . , MQ, . . , .
:50 RNflLYST
SUB
120 132 154
...^ .../ ^
r : | , 1 1 | 1 1
120
LHK D12-CHRYSENE (? 20 Scan 1653
34 .28 min .
156
s*^
160
180 194 212
/ / /
200
Chrysene d-12
.
66 90 1-96
' 1 '' ' ' 1 ' ' ' ' ' 1 ' ''
30
Benzenamine ,
B
69 77
/ ^
\ ' ' i i
SO
2-Propen-l-one ,
65 79 93
/ / .-'
\ I1 'l"l | 1 '!" 1 | 1
30
1?^ 132 154
' 1 i
120
156
I ' *
160
184 194 203
^ / f .
i ' i ' i
200
4-<6-methyl-2-berizothiazolyl >-
118
/ 123
120
160
l-<2-hydroxyphenyl
1^ 123 147
l! """""^ i
r i ' I1 i i i"'i i
180
162
f
160
j ... I ... |
200
240
232 t
.. .T/.l..ll, Frj
240
Scan 437
0 .00 min .
240
236 F
240
Scan 85422
0.00 min.
240
:L
240
)-3-<4-hydroxyp Scan 35532
0.00 min .
1?3 194 211
fit
200
240
2£3 || E
^ (1 C
i -' rQ
240
Figure B-22.
Mass spectrum of fifth peak in spiking sample
chromatogram (Fig. B-17). Probability of 0.93 for match
with library spectrum of chrysene-d^-
84
-------
Fil« >B
78000-
70000-
*BOOO-
60000-
55000
50000-
45000-
40000-
35000-
30000-
88000-
&0000-
1BOOO-
10000-
5000-
ec
Fii* >e
75000-
70000-
1 *ooo
60000
55000-
50000-
4500O
40000-
35000-
30000-
86000-
eoooo-
18000-
1000O-
8000-
0-
30
7314 45.0-449.0 UIU. T2-7S RNRLYST LHK D1Z-CHRYSEN
WC TIC
990 i 1000 t 1100 t 1200 § 1300 t 1400
1
I
1
.0 ei'.o ea'.o 23.0 24.0 ss
i
8
5
1
3
A ir
" r --'if- J
0 26.0 27.0 26.0 29'. 0 30.0
7314 45.0-449.0 MU. T8-72 OHflLYST LHK D12-CHRYSEM
ODC TIC
15(00 t 1600 f 1700 ( 1800 f 1900
2
14
7-1819
-M/^
0
'.0 31.0 32.0 33.0 34.0 35.0 36.0 37.0 38.0 39.0 40.0
Figure B-23.
Total ion chromatogram of 72-h, replicate 2 sample.
85
-------
File
Bpk
Q
Bpk
File
Bpk
File
Bpk
>B7314 T2-72
fib 9999.
1 50 63
ol ., 1
40 60
>BIGDB
fib 9999.
-j
j 51 63
Qi ^J, ^i
40 60
>BI6DB
flb 9999.
1 51 63
0 ~*~~pi -i'
40 60
>BI6DB
Rb 9999.
1 51 63
40 60
RNflLYST LHK D12-CHRYSENE 1? 20 Scan 996
SUB
-_ 77 94 102
80 100
126 127 151
120 140
1 ,1 '-Bipheny 1 , 2-chloro-
'-^ 77 94 102
80 100
126 127 1?^
"^r-" y
120 140
1 ,1 '-Biphenyl , 2-chloro-
76
--^. 77 94 102
. | . 1 1 . | , . i . , . 1 . r , i . . , , ,
80 100
126 127 1?^
120 140
1 ,1 '-Bipheny 1 , 4-chloro-
7^ 77 94 113
80 100
126 127 15_^
120 140
22.63 win.
188
152 """
,r*t
160 180 200
Scan 65656
0 .00 min .
188
152 ^i i-
164 .--^ t
160 180 200
Scan 65655
0 .00 min .
188
"" 164
i . r n
160 180 200
Scan 65625
0 .00 min .
188
152 ' -
173
't '
190 t
i In
160 180 200
Figure B-24.
Mass spectral matching for peak 1 (Fig. B-23) indicated
0.94-0.99 probable agreement with isomers of
monochlorobiphenyl.
86
-------
File >B7314 T2-72
Bpk fib 9999.
1 51 63
nl ...L. ,../...
40
.'e >BIGDB
Bpk fib 9999.
j
\j i i i i i i i
40
File >BI6DB
Bpk fib 9999.
^1 / /
40
File >BIGDB
Bpk fib 9999.
1 51 63
.-. 1 .. . u -.1
U '| i I1 i I1" i i
40
flNRLYST LHK D12-CHRYSENE P 20 Scan 1133
SUB £5
75 93 ±11 126
f, ( , ,^ , ,>
80 120
1 ,1 '-Biphenyl
80 120
1 ,1 '-Biphenyl ,
75 93 in 126
80 120
1 ,l'-Biphenyl
/ 9,3 HI 126
.11 ( .. >, V,
80 1£0
152
135 1 170 186196 220
160 200
.06 min .
222
^^^-
I , fo
, 2 ,6-dichloro- Scan 80806
0
152
' I
j. r 1
160 200
.00 min.
£££
I I
4,4'-dichloro- Scan 80196
0
152
136 1 162 186 207
/ ,,1. ^ / ^- 1
160 ' 200 '
.00 min.
222
r.
' n
, 3,4-dichloro- Scan 80199
0
152
135 161 186196 £09
/ |ll. - .' / X- I
160 £00
.00 min.
222
1 , 1.
Figure B-25.
Mass spectral matching for peak 2 (Fig. B-23) indicated
0.93-0.96 probable agreement with isomers of
dichlorobiphenyl.
87
-------
File
Bpk
>B7314 T2-72 ONflLYST LHK D12-CHRYSENE 9 20 Scan 1163
Rb 9999. SUB
1
1 50 63 /5 93 111 126136
J / / i ..{....(. f ' ii
60 ' 80 100 120 140
25.59 min.
222
53 j-**
168 186 1 f
/ f \\f Fc
160 180 200 220
e >BI6DB l,l'-Biphenyl , 2 ,6-dichloro- Scan 80206
Bpk
flb 9999.
0.00 min.
222
152 ^-
File
Bpk
1
0 i i | i i | i i i i i ' I ' | ' i
60 80 100 120 140
ll f
. M.i r n
'i | i | i | ' | i | i | i |"'l' | i1 O
160 180 200 220
>BI6DB l,l'-Biphenyl , 2 ,4-dichloro- Scan 80208
flb 9999.
0.00 min.
222
1 152 f -i-
j 51 63 75 93 m 126i35 \ 170 186196 220 I f
J / / / ,' i. .ft .1. / t. / ^11- L
File
Bpk
W r- [ i | i | | | | | | | | |
60 SO 100 120 140
i i i i i i i i w
160 180 200 220
>BI6DB l,l'-Biphenyl , 4 ,4 '-dichloro- Scan 80852
flb 9999.
0.00 min.
152 ZSZ
1 70 /5 ^ "-1 126 139
' Vo' ' ' SO' 'lOO ' 'l20 ' 140
165 185 201 I F
i ('. ^ |-. / llljl. FQ
160 180 200 220
Figure B-26.
Mass spectral matching for peak 3 (Fig. B-23) indicated
0.87-0.96 probable agreement with isomers of
diclorobiphenyl. Relative retention time distinguishes
this isomer from that shown in Fig. B-25.
88
-------
File >B7314 T2-72 flNflLYST LHK D12-CHRYSENE 9 20 Scan 1277
Bpk fib 9999. SUB
J
47 68 69 97 3.15 141
f "Tir" i » "^i n
40 80 120 160
27.61 min.
204
/
236 F
200 240
.e >BIGDB 2-Hydroxy-5-chlorobiphenyl Scan 72861
Bpk flb 9999.
j
rJ
51 63 75 89 115 126 ±/39 151 *-6i
^ / . / ../ ^j / LI **- r
40 80 120 160
0.00 min.
204
3 189 207 E
/ At - f0
200 240
File >BI6DB Cl ,1 '-Bipheny l]-4-ol , 4'-chloro- Scan 72880
Bpk flb 9999.
j
11R 141 170
39 55 69 85 11^ 119 / 151 /
v- ,...,...,...,...,.., , , . , ,
40 80 120 160
File >BISDB 2-Chloro-4-bipheny lol
Bpk Ob 9999.
^
j
139
51 63^69 39 ng 126 ' 151 16J
40 80 120 160
0 .00 min .
204
184 210 E
' ' ' 200 ' ' ' ' ' 240 °
Scan 72902
0 .00 min .
204
/ i_
3 208 E
200 240
Figure B-27.
Mass spectral matching for peak 4 (Fig. B-23).
Identification as a hydroxymonochlorobiphenyl isomer with
probability 0.36-0.41.
89
-------
File
Bpk
>B7314 T£-72 RNflLYST LHK D12-CHRYSENE 9 20 Scan 1284
flb 9999. SUB
"j 50 74 J? 93 111 129 «°
J f. ... !>J ... ,L^.. . .y^.....-^. Jl ... . ]. .
£7.73 min.
220 1 F
-" I,L F«
' | 1 1 1 | 1 1 1 | 1 1 1 | 1 1 ' 1 | 1 1 > | 1 1 1 | 1 1 1 | 1 1 1 | 1 1 1 | 1 1 1 | 1 1 M
40 80 1£0 160 800 240
e
Bpk
>BIGDB 1, I'-Biphenyl , trichloro-
flb 9999.
'H 186
\ 50 74 JJ> 93 m ±29 150 -""
{ if . ... ^il i it. '....r~^ .«. .
Scan 90258
0.00 min.
256
- f f
,- 11 F.
V | | | | | | | | . . | . . . | . . | V
40 80 120 160 200 240
File
Bpk
>BIGDB 1, I'-Biphenyl , 2 ' ,3 ,4-tr ichloro- Scan 90269
flb 9999.
186
1 ->* *s
j 50 74 75 JK 11± 12g U>Q 1?2
0.00 min.
256
_^^-
207 2£0 E
/ ^ 111 L
iii "i i i i i i i""i" i i1 u
40 30 120 160 200 £40
File
Bpk
>BIGDB 1, I'-Biphenyl , £ ,3 ' ,5-tr ichloro- Scan 90268
flb 9999.
_j 186
1 50 74 J.5 93 m 129 150 1?0 ^
nl .if if .. rT l.-^....-^ .1 , / 1,
0.00 min.
258
^ 1
215 220 054 F
^^ ^ " -LI.. r«
40 SO 1£0 160 £00 £40
Figure B-28.
Mass spectral matching for peak 5 (Fig. B-23). Excellent
match as isomer of trichlorobiphenyl, probability 0.96-
0.99.
90
-------
File >B7314 T2-72
Bpk fib 9999.
] 50 92 11£?
j / --5 i /
nl j ki..i< i l i .1 >
80 120
flNHLYST LHK D12-CHRYSENE BIGDB 1 ,1 '-Biphenyl , te trachloro-
Bpk fib 9999.
292
220 ,
1 50 9^ 1>° 146^ ^50 184 | 254J?5 it 29?
0 ' t '|* '"lf'*f 1 * I
80 120
160 200 240 280 320
File >BIGDB 1 ,1 '-Biphenyl , 3 ,3 ' ,5 ,5 '-te trachloro-
Bpk fib 9999.
1 9£ ^ 127
.-.I ji n .. .{
80 120
292
220 /
^° 184 l' || 296
160 200 240 280 320
File >BI8DB 1 ,1 '-Biphenyl , 2 ,3 ,4 ' ,6- te trachloro-
Bpk fib 9999.
1 92 11012E
,-,1 ^~; . f . /
U 1 | 1 | I1 1 f I I
80 120
?9?
220 f
I J.50 185 I || 296
160 £00 240 280 320
20 Scan 1301
£8.04 min.
37i
' '360' ' °
Scan 100686
0 .00 min .
I
360
Scan 100695
0 .00 min .
f
360
Scan 100720
0 .00 min .
I
1 i 1 ' 1 0
360
Figure B-29.
Mass spectral matching for peak 6 (Fig. B-23) indicated
0.87-0.95 probable agreement with isomers of
tetrachlorobiphenyl.
91
-------
File >B7314
Bpk flb 9999.
3 w
.' '1 | V t
.e >BIGDB
Bpk flb 9999.
0' I1' '
File >BIGDB
Bpk flb 9999.
j
File >BI6DB
Bpk Pb 9999.
T2-72 RNPLYST LHK D12-CHRYSENE 1? 20 Scan 1308
SUB POO NRM NSP 28.16 min.
280 290
/ / ,
SB 111 150 255 i f
/ / 121 / 166 188 233 ' j 300 322 t
80 ' ' f 120rtn 160 '
200 ' 240 ' 280 ' 320
TETRPCHLOROBIPHENYL Scan 100676
0 .00 min .
220 292
/ ,
98_ 110 123 / 185 ^£55 I 2g6 f
if, i .1 .|l ^"",, ,4 ^ i \ ll. lilr^' hn
80 120 160
l,l'-Biphenyl , 2,2
149
110 123 f 184
80 1£0 160
1 ,1 '-Biphenyl
?2 ""is? S50 184
.{, j. ./ rf ^.
1 fill ' ' ' 1 ' ' 1 1 ' ' 1 1 ' ' 1 1 '
80 120 160
£00 240 280 320
',4,5'-tetrachloro- Scan 100710
0.00 min.
220 292
/ / .
1 255 I, £96 f
1, f \l\r-- F0
200 240 £80 320
, tetrachloro- Scan 100686
0 .00 min .
£20 2J2
' 255 || f7
194 £54 ^" . i £97 t
£00 ' £40 ' £80 ' 3£0
Figure B-30.
Mass spectral matching for peak 7 (Fig. B-23). Moderate
agreement with isomers of tetrachlorobiphenyl
(probability 0.52-0.79)
92
-------
File >B7314 T2-72
Bpk Rb 9999.
1 5°
QJ ' .. it .
40 80
.e >BI8DB 1
Bpk fib 9999.
1
n' i '1 L
U | i i i | i i 1 | i
40 80
File >BIGDB
Bpk fib 9999.
j 50 75^
n u i Jl
40 80
File >BIGDB 1
Bpk fib 9999.
_i
] 50 74
-\ ' ^,
u i i r i i i > 1 i i
40 80
flNRLYST LHK D18-CHRYSENE 9 20 Scan 1393
98 110 12?
{. .. h .{.
120
,1 '-Bipheny 1
92 1/0 128
120
SUB
£9.67 win.
292
220 r" h
150 17Q 184 -X 225 256 885 I, f
r( { .fT . ..L-< ./ -^ ill, F«
' 166 ' 266 ' '
i i i i i i i i i i i i i i -
240 280
, 2,3' ,5,5'-tetrachloro- Scan 100714
0.00 min.
£92
220 \' h
150 184 x. 1 b
.( r (.. IL L
160 ' 200
240 ' 280 ' ' ~
l,l'-Biphenyl , tetrachloro- Scan 100687
92 110 128
t i, ,. ii. .( .
120
,l'-Biphenyl
820
150 184 ^
160 £00
0.00 min.
292
f |_
i, r 2,si L
\ i i | i i i"| i i i | i i"i' | .' g
£40 £80
, 2,2',5,5'-tetrachloro- Scan 100700
0 .00 min .
292
£20 ^ L
92 110 123
i i i i 1 i i i i i
1£0
,7 ''" " 4 ^
160 £00
i 225 £/55 I E
(^ u, 11., F0
£40 280
Figure B-31.
Mass spectral matching for peak 8 (Fig. B-23). Excellent
agreement with isomers of tetrachlorobiphenyl
(probability 0.96-97).
93
-------
File >B7314
Bpk flb 9999
h
nl , .
V 1 i l r
40
. .e >BIGDB
Bpk flb 9999
150
-/
1 ' '
40
File >BIGDB
Bpk fib 9999
j 50
40
File >BIODB
Bpk flb 9999
0- nr-
40
T2-72
74
1 /
80
74
SO
74
1 ' ' 'so
1
74
80
109 127
. ,r^ . L r
120
ONflLYST LHK D12-CHRYSENE
-------
File >B7314 T2-72 RNRLYST LHK D12-CHRYSENE 9 20 Scan 1436
Bpk fib 9999.
1"
^ i f\
J \ \ f'|
\
. .'e >BISDB
Bpk Rb 9999.
-i
i |
File >BI6DB
Bpk flb 9999.
J
File >BI6DB
Bpk fib 9999.
J
SUB
73 98 110 149i59
4 t J '1 '' /' f
80 120 160
l,l'-Biphenyl
74 109 127 145
80 120 160
l,l'-Biphenyl , 2,2'
109 127 135 163
Y "i -1 1"
80 120 160
l,l'-Biphenyl , 2,3»
80 120 160
184
' 266'
256
219 \ 29°
(1. ")
240 280
, pentachloro-
184
/
200
,3, 4', 5
184 gj
/
\. . .
200
254
T (L
1 1 ' i ' i i '
240 280
'-pentachloro-
,1 2r 2*i
£40 280
,4,4' ,5-pentachloro-
£00
256
240 £80
30.43 min.
326
306 | f
\*\ 10
'320 '
Scan 108107
0.00 min.
326
f h
,1,0
1 . . . 1 I'l i [ U
320
Scan 108115
0.00 min.
326
ft :o
'320
Scan 108108
0.00 min.
326
1, t
320
Figure B-33.
Mass spectral matching for peak 10 (Fig. B-23)
agreement with isomers of pentachlorobiphenyl
(probability 0.36-0.58).
Moderate
95
-------
File >B7314
Bpk flb 9999
1 58
J ^
e >BI8DB
Bpk flb 9999
j
File >BIGDB
Bpk flb 9999
1
nl
File >B!SDB
Bpk flb 9999
T2-72 flNflLYST LHK D12-CHRYSENE 9
SUB
75 86 110 137 iJ3186
I,'!../ ^1 I L
80 120 160 £00
20 Scan
1451
30.70 min.
"" M /'
i i i i-i if
240
3-t1ethoxy-4,5)4'-trichlorobiphenyl
£86
i
i i i I i'"r i
280
Scan 99756
0 .00 rain .
173
100 111 13^ 143 { 186 207
~^r N/. .,. ,1. JT^i. .1,1 .1^ . '. ...f.
[ ' ' ' \ ' ' | * !(. |..f|l..|...[
80 120 160 200
243
221 f
240
3-nethoxy-2,£' ,5 '-trichlorobipheny 1
286
/
I
iii) i»'h
280
=.
Scan 99782
0 .00 min .
173
100 110 13^ 15° \ 186 207
"^. 4- y.,.^.,1- H- ,- -,'4 .1^, (.. ',- -i.j.
80 120 160 200
236245
' '"
i Vl-K
240
PHENYL-DI-D5-PHENYLPHOSPHIHE OXIDE
1 54 77 J32 W6 128 159 163 190 206
SO 120 160 £00
240
288
273 1
..' 11,1.
i i i 1 I'-'f i
280
^0
Scan 99203
0 .00 min .
286
...1.
£80
Figure B-34.
Mass spectral matching for peak 11 (Fig. B-23). Moderate
to poor agreement with isomers of
methoxytrichlorobiphenyl (probability 0.27-0.42).
96
-------
File >B7314
Bpk fib 9999.
1"
J r,
. e >BI6DB
Bpk fib 9999.
H
J "
File >BIGDB
Bpk Ob 9999.
-\
0 ,?M-
File >BI6DB
Bpk fib 9999.
^
J ,
T2-72
73
f
SO
RNOLYST LHK D12-CHRYSENE @
SUB
105 12714Q 184 218 254
i r^ \ f f { \\
120 160 200 240 280
1 ,1 '-Biphenyl , pentachloro-
20 Scan 1516
31 .85 min .
326
ill
||]|
320
-
-0
Scan 108107
0 .00 min .
74
j ..
80
254
109 127 145 ^?4 218 / 291
120 160 200 240 280
1 .1 '-Biphenyl , pentachloro-
3ii6
I
1
1 1 ' ' ' 1 ' ' '
320
u
'
Scan 108104
0.00 min.
326
74
. A , '»
80
1
74
80
254
109^ 127 ±45 163 184 218 / 2g9
120 160 200 240 280
254
1Q2 ±?! 146 163 I?4 218 ,<
, >| 1i / {. f ' l|i
120 160 £00 £40 £80
, ,
i I i i i I i i i
320
_
\
Scan 108109
0 .00 min .
326
{
320
-n
Figure B-35.
Mass spectral matching for peak 12 (Fig. B-23).
Identified as isomer of pentachlorobiphenyl (probability
= 0.83-0.86).
97
-------
File >B7314 T2-72 flNRLYST LHK D12-CHRYSENE 9 20 Scan 1535
Bpk flb 9999.
1 4&
Q| ,"\ ,
40
. .e >BIGDB
Bpk fib 9999.
jr
40
File >BIGDB
Bpk fib 9999.
.1...
40
File >BIGDB
Bpk fib 9999.
1 39
40
SUB
62 86 101 us 139 173 209
l i it it i
74 85 1*° 132 144 155 176 £/°1 228
80 120 160 200
5-BROMO-3-<4-PYRIDYL>INDOLE
139 167 182 193
' 80 ' 120 160 ' 200 '
!H-PyrroloC2 ,3-b]pyridine, 3-broi»o-2-phe
63 90 96 117 139 166 ***
/, WIT* ,'"n i ^> , A i , , i 'i', , i , A i , , ,
80 120 160 200
238
{
240
237
(i,
i iiy 1 1
240
240
nyl-
240
32 .19 min .
278
H
'i FO
I | i I "f | 1 V
280
Scan 95852
0.00 min.
274
lif.
i 1 > r"f | t y
28^
Scan 95791
0.00 win.
272
,u.
280
Scan 95769
0 .00 min .
272
, u
280
Figure B-36.
Mass spectral matching for peak 13 (Fig. B-23)
Identification uncertain.
98
-------
File >B7314
Bpk fib 9999
_i
j &^
40
. e >BIGDB
Bpk flb 9999
J
40
File >BI6DB
Bpk fib 9999
_j
^ j i
40
File >BI6DB
Bpk flb 9999
.1 ,
40
T2-72 flNflLYST
SUB
74 109126 ^145 182 218
80 120 160 200
LHK D1S-CHRYSENE 9 20 Scan 1546
290
222 253 .' 325
240 280 320
2,2' ,3,4,4 ' ,5-Hexachlorobiphenyl
80 ' 120 ' 160 ' 200
l,l'-Biphenyl , 2,2', 3, 4,
B
218
162 I?2 "--
,,,,.,,,,,,,,,,,,, ,.i», ft,, fs,,
80 120 160 200
240 280 320
4 ' ,5'-hexachloro-
290
222 254 / 323
240 280 320
32.38 min.
360
i t
^? i I
360
Scan 115163
0.00 min.
360
If.
1 1 1 I'fw-i i1 o
360
Scan 115162
0.00 min.
360
"' 1 f.
360
l,4,5,6-Tetrachloro-7-<2,5-dichlorophenyl>bicycloScan 115179
0.00 min.
80 ' 120 ' 160 ' 200
240 280 320
360
, u
360
Figure B-37.
Mass spectral matching for peak 14 (Fig. B-23).
Excellent match with isomers of hexachlorobiphenyl
(probability = 0.86-0.95); moderate match (probability
0.50) with tetrachlorobiphenyl.
99
-------
File >B7314 T2-72 flNflLYST LHK D12-CHRYSENE 1?
Bpk flb 9999. SUB
]o*?o
53 85 103 135 147 171 207 216 243 , 307
,?,.... ~.^ L. ...^L ./... L . ^s- .>{. iL . ^
40
. .e >BIGDB
Bpk flb 9999
.1
40
File >BIGDB
Bpk flb 9999
0- i i i
40
File >BIGDB
Bpk flb 9999
-\
j 43
0iUi
v i 1 1
40
80 120 160 200 240 280
20 Scan 1554
32.53 min.
342
/ ,
ll I
320
2 ,5-Dibromo-3 '-me thoxybipheny 1 Scan 111808
0.00 min.
342
r 7 ili f.
80 120 160 200 240 280
2 ,5-Dibromo-4 '-me thoxybipheny 1
29<
/
80 ' 120 ' 160 ' 200 ' 240 ' 280
320 '
Scan 111809
0.00 min.
342
' T i 1
h p "" o
Methanone, <5-bromo-6-methoxy-2-naphthalenyl JphenScan 111715
0 .00 min .
340
\ t"
77 105 ±26 156169 189 220 ~J t
.{{.{{.' ^- ( l| IF.
80 ISO 160 £00 £40 £80
320
Figure B-38.
Mass spectral matching for peak 15 (Fig. B-23). Library
search yielded poor to moderate tentative identification
as methoxy derivative of brominated biphenyl.
100
-------
File >B7314 T2-72 ONOLYST LHK D12-CHRYSENE (? 20
Bpk Ob 9999.
Scan 1559
SUB 32.61 min.
342
1 53 87 103 ^21 1J1 208 21? 2«3 */ 305 31;
01 ( ., "]i ., .|' ,1.. [ )l "^ r-" ll, ll ^|, r^
""'''' 80 ' ' 120 ' 160 ' 200 ' 240 ' 280 ' 320
Mil
.e >BieDB 3.4-DIMETHYL-6-PHENYL-1 ,60 .LfiMBDO .*4-DISELENfi-5 ,Scan 112166
Bpk Ob 9999.
\ !
n' i1- -i
0 .00 min .
342 F
104 159 173 235 250 263 329 / fc
i w
80 120 160 200 240 280 320
File >BIGDB
Bpk fib 9999.
2 ,5-Dibromo-4 '-methoxybipheny 1 Scan 111809
0 .00 min.
342
1 *r Ti 1
V 1
1 i i i I i .. . ....... . ... . ... I i , . I i i i , i . i i i . i I i i i-f i i i , i'
i 11 r 0
1 ' i * ' y
80 120 160 200 240 280 320
File >BISDB l-<7 '-BROMO-4 '-METHOXY-3 '-METHYL-2 ' .3 '-DIHYDROBENScan 112005
Bpk Ob 9999.
j
0 .00 min .
327342
271 £99
if /
SO 120 160 200 240 280 320
'f
i» fi
1 i | r i V
Figure B-39.
Mass spectral matching for peak 16 (Fig. B-23). Library
search yielded poor tentative identification as methoxy
derivative of brominated biphenyl (probability = 0.25-
0.36) .
101
-------
File >B7314 T2-72
Bpk fib 9999.
1 55 85
J / ---
n1 . ( V 11,. (
40 80
. .e >BIGDB
Bpk fib 9999.
H
.1
y 1 i i | i r i | i i
40 80
File >BISDB
Bpk fib 9999.
1 85
40 80
File >BISDB
Bpk fib 9999.
1 85
ni 1
V | 1 I 1 | 1 1 1 | 1 1
40 80
flNRLYST LHK
SUB
97 153 iji 2Q8
/ ' i --^
i . j i 1 <
120 160 200
D12-CHRYSENE 9
306
~~--~-.
£/4± 870 1
,,.{ |
240 280
2 ,4 ,6 ,3-Te trachlorodibenzof uran
153 171
1 1
120 ' 160 ' 200
306
|
II
i . l". i . 1 , , i 1 , i i 1 1
240 280
2,3,6 ,8-Tr if luorodibenzof uran
171
132 153 /
120 ' ' 'l60 ' 200 '
306
r ii
, . { , . . | i i i | i i r | 1
240 280
1, 2, 7, 8-Te trachlorodibenzof uran
153 171
< 1
1£0 160 £00
306
|
ii
£40 £80
20 Scan 1569
32.79 min.
309 ,
i'"'' 344 ;
1 ^ "0
320
Scan 103979
0 .00 min .
303 h
r^ 314 :
320
Scan 103978
0 .00 min .
!308 h
i^-^'t,
320
Scan 103980
0 .00 min .
!""" 314 1
i ~ ^ *~. L " n
1 i | i i i ] i i t
-------
File >B7314 T2-72 RNflLYST LHK D12-CHRYSEHE
Bpk flb 9999. SUB flDD NRM HSP
11000;
10000-
5000;
800*
7000-
6000^
5000;
4000;
300*
2000-
100*
/v_
55 /"
77
' 209
126 154
X X
! 1 ! B 1 JHjBfl HjBBUHjMjKJKAUBflUBJMflUBJW
i | i | i | rTT~i i i T | i |
80 120 160 200
211 239
UHJUBUUHJH^KMUI
r ~~ r~ r '
240
272
/
295
x
m^f^fmfmi^mmimmmi^mi
\ r r~ T~
230 320
9 20 Scan 1587
33.11 win.
rllO
356
41
*U
3
y-
380
s
URJUUHflUU
60 40C
I
;100
-90
-80
-70
-60
-50
-40
-30
-20
-10
-0
Figure B-41.
Mass spectral matching for peak 18 (Fig. B-23). Library
search yielded no matches.
103
-------
File >B7314
Bpk flb 9999
1"
nl f
. /e >BIGDB
Bpk flb 9999
J
File >BI6DB
Bpk fib 9999
_,
1
j
n 1
File >BIGDB
Bpk flb 9999
I
T2-72
73 121 122
80 120
4-Methoxy-2,
9 135
80 120
3-Methoxy-2,
HO 145
/ -x.
..i* i.. -, . ...
i i i i
SO 120
C2.0
80 120
171
160
5,2
171
--.
160
5,2
171
160
flNRLYST LHK D12-CHRYSENE i? 20
SUB
183 206 £t1 / 3
n ' A 1*1 ^'
260' ' ' 240' ' ' ' '280 320'
' ,4' ,5'-pentachlorobiphenyl
241 313
173 205 -^ 251 271 / 3
J I'fi i'i | f'i i«f i i i fl ^-| i v i1) v i I
' I ' ' 'T 0
360
Scan 114190
0.00 min.
356
/ ,
173 206 211 256 277 3}3 321 i, :
" ' ' \ ' f i ^
i.fi..lr..(-...li.-[l.\ti'lii.l^l'\{..
200 240 280 320
. 2]Me tacyc 1 ophaned i ene
160
182
|".i TTI ~IT r\ t
£00 £40 £80 320
1 '
'H- o
1 l ' r 1 ' v
360
Scan 114166
0.00 min.
356
/ ,
339 f
f r Fo
360
Figure B-42.
Mass spectral matching for peak 19 (Fig. B-23). Library
search yielded good tentative identification as isomer of
methoxypentachlorobiphenyl (probability = 0.73-0.93)
104
-------
File >B7314 T2-72
Bpk flb 9999.
| 66 90 106
80 1<
. .e PR I POL
Bpk fib 9999.
1 44 90 106
n1\, ,11,1 . y "i"!" rV 'f^
80 1<
RNflLYST LHK Dl
SUB
I20 156 180208 232
/ / "~>y ^ 1.
2-CHRYSENE (» 20 Scan
34.26
240
272 342
20 ' 160 200 240 ' 280 ' 320
Chrysene d-12
^?° 156184 208 212
JO ' '' 'l6b' ' '200 '' 2'
File >BI6DB NaphthoC2 ,1-b :7 ,8-b ' ]dif uran , 1,2,
Bpk flb 9999.
1 211
j 63 76 99 115 152 181 . _ 225
Scan
0.00
£40
242
JO ' 280 ' 326"'"
1652
min .
L
437
min .
I
9,10-tetrahydro Scan 85492
0.00 min.
240
242 E
i r r,
SO ' 120 ' 160 ' 200 ' 240 ' 280 ' 320
File >BISDB £-Propen-l-one . l-(2-hydroxyphenyl >-3-<4-hydroxyp Scan 85532
Bpk flb 9999. 0.00 min.
240
L^3 ^J 1/°7l/Cl "? 165 211 2f3 || 242 f
SO 1£0 160 £00 £40 £80 320
Figure B-43.
Mass spectral matching for peak 20 (Fig. B-23). Library
search yielded good match (probability = 0.85) with
internal standard, chrysene-d^-
105
-------
! Fi 1. -8
4800O-
44000-
.OOOO-
>£ nnri.
52000-
£8000-
£4000-
20000-
16000-
12000^
8000-
4000^
0-
File >6
48000^
40000^
3£OOO;
20000-
16000-
8000-
4000-
0-
3C
'411=. 4'.0-4«iO.O »«'j. UN-?£ 44323-52-29 OIL 1:10 BuBLVST LHk D
flDC TIC
900 1000 1100 1200 1300 1400
z
1
'.0 21. 0 22,0 23.0 24.0 £5.0 26.0 27.0 £8.0 29.0 3o'.0
7416 4". .0-450.0 »BIU. UN-72 44323-52-29 OIL 1:10 ONflLYST LHK D
BOC TIC
1500 1600 1700 1800 1900
,.. i .... i ... .... i .... i .... i .... i .... i .... i .... i .... i ,.
1
4
V f ^ t ^ . --.!^ .1 >.ll.
'.0 n'.O 32.0 33.0 34.0 35.0 36.0 37.0 38.0 39.0 40.0
Figure B-44.
Total ion chromatogram of extract from untreated, 72 h
control sample.
106
-------
APPENDIX C - VOLATILIZATION CALCULATIONS
INTRODUCTION
Estimates of volatile emissions that could accompany quicklime treatment
of PCB-laden soils were made by Louis J. Thibodeaux of Louisiana State
University. The equations were patterned after models he developed for Region
1 of EPA in relation to the New Bedford Harbor (Mass.) Superfund site .
Further calculations based on these equations were made by the authors to
improve the agreement between calculated and observed PCB losses in open-
vessel reactions and to examine the model's sensitivity to various parameters.
MODEL DEVELOPMENT
For the conditions of the open-vessel experiments, it was assumed that
most of the volatile emissions would occur during heat evolution caused by
quicklime slaking. It was further assumed that there was no resistance to PCB
transport (no buildup of PCB concentration) in the vapor phase since strong
ventilation was used in the glove box where the experiment was conducted.
Thus, volatile emissions would depend on diffusive transport of vapor-phase
PCB from a porous medium.
The evaporation rate was calculated as:
W = AK/3 (1)
where W is the evaporative loss rate (g/h), A is the surface area (cm ), K is
the transport coefficient (cm/h), and /o is the PCB congener vapor
concentration (g/cm ) .
Thibodeaux, L. J. Theoretical Models for Evaluation of Volatile
Emissions to Air During Dredged Material Disposal with Applications to New
Bedford Harbor, Massachusetts. U.S. Army Corps of Engineers, Miscellaneous
Paper EL-89-3. Prepared under Contract No. DACW39-87-M-2487, U.S. Army
Engineer Waterways Experiment Station, Vicksburg, MS, 39181.
107
-------
/"> is calculated from the ideal gas law as:
f> = ZJi (2)
RT
*
where P is the temperature-dependent pure component vapor pressure (mm Hg) , M
is the molecular weight (g/mol), R is the gas constant (cm mm Hg kelvin mol
), and T is temperature (kelvin). We assumed that the experiment was
conducted at an atmospheric pressure of 760 mm Hg, although the glove box was
*
actually under slightly reduced pressure. P was calculated as:
P* = exp(A+B/T) (3)
*
using literature values of P at various temperatures to evaluate the
empirical constants A and B. Literature values were not found for the
*
specific congeners used in this study. Consequently, P values for Aroclors
approximating the chlorine content of DCBP, TCBP, and HCBP were used.
The transport coefficient, K, cm/hr, was calculated according tot
K = De1-33(l-e) + S (4)
H
where D is diffusivity (cm /h), e is matrix porosity (cm /cm ) , H is the
height of the solid matrix at 0 porosity (cm), S/t is the mass of solvent lost
over time (g/h) andyOsis the vapor density of the solvent (g/cm ). The term
(1-e) effectively converts the solids height to total height (bed depth) for
any porosity value. The first term of the equation represents transport by
evaporation while the second term represents steam stripping.
The diffusivity of PCB congeners has not been reported in the
literature. However, Thibodeaux used a value of 0.036 cm /s for Aroclor 1242
at 25 C. Knowing this value and the relationship between diffusivity,
temperature and molecular weight allows calculation of estimated diffusivities
for pure PCB congeners. According to the Chapman-Enskog equation:
D
-------
D oCT1'75 x (1/Mp^ + l/Ma1r)'5 (6)
Given known molecular weights and a value of D for Aroclor 1242, D can be
estimated for any congener by ratio of Daroclor1242 to D ., rearranging:
Dpcb = 129.6/jrV-5 (5.114)/_L_ + _1_V5 (7)
\ 298/ V Ma1r M^J
where the value 129.6 is Darociori242 in cm2/h, and 5.114 is the value for the
molecular weight term for Aroclor 1242. Evaluating D by the Fuller-Schettler-
Giddings equation can be performed in the same manner.
CALCULATIONS
Constant and Variable Values
The molecular weights of DCBP, TCBP, and HCBP are 223, 292, and 361
g/mol, respectively. The molecular weight of air was approximated at 29
g/mol. Open vessel experiments were conducted in 9.93-cm diameter beakers,
yielding a surface area of 77.4 cm . The height of the spiked solid phase (50
g silica matrix plus 120 g CaO) prior to water addition was measured at 2.9
cm. Assuming a density of 2.65 g/cm for the matrix and 2.2 g/cm for
hydrated lime (120 g CaO = 159 g Ca(OH)2), the solids volume would be 91.1
cm , with a solids height of 1.18 cm at 0 porosity. Comparison of the solids
volume and measured volume yields a porosity of 0.59.
We did not measure the weight loss of either the spiking solvent or
excess water following the slaking step. Consequently, it is difficult to
estimate the contribution of steam stripping to the transport coefficient.
Assuming that steam stripping only occurred during slaking, when steam
evolution was observed, the following estimate can be made. Addition of 50 mL
water to 120 g quicklime yields a 2.8:2.1 mole ratio. Thus, after complete
slaking, 0.7 mol or 13 g water available for evaporation. Since the sample
appeared nearly dry after the peak temperature was observed, most of this
water must have evaporated in the short time the sample was heated above
100°C.
P values published for Aroclors 1232, 1248, and 1260 were used to
*
calculate the coefficients for evaluating P at experimental temperatures for
DCBP, TCBP, and HCBP, respectively. The values used and resultant calculated
coefficients are presented in Table C-l.
109
-------
Table C-l
* *
P and Calculated Coefficients for P = exp(A+B/T)
Aroclor
ID
P
mm Hg
T
kelvin
1232
1248
1260
4.06x10
2.2 xlO °
-4
4.94x10
5.3 xlO"1
->;
4.05x10
7.6 xlO 2
7.6 xlO 2
.5
9. xlO
1.5 xlO"1
298
373
298
373
298
693
6.73
293
373
2.58x10
2.71x10
1.93x10
1.99x10
1.89x10
2.53x10
3.08x10
-9.33x10
-1.03x10
-8.76x10
-8.96x10
-8.28x10
-1.01x10
-1.22x10
Values of A and B for Aroclor 1260 are calculated,
in the order shown in the Table, from the following
data pairs: 298 and 693 kelvin, 298 and 673 kelvin,
293 and 673 kelvin, 293 and 373 kelvin, and 298 and
373 kelvin.
Model Sensitivity
Calculated evaporation rates are quite sensitive to several variables in
*
the equation. P , which is linearly related to/>, varies widely with
* '
selection of the literature values of P used to calculate the temperature-
dependence coefficients. Figure C-l shows vapor pressure-temperature curves
for Aroclor 1260 used in this exercise to approximate HCBP. The differences
increase greatly in the temperature range of interest in this study (about 185
*C or 458 kelvin) .
110
-------
500
400
300
200
100
VAPOR PRESSURE, mm Hg
-83EBSi-
100
200 300
TEMPERATURE, kelvin
400
500
293,673
293,373
298,373
Figure C-l. Variation of vapor pressure with temperature depends strongly on
*
which literature values of P are used to calculate constants
for equation (3).
The transport coefficient, K, is composed of an evaporation term and a
steam stripping term (eq. 4). The former depends strongly on diffusivity, D,
and less strongly on porosity. Diffusivity was estimated by two formulas, the
Chapman-Enskog and Fuller-Schettler-Giddings equations. Figure C-2 shows the
variation of transport coefficient with porosity for both equations. In both
cases, the transport coefficient reaches a maximum at approximately 60 percent
porosity (balance of increasing pore space with increasing diffusion path
length); values in the range of 40 to 80 percent porosity are within about 25
percent of the maximum value. The assumed temperature dependence of
diffusivity has a much greater effect, with the transport coefficient
increasing more than 400 percent at a porosity of 0.6 as the exponent on the
temperature term is increased from 1.5 (Chapman-Enskog) to 1.75 (Fuller-
Schettler-Giddings) .
Ill
-------
120
100
80
60
40
20 h
0
TRANSPORT COEFFICIENT, cm/h
0
-a B-
0.2
0.4 0.6
POROSITY
0.8
Chapman-Enskog
Fuller et al.
Figure C-2.
Effect of diffusivity estimation method on evaporation term
of transport coefficient, as a function of porosity.Figure
C-2.
The steam stripping term is equally sensitive to solvent evaporation
rate, surface area, arid solvent vapor density (eq. 4). When the water loss by
steam stripping was estimated to be 7-11 g over 8 min, the steam stripping
term varied from 1373 - 2158 cm/h, and the associated loss of PCBs was in the
range of grams per hour. Clearly, much less steam stripping occurred in the
open vessel experiments conducted in this study, because substantial fractions
of the PCB congener masses were still present at the end of the experiment.
Further modeling of steam stripping would require actual measurement of vapor
losses under the various temperature conditions used in the experiment. It
might also require consideration of solvent and PCB interactions with the
matrix that could hinder PCB transport. For these reasons, the steam
stripping term was generally set equal to zero for calculations of PCB loss by
volatilization. The interesting point is that these limited calculations show
that steam stripping is capable of removing large quantities of PCBs from non-
interacting matrices.
Comparison of Calculated Evaporation Rates with Experimental Data
Calculated versus observed PCB losses are shown in Figure 5 of the main
body of this report. The calculated loss curves were constructed by
112
-------
calculating evaporation rates at several temperatures and multiplying the
rate by the estimated time a sample was held within a temperature range
(Table C-2). To obtain calculated values that bracketed observations,
measured values of solids height and porosity were used with the Chapman-
Enskog equation for diffusivity and steam stripping was omitted.
Table C-2
Calculated PCB Losses for Open-Vessel Reactions
STEPWISE PCB LOSS CUMULATIVE
DCBP TCBP HCBP REACTION TIME, h
ACTION*
Spike
Slake
Cool
Slurry
Cool
Ambient
TIME
min
2
60
180
50
420
720
1440
1440
TEMP
kelvin
453
383
358
323
298
298
298
298
0.
150.
96.
51.
0.96
0.57
0.98
1.96
1.96
0.
84.
36.
16.
0.23
0.10
0.18
0.36
0.36
0.
26.
12.
5.7
0.08
0.04
0.07
0.14
0.14
5
12
24
48
72
The time and temperature for each reaction phase are estimated
from the range of measured values.
Figure C-3 shows a set of calculated loss versus reaction time curves
for HCBP using other assumptions. In all cases, varying the solids height
and porosity by ±10 percent had little effect on the loss curve. Changing
*
the temperture-dependence coefficients for P had a greater effect, in one
case yielding a prediction of essentially no HCBP loss. The greatest
influence was exerted by the method of estimating diffusivity. When the
Chapman-Enskog equation was replaced by the Fuller-Schettler-Giddings
equation, predicted HCBP losses for the 72-h open vessel reaction increased
from about 20 percent to 100 percent.
The model used in this comparison, by omitting the steam stripping
term, is not completely appropriate since it is intended for estimating
evaporation from dry materials. In our experiments, the materials were
briefly wet during quicklime slaking and were slurried with water for a 3-h
heating period. The model can be extended to wet matrices, as shown in
equation (4) and as Thibodeaux did for the New Bedford Harbor sediments.
However, the uncertainty due to unknown diffusivities and vapor pressures
of pure PCB congeners, described above, would severely limit confidence in
113
-------
predicted losses. Further work in this area must be preceded by
experiments to generate vapor pressure and diffusivity data.
120
HCBP REMAINING, percent
80
60
40
20
0
Figure C-3.
20
40
TIME, h
60
80
baseline
helght-10%
poroslty+10%
293,673
poroslty-10%
293,393
helght+10%
Fuller et al.
Variation of calculated evaporative losses as equation
parameters are altered.
114
-------
Environmental Protection Information BULK RATE
Agency Cincinnati OH 45268-1 072 POSTAGE & FEES PAID
EPA
No. G-35
Official Business
Penalty for Private Use, $300
EPA/600/2-91/052
-------