EPA 600/R-07/091 I August 2007 I www.epa.gov/ada
United States
Environmental Protection
Agency
Effects of Thermal Treatments
on the Chemical Reactivity of
Trichloroethylene
Abiotic Dearadation
c=c
Trichloroethylene
Reduction
Oxidation
DCE, VC, Ethene
CO, CO2, Organic Acids
Office of Research and Development
National Risk Management Research Laboratory, Ada, Oklahoma 74820
-------
Effects of Thermal Treatments
on the Chemical Reactivity of
Trichloroethylene
Jed Costanza, James Mulholland, and
Kurt Pennell
Georgia Tech University
Eva Davis
Project Officer, Robert S. Kerr Environmental
Research Center
Office of Research and Development
National Risk Management Research Laboratory, Ada, Oklahoma 74820
-------
Notice
The U.S. Environmental Protection Agency through its Office of Research
and Development managed the research described here under EPA Coopera-
tive agreement Contract No. R-82947401 to Georgia Institue of Technology,
Atlanta, Georgia, through funds provided by the U.S. Environmental Protection
Agency's Office of Research and Development, National Risk Management
Research Laboratory, Ada, Oklahoma. It has been subjected to the Agency's
peer and administrative review and has been approved for publication as an
EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
All research projects making conclusions or recommendations based on
environmental data and funded by the U.S. Environmental Protection Agency
are required to participate in the Agency Quality Assurance Program. This
project was conducted under an approved Quality Assurance Plan. Information
on the plan and documentation of the quality assurance activities and results
are available from the lead author.
-------
Foreword
The U.S. Environmental Protection Agency is charged by Congress with protecting the Nation's land, air, and water
resources. Under a mandate of national environmental laws, the Agency strives to formulate and implement actions
leading to a compatible balance between human activities and the ability of natural systems to support and nurture
life. To meet this mandate, EPA's research program is providing data and technical support for solving environmental
problems today and building a science knowledge base necessary to manage our ecological resources wisely,
understand how pollutants affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for investigation of technological
and management approaches for preventing and reducing risks from pollution that threatens human health and the
environment. The focus of the Laboratory's research program is on methods and their cost-effectiveness for prevention
and control of pollution to air, land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites, sediments and ground water; prevention and control of indoor air pollution;
and restoration of ecosystems. NRMRL collaborates with both public and private sector partners to foster technologies
that reduce the cost of compliance and to anticipate emerging problems. NRMRL's research provides solutions
to environmental problems by: developing and promoting technologies that protect and improve the environment;
advancing scientific and engineering information to support regulatory and policy decisions; and providing the technical
support and information transfer to ensure implementation of environmental regulations and strategies at the national,
state, and community levels.
This report describes laboratory experiments conducted to determine the reactivity of trichloroethylene (TCE), a
commonly-used industrial solvent and a groundwater contaminant at many Superfund sites, under the conditions used
for in situ thermal remediation. It was found that at temperatures below 420°C, TCE is essentially unreactive without
the presence of some type of catalyst, such as a base or mineral. Thus, during in situ thermal remediation at these
temperatures, TCE is recovered by volatilization and vapor extraction. At higher temperatures, significant reaction of
TCE may occur; however, the products of these reactions may include larger molecular weight chlorinated compounds
as well as carbon dioxide and hydrochloric acid, which would be the expected products when TCE is completely
mineralized.
Director
Water and Ecosystems Restoration Division
National Risk Management Researori Laboratory
-------
Acknowledgment
EPA would like to thank the peer reviewers, Dr. Bill Mabey, Dr. Gorm Heron, and Dr. Rick Wilken, for their insightful
and useful comments.
-------
Contents
Project Summary 1
1.1 Introduction 1
1.2 Research Objectives 1
1.3 Experimental Systems 2
1.4 Conclusions and Recommendations 2
1.4.1 Quartz Tube Reactor Experiments 2
1.4.2 Sealed Ampule Experiments 3
1.4.3 Implications to Field Applications 3
1.5 Report Organization 6
Background Information 7
2.1 Trichloroethylene Properties 7
2.2 TCE-Water Phase Behavior 8
2.3 Selected Experiments on the Stability of TCE 8
2.3.1 TCE-NAPL Degradation by Oxygen 8
2.3.2 Hydrolysis of TCE-NAPL Degradation Products 10
2.3.3 Degradation of TCE Dissolved in Water at Elevated Temperatures 11
2.3.4 Thermal Degradation of TCE in a Water-Filled Reactor 12
2.3.5 Degradation of Gas-Phase TCE within Heated Quartz Tubes 15
2.3.6 TCE Degradation Products as a Function of the C1:H Ratio 19
2.4 Operational Conditions of In Situ Thermal Treatment Technologies 21
2.4.1 Steam Flushing 21
2.4.2 Thermal Conductive Heating 22
2.4.3 Electrical Resistive Heating 23
2.4.4 Hybrid Thermal Technologies 24
TCE Degradation in Flow-Through Quartz Tube Reactors 25
3.1 Introduction 25
3.2 Experimental Materials and Methods 26
3.2.1 Materials 26
3.2.2 Quartz Tube Apparatus 26
3.2.3 Experimental Procedures 27
3.2.3.1 Experimental Series 1 27
3.2.3.2 Experimental Series 2 27
3.2.3.3 Experimental Series 3 27
3.2.3.4 Experimental Series 4 28
3.2.3.5 Experimental Series 5 28
3.2.4 Analytical Methods 29
3.3 Experimental Results 29
3.3.1 Results of Experimental Series 1-4 29
3.3.2 Results of Experimental Series 5 31
3.3.2.1 TCE Recovery 32
3.3.2.2 Compounds in the DCM Trap 33
3.3.2.3 Compounds Detected in Tedlar® Bags 35
3.3.2.4 Compounds Detected in the Water Rinse 36
3.3.2.5 Compounds Detected in the Iso-Octane Rinse 37
3.3.2.6 Mass Balance 39
3.4 Discussion 40
3.4.1 Nitrogen as the Carrier Gas at 420°C 40
3.4.2 Air as the Carrier Gas at 420°C 42
3.4.3 Experiments Conducted at 120 and 240°C 43
3.5 Summary 45
-------
3.6 Quality Assurance Summary for the Flow-Through Experiments 45
TCE Degradation in Heated Ampules 47
4.1 Introduction 47
4.2 Experimental Materials and Methods 47
4.2.1 Preparation of Solids 47
4.2.2 Preparation of Aqueous Solutions 48
4.2.3 Preparation of Ampules 48
4.2.4 Description of Ampule Experiments 49
4.2.4.1 Ampule Experiment 1 49
4.2.4.2 Ampule Experiment 2 49
4.2.4.3 Ampule Experiment 3 50
4.2.3.4 Ampule Experiment 4 50
4.2.4 Ampule Sampling Methods 54
4.2.5 Analytical Methods 55
4.3 Experimental Results 56
4.3.1 Results of Ampule Experiment 1 56
4.3.2 Results of Ampule Experiment 2 56
4.3.3 Results of Ampule Experiment 3 58
4.3.2 Results of Ampule Experiment 4 61
4.3.2.1 Change in TCE Content 61
4.3.2.2 Change inpH 64
4.3.2.3 CO and CO2 in the Gas Phase 67
4.3.2.4 Other Gas Phase Compounds 69
4.3.2.5 Aqueous Phase Compounds 69
4.3.2.6 Mass Balance 74
4.4 Discussion 76
4.4.1 Oxygen Initiated TCE Degradation 76
4.4.2 Hydrogen Elimination Initiated TCE Degradation 77
4.4.3 Oven Explosion 78
4.4.4 Comparison to Knauss et al. (1999) Results 78
4.5 Summary 79
4.6 Quality Assurance Summary for the Ampule Experiments 80
References 81
APPENDIXA 87
Detailed Experimental Methods for Flow-Through Quartz Tube Reactors 87
A.I Quartz Tube Preparation 87
A.2 Quartz Tube Temperature Profile 87
A.3 Modified TCE Introduction Method: Experimental Series 5 87
A.4 Effluent Trapping Procedures and Analytical Methods 88
A.5 Quartz Tube Rinse Procedure 89
APPENDIX B 91
Change in Dissolved Oxygen 91
APPENDIX C 93
C.I Rates of TCE Degradation 93
C.2 Methods used to Determine TCE Dedradation Rate Parameters in Ampule Experiments 94
-------
Figures
Figure 2.1 Water and TCE-NAPL vapor pressure as a function of solution temperature at 1 bar or
1.02 atm of total gas phase pressure 8
Figure 2.2 Measured concentration of TCE vs. time along with the zero-order reaction model fit 14
Figure 2.3 Natural log of the measured TCE concentration normalized by the initial TCE
concentration vs. time along with the first-order reaction model fit 14
Figure 2.4 Reciprocal of the measured concentration of TCE vs. time along with the second-order
reaction model fit 15
Figure 2.5 Measured concentration of TCE vs. time for Experiment TCE-40 and radical chain
reaction model fit 15
Figure 3.1 Quartz tube experimental apparatus 26
Figure 3.2 Amounts of TCE, PCE and carbon tetrachloride (CC14) recovered during the first
experimental series 30
Figure 3.3 Amounts of TCE, tetrachloroethylene (PCE) and carbon tetrachloride (CC14) recovered
during the second experimental series 30
Figure 3.4 Recovery of TCE with nitrogen or air as the carrier gas, averaged over three relative
humidities (inlet temperatures) 32
Figure 3.5 Amount of chloride detected in the post experiment water rinse 37
Figure 3.6 Amount of dichloroacetate detected in the post experiment water rinse 37
Figure 4.1 Picture of the 50 mL funnel-top ampule before and after sealing 49
Figure 4.2 Explosion-resistant ampule incubation apparatus 53
Figure 4.3 Illustration of the ampule gas sample collection method 55
Figure 4.4 Amount of TCE in the 100 mg/L anoxic (Batch 4) and oxic (Batch 3) ampules stored
at 22°C and incubated at 120°C 62
Figure 4.5 Amount of TCE in anoxic 100 mg/L ampules with Ottawa sand (Batch 6) and Ottawa
sand+1% goethite (Batch 7) stored at 22°C and incubated at 120°C 63
Figure 4.6 Amount of TCE in 100 mg/L anoxic ampules amended with NaOH (0.26 mM) to
pH 10 (Batch 5) stored at 22°C and incubated at 120°C 64
Figure 4.7 The pH of anoxic (Batch 1) and oxic (Batch 2) ampules with 1,000 mg/L of TCE and
TCE-free controls stored at 22°C and incubated at 120°C 64
Figure 4.8 The pH of anoxic (Batch 4) and oxic (Batch 3) ampules with 100 mg/L of TCE and
TCE-free controls stored at 22°C and incubated at 120°C 65
Figure 4.9 The pH of anoxic ampules that contained acid-washed Ottawa sand with 100 mg/L
of TCE (Batch 6) and TCE-free controls stored at 22°C and incubated at 120°C 65
Figure 4.10 The pH of anoxic ampules that contained acid-washed Ottawa sand+l%goethite with
100 mg/L of TCE (Batch 7) and TCE-free controls stored at 22°C and incubated at
120°C 66
Figure 4.11 The pH of anoxic ampules amended with NaOH (0.26 mM) to pH 10 and 100 mg/L
of TCE (Batch 5) along with TCE-free controls stored at 22°C and incubated at 120°C 67
Figure 4.12 Amounts of CO and CO2 in anoxic (Batch 1) and oxic (Batch 2) ampules with
1,000 mg/L of TCE and incubated at 120°C 67
Figure 4.13 Amounts of CO and CO2 in anoxic (Batch 4) and oxic (Batch 3) ampules with
100 mg/L of TCE and incubated at 120°C. No CO or CO2 was detected in ampules
stored at 22°C 68
Figure 4.14 Amounts of CO and CO2 in anoxic ampules with Ottawa sand and 100 mg/L of
TCE (Batch 6) stored at 22°C and incubated at 120°C 68
Figure 4.15 Amounts of CO and CO2 in anoxic ampules with Ottawa sand+l%goethite and
100 mg/L of TCE (Batch 7) stored at 22°C and incubated at 120°C 69
Figure 4.16 Amounts of CO and CO2 in anoxic ampules amended with NaOH (0.26 mM) to pH 10
(Batch 5) and 100 mg/L of TCE incubated at 120°C 69
Figure 4.17 Mass spectrum of the 2.05 min chromatogram peak from the analysis of 1 mL of gas
from Ampule 83 69
-------
Figure 4.18 Amount of chloride in anoxic (Batch 1) and oxic (Batch 2) ampules with 1,000 mg/L
of TCE stored at 22°C and incubated at 120°C 70
Figure 4.19 Amount of chloride in anoxic (Batch 4) and oxic (Batch 3) ampules with 100 mg/L of
TCE stored at 22°C and incubated at 120°C 70
Figure 4.20 Amount of chloride in anoxic ampules with Ottawa sand and 100 mg/L of TCE (Batch 6)
stored at 22°C and incubated at 120°C 70
Figure 4.21 Amount of chloride in anoxic ampules with Ottawa sand+l%goethite and 100 mg/L
TCE (Batch 7) stored at 22°C and incubated at 120°C 71
Figure 4.22 Amount of chloride in anoxic ampules amended with NaOH (0.26 mM) to pH 10 with
100 mg/L of TCE (Batch 5) stored at 22°C and incubated at 120°C 71
Figure 4.23 Concentration of dichloroacetate (DCAA) in the anoxic (Batch 1) and oxic (Batch 2)
ampules with 1,000 mg/L of TCE stored at 22°C and incubated at 120°C 71
Figure 4.24 Concentration of chloroacetate in ampules amended with NaOH (0.26 mM) to pH 10
and 100 mg/L TCE (Batch 5) incubated at 120°C 71
Figure 4.25 Amount of glycolate and formate in anoxic (Batch 1) and oxic (Batch 2) ampules with
1,000 mg/L of TCE incubated at 120°C 72
Figure 4.26 Amount of glycolate and formate in anoxic (Batch 4) and oxic (Batch 3) ampules with
100 mg/L of TCE incubated at 120°C 72
Figure 4.27 Amount of glycolate and formate in anoxic ampules with Ottawa sand and 100 mg/L of
TCE (Batch 6) stored at 22°C and incubated at 120°C 72
Figure 4.28 Amount of formate in anoxic ampules with Ottawa sand+l%goethite and 100 mg/L of
TCE (Batch 7) stored at 22°C and incubated at 120°C 73
Figure 4.29 Amount of sulfate in anoxic ampules with Ottawa sand and 100 mg/L of TCE (Batch 6)
stored at 22°C and incubated at 120°C 73
Figure 4.30 Amount of glycolate in anoxic ampules amended with NaOH (0.26 mM) to pH 10 and
with 100 mg/L of TCE (Batch 5) incubated at 120°C 73
Figure A. 1 Temperature profile within the quartz tube heated to 120°C 87
Figure A.2 Quartz tube apparatus for the fifth quartz tube experimental series 88
Figure C. 1 The change in anoxic ampule TCE content as a function of incubation time 95
Figure C.2 The change in oxic ampule TCE content as a function of incubation time 95
Figure C.3 The change in anoxic ampule TCE content, ampules with Ottawa sand, as a function of
incubation time 95
Figure C.4 The change in anoxic ampule TCE content, ampules with Ottawa sand and 1%
Goethite, as a function of incubation time 95
Figure C.5 The change in anoxic ampule TCE content as a function of incubation time 95
Figure C.6 The change in oxic ampule TCE content as a function of incubation time 95
Figure C.7 The change in anoxic ampule TCE content, ampules with Ottawa sand, as a function of
incubation time 96
Figure C.8 The change in anoxic ampule TCE content, ampules with Ottawa sand and 1%
Goethite, as a function of incubation time 96
Figure C.9 The change in anoxic ampule TCE content based on detected reaction products 96
Figure C. 10 The change in oxic ampule TCE content based on detected reaction products 96
Figure C. 11 The change in anoxic ampule TCE content, ampules with Ottawa sand, based on
detected reaction products 96
Figure C. 12 The change in anoxic ampule TCE content, ampules with Ottawa sand and 1%
Goethite, based on detected reaction products 96
-------
Tables
Table 2.1 Selected Properties of TCE (McNeill, 1978) 7
Table 2.2 Selected Properties of Water (Gebhart et al., 1988) 8
Table 2.3 Selected TCE Stability Test Results 24 hour test (Carlisle and Levine, 1932) 9
Table 2.4 Oxygen and TCE Gas-Phase and NAPL Degradation Products (McKinney et al., 1955) 10
Table 2.5 Hydrolysis of TCE NAPL Degradation Products (McKinney et al., 1955) 10
Table 2.6 Rate of Dichloroacetic Acid (DCAA) Disappearance from Heated Water (Prager et al., 2001)... 11
Table 2.7 Summary of Knauss et al. (1999) Experimental Results 13
Table 2.8 Amount of TCE Degraded after Passing Through a Heated Quartz Tube Residence
Time of 2 Seconds (Graham et al., 1986) 16
Table 2.9 Selected Degradation Products after Passing TCE through a Heated Quartz Tube
(Yasuhara and Morita, 1990) 17
Table 2.10 Selected Degradation Products after Passing TCE through a Heated Quartz Tube
Containing Fly Ash, 0.9 to 1.5 second Residence Time (Froese and Hutzinger, 1994) 17
Table 2.11 Selected Degradation Products at 600°C as a Function of Quartz Tube Solids Content,
0.9 to 1.5 second Residence Time (Froese and Hutzinger, 1994) 18
Table 2.12 Selected Compounds in a TCE Flame with C1:H Ratio of 3 (Chang and Senkan, 1989) 19
Table 2.13 Selected Compounds after Passing TCE through a Flame with C1:H Ratio of 0.09
(Werner and Cool, 2000) 20
Table 3.1 Summary of Flow Through Quartz Tube Experiments 25
Table 3.2 Fifth Experimental Series Matrix 28
Table 3.3 Amounts of TCE, PCE, and CC14 from the Sand Filled Quartz Tube at 420°C 31
Table 3.4 Amount of TCE, PCE, CC14, and CO2 from an Empty Quartz Tube at 420°C 31
Table 3.5 Experiments Completed for the Fifth Experimental Series 31
Table 3.6 TCE Recovery with Nitrogen as the Carrier Gas 32
Table 3.7 TCE Recovery with Air as the Carrier Gas 32
Table 3.8 Concentration (ppmv) of Compounds Detected in the DCM Trap for the 420°C
Experiments with Nitrogen as the Carrier Gas 33
Table 3.9 Concentration (ppmv) of Compounds Detected in the DCM Trap for the 420°C
Experiments with Air as the Carrier Gas 34
Table 3.10 Concentration of Compounds Detected in the Tedlar® Bag and Phosgene Trap for
the 420°C Experiments with Air as the Carrier Gas 35
Table 3.11 Change in the Amount of CO2 and Phosgene Detected with Increase in Water Content
for the 420°C Experiments with Air as the Carrier Gas 36
Table 3.12 Amount of Haloacetic Acids in the Water Rinse from 240°C and 420°C Experiments
with Air as the Carrier Gas 38
Table 3.13 Estimated Amounts (umol) of Compounds in Water Rinse from 420°C Experiment
with Air as the Carrier Gas 38
Table 3.14 Amounts (umol) of Compounds in the Iso-Octane Rinse from the 420°C Experiments
with Nitrogen as the Carrier Gas 38
Table 3.15 Amounts (umol) of Compounds in the Iso-Octane Rinse from the 420°C Experiments
with Air as the Carrier Gas 39
Table 3.16 Distribution of Carbon and Chlorine for the 420°C Experiments with Nitrogen as the
Carrier Gas 40
Table 3.17 Distribution of Carbon and Chlorine for the 420°C Experiments with Air as the Carrier
Gas 40
Table 3.18 Carbon in the DCM Trap and Iso-Octane Rinse, and Chloride in the Water Rinse for
the 420°C Experiments with Nitrogen as the Carrier Gas 41
Table 3.19 Carbon in the DCM Trap and Iso-Octane Rinse, and Chloride in the Water Rinse for
the 420°C Experiments with Air as the Carrier Gas 43
Table 3.20 Amount of Chloride and DCAA in the Water Rinse from the 120°C Experiments 44
-------
Table 3.21 Amount of Chloride, DCAA, and TCAA in the Water Rinse from the 240°C
Experiments 44
Table 4.1 Ampule Experimental Matrix 47
Table 4.2 Experimental Matrix Used for the Fourth Ampule Experiment 50
Table 4.3 Summary of Ampules Prepared for the Fourth Ampule Experiment 51
Table 4.4 Schedule for Convection Oven Ampule Incubation 52
Table 4.5 Schedule for Room Temperature Ampule Storage 53
Table 4.6 Schedule for Explosion-Resistant Apparatus Ampule Incubation 54
Table 4.7 Results of the First Ampule Experiment After 6 Days at 120°C 56
Table 4.8 Results of Second Ampule Experiment After 10 Days at 120°C 57
Table 4.9 Initial and Final TCE Concentrations in the Second Ampule Experiment 58
Table 4.10 Initial and Final TCE Concentration in Anoxic Ampules from the Third Ampule
Experiment (1.4 uL of TCE added to each ampule, Initial DO = 0.79 mg/L) 58
Table 4.11 Initial and Final TCE Concentration in Oxic Ampules from the Third Ampule
Experiment (TCE mixed into 250 mL of water prior to addition to each ampule,
Initial DO = 8.22 mg/L) 59
Table 4.12 Third Ampule Experiment Results for Anoxic Ampules (1.4 uL of TCE added to
each ampule, Initial DO = 0.79 mg/L) 60
Table 4.13 Third Ampule Experiment Results for Oxic Ampules (TCE mixed into 250 mL of
water prior to addition to each ampule, Initial DO = 8.22 mg/L) 60
Table 4.14 Amount (Aqueous and Gas Phases) of TCE in 1,000 mg/L (nominal) Anoxic (Batch 1)
and Oxic (Batch 2) Ampules 62
Table 4.15 Amount (Aqueous and Gas Phases) of TCE in 100 mg/L (nominal) Anoxic (Batch 4)
and Oxic (Batch 3) Ampules 63
Table 4.16 Amount (Aqueous and Gas Phases) of TCE in 100 mg/L (nominal) Anoxic Ampules
with Sand (Batch 6) and Sand+1% Goethite (Batch 7) 63
Table 4.17 Amount (Aqueous and Gas Phases) of TCE in 100 mg/L (nominal) Anoxic Ampules
Amended with NaOH (0.26 mM) to pH 10 (Batch 5) 64
Table 4.18 Elements Present in Ottawa Sand and Ottawa sand+1% Goethite 66
Table 4.19 Mass Balance Between Carbon and Chloride Lost as TCE and Detected as Degradation
Products in Ampules Incubated at 120°C 74
Table 4.20 Amount of Carbon and Chloride Detected as Degradation Products and the C1:C Ratio
for Ampules Incubated at 120°C 75
Table B. 1 Dissolved Oxygen (DO) Concentration Range for Anoxic and Oxic Ampules with
1,000 and 100 mg/L of TCE (Batches 1-4) 91
Table B.2 Dissolved Oxygen (DO) Concentration Range for Anoxic Ampules that Contained
Solids (Ottawa sand and Ottawa Sand+1% Goethite) or NaOH (pH 10) (Batches 5-7) 92
Table C. 1 Rate of TCE Disappearance from the 100 mg/L Ampules at 120°C 93
-------
List of Abbreviations
A, B, C constants in equations hrs
ACS American Chemical Society 1C
AIBN Azo-bis-isobutyronitrile ICP-MS
ASTM American Society of Testing and
Materials ID
ATSDR Agency for Toxic Substances and ISE
Disease Registry IR
atm atmosphere J
C Celsius k or k
C Gas phase concentration of k
g * o
trichloroethylene kl
Co Initial concentration of &2
trichloroethylene k*, k*, k3*
Cw Aqueous phase concentration of
trichloroethylene kg
CTCE Concentration of trichloroethylene kJ/mol
cfm cubic feet per minute L
cis-l,2-DCE c/s-l,2-Dichloroethylene LLNL
cm centimeter
cP centiPoise Ibs/day
DCA Dichloroacetylene M
DCAA Dichloroacetic acid M*
DCM Dichloromethane MCL
DI Deionized MDL
DNAPL Dense, non-aqueous phase liquid MPa
DO Dissolved oxygen MS
d day MSD
Ea Activation Energy MTBE
BCD Electron Capture Detector MQ-cm
EGDY East Gate Disposal Yard m
EPA Environmental Protection Agency mg
ERH Electrical Resistance Heating nig/kg
e" Electron nig/L
FID Flame lonization Detector min
FTIR Fourier-Transform Infrared mol/kg
ft feet mL
ft3 cubic feet niL/g
ft/min feet per minute mL/hr
GC Gas Chromatography mL/min
GC/MS Gas Chromatography/Mass mM
Spectrometry mm
g grams |am
g/mL grams per milliliter mmol
g/mol grams per mole mmolal
H Henry's Law constant N
HCB Hexachlorobenzene NA or na
HCDD Hexachlorodibenzo-p-dioxin NAPL
HP Hewlett-Packard NM
HPO Hydrous pyrolysis/oxidation NPL
HPLC High Performance Liquid No.
Chromatography
hours
Ion Chromatography
Inductively Coupled Plasma - Mass
Spectrometry
Inner diameter
Ion Selective Electrode
Infrared
Joule
rate constant
Zero-order rate constant
First-order rate constant
Second-order rate constant
Rate constants for radical-initiated
reactions
kilograms
kilojoules per mole
liter
Lawrence Livermore National
Laboratory
pounds per day
Molarity
Radical initiator
Maximum Contaminant Level
Method Detection Limit
Million Pascals
Mass Spectrometry
Mass Select Detector
Methyl-tert-butyl ether
Mega ohms per centimeters
meter
milligrams
milligrams per kilogram
milligrams per liter
minute
moles per kilogram
milliliter
milliliters per gram
milliliters per hour
milliliters per minute
millimolar
millimeter
micrometer
millimoles
millimolality
Normality
Not analyzed
Non-aqueous phase liquid
not measured
National Priorities List
Number
-------
nd
nm
nmol
nr
OD
po
po
water
po
TCE
PEEK
PCE
pH
ppmv
psi
R
R2
RH
RSD
SEE
S.D.
s
T
TCAA
TCE
number of ampule results used to
calculate statistics
amount not evident in graph or below
analysis detection limit
nanometers
nanomoles
not reported
Outer diameter
Atmospheric pressure
Total pressure
Pure phase vapor pressure
Water vapor pressure
Trichloroethylene vapor pressure
Poly ether-ether ketone
Tetrachloroethylene
Hydroxide ion content
parts per million volume
pounds per square inch
Universal Gas Constant
correlation coefficient
Relative Humidity
Relative standard deviation
Steam Enhanced Extraction
standard deviation
second
Temperature, Kelvin
Trichloroacetate
Trichloroethylene
TCE0
TCE*
TCE-O2*
TCD
TCH
/
trans-l,2-DCE
U.S.
UV
UZA
ug/L
uL
uL/L
uM
umol
VOCs
Vol
vs.
wt
XRD
(g)
(aq)
t1/2
<
<
%
~
Trichloroethylene, initial moles
Trichloroethylene radical
Trichloroethylene-oxygen radical
Thermal Conductivity Detector
Thermal Conduction Heating
Time
trans- 1 ,2-Dichloroethylene
United States
ultraviolet
Ultra zero grade air
micrograms per liter
micro liter
micro liter per liter
micromolarity
micromoles
Volatile Organic Compounds
volume
versus
weight
X-ray Diffraction
gas phase
aqueous phase
half life
less than
greater than
percent
approximately
-------
Chemical Formulas
Al
A1203
Ar
C
CHOO"
CH2O
CH3'
CH4
CH0
CHC13
CH2C12
CHC1
C2HC13
(also C12C=CHC1)
C2HC1O
C2H2C12
C2H3C1
C2H4C12
C3HC15
C3H2C14
C4HC15
C4H2C14
C6HC15
C6H5C17
:CC12
CC14
•C Cl
.^2^12
C2C12
(also C1C= CC1)
C2C13
C2C13-
C2C14
C2C14H2
C,CL
Aluminum
Aluminum oxide
Argon
Carbon atom
Formate
Formaldehyde
Methyl radical
Methane
Ketene
Ethylene
Dimethyl ether
Chloroform
Dichloromethane
Chloroacetylene
Ethene
Ethane
Trichloroethylene
Chloroketene
Dichloroethylene
Vinyl Chloride
Dichloroethane
Propane
Pentachlorocyclopropane
Tetrachloropropene
Pentachlorobutadiene
Tetrachlorobutadiene
Hexachlorobutene
Pentachlorobenzene
Heptachlorocyclohexane
Dichlorocarbene
Carbon Tetrachloride
Dichlorovinylidene
Dichloroacetylene
Trichloroethylene hydrolysis
product
Trichloroethylene radical
Tetrachloroethylene
1,1,2,2-Tetrachloroethane
Hexachloroethane
C
^
C4C12°H4
C.Cl.O
4 4
C6C15OH
C Cl
^^1
C Cl
SlMs
CO
C02
CO/
COC12
CaSO4
Cl
cr
Cl'
OHO''
C1H2C2OC1
C1H2C200'
C12
C12-
C12COCHC1
C12C2HOH
C12C20
C12HC2C12'
C12HC2OQ-
C12HC2C1200'
C1HCOC1
Pentachloro- 1 -propene
Hexachloropropene
3 ,4-Dichloro-3 -butene-2-one
Perchlorocyclobutenone
Hexachlorobutadiene
Pentachlorophenol
Hexachlorobenzene
Tetrachloro- 1 ,3 -cyclopentadiene-5-
dichloromethylene
Hexachlorophenylacetylene
Octachlorostyrene
Octachloronaphthalene
Carbon monoxide
Carbon dioxide
Carbonate ion
Phosgene
Calcium sulfate
Chlorine atom
Chloride ion
Chlorine radical
radical formed by the reaction of a
chlorine radical with water
Chloroacetyl chloride
Chloroacetate
Chlorine gas
Dichlorine radical anions
Trichloroethylene epoxide
Dichloroethenol
Dichloroketene
Tetrachloroethyl radical
Dichloroacetate
Peroxy radical
Dichloroacetyl chloride
(also Cl,C=CCr) Trichlorovinyl anion
ci3c2Hci'
C13C2OC1
C13C2OO'
C13C202CH3
CIO'
CoCl2
Cu
1,1,1,2-Tetrachloroethyl radical
Trichloroacetyl chloride
Trichloroacetate
Trichloroacetic acid methyl ester
Chlorine-oxygen radical
Cobalt dichloride
Copper
-------
CuO
Fe (also Fe°)
Fe2+
Fe3+ (also Fe(III))
FeOOH
FeNH4(S04)2
FeS2
FeS2
Fe(SCN)2+
FeA
H
H'
H+
HCO;
HCOOH
HC2C1
HC2C12
HC1C2C1'
HC12C2OOH
HOCH2COOH
HOC2OOH
HOC2OCT
HOH2C2OO"
HOOC2OQ-
Copper oxide
Iron, zero valent
Ferous iron ion
Ferric iron ion
Goethite
Ferric ammonium disulfate
Pyrite
Marcasite (polymorph of pyrite)
Iron-thiocyanate complex
Hematite
Hydrogen atom
Hydrogen radical
Hydronum ion
Bicarbonate ion
Formic acid
Chloroacetylene
TCE hydrolysis product
Vinyl radicals
Dichloroacetic acid
Glycolic acid
Oxoacetic acid
Oxoacetate
Glycolate
Oxalate
HC1
HNO3
HS"
H2C1C2OOH
H0
H2S04
H4Si04
H3O+
H3SiO;
HgCl2
Hg(SCN)2
NO3'
NaHCO3
NaOH
Na2C03
02
OH-
Si02
so/-
TiCl,
Hydrochloric acid
Nitric acid
Hydrogen bisulfide ion
Hydrogen gas
Carbonic acid
Monochloroacetic acid
Water
Phosphate ion
Sulfuric acid
Silicic acid
Water, protonated
Silicic acid, dissociated
Mercuric chloride
Mercuric thiochanate
Nitrogen
Nitrate ion
Sodium bicarbonate
Sodium hydroxide
Sodium carbonate
Oxygen
Hydroxide ion
Silica
Sulfate ion
Titanium tetrachloride
-------
Abstract
A series of experiments were completed to investigate abiotic degradation and reaction product formation of
trichloroethylene (TCE) when heated to temperatures ranging from 60 to 480°C. The experimental systems were
designed to simulate conditions anticipated during the thermal treatment of subsurface environments, most notably
the inclusion of a gas phase which is essential because of the strong dependence of TCE vapor pressure and Henry's
Law constant on temperature. The two experimental systems, a 0.5 L quartz tube flow-though reactor and 50 mL
borosilicate glass ampules, provided for the quantification of TCE degradation in the presence of three phases (solid,
liquid, gas). The quartz-tube apparatus was used to study short residence time (<10 minutes) conditions that are
thought to occur during thermal conductive heating and during the recovery of contaminants by vapor phase extraction.
The glass ampules were used to study longer residence time conditions (>1 day) that are thought to occur during steam
flushing and electrical resistive heating. No electrical potential was applied during the experiments, and hence, these
experiments do not directly simulate electrical resistive heating.
The quartz tube experiments were conducted at the temperatures of 120, 240 and 420°C, in the presence of water
vapor, and with either nitrogen or air as the carrier gas. Free chloride ions were detected at all three temperatures
considered, which was interpreted as evidence of gas-phase TCE degradation. The amount of chloride formed in the
120°C experiments was small, representing less than 0.01% of the TCE that passed through the quartz tube. Passing
TCE through the quartz tube heated to 420°C with nitrogen as the carrier gas resulted in substantially greater amounts
of chloride (up to 6.5% of TCE). Chlorinated compounds (up to 7% as TCE) with 4 and 6 carbon atoms and at least
5 chlorine atoms were also detected at 420°C. Introducing air containing 21% oxygen into the quartz tube heated
to 240°C resulted in the detection of chloride representing up to 0.4% of TCE introduced, as well as the detection
of dichloroacetate and trichloroacetate. At 420°C, the presence of oxygen in the carrier gas resulted in significant
increases in the number and amount of reaction products detected. Under these conditions, more than 20% of the
carbon introduced as TCE was transformed into carbon monoxide and carbon dioxide, while up to 22% of the chlorine
introduced as TCE was detected in the form of chlorinated carbon compounds. Increasing the quartz tube water content
resulted in an increase in TCE recovery concurrent with a decrease in TCE degradation products with nitrogen as the
carrier gas. With air as the carrier gas, increasing the quartz tube water content in the 420°C experiments may have
served to hydrolyze phosgene and remove reactive chlorine from the gas phase while not impacting the amount of TCE
degraded.
The ampule experiments were conducted in borosilicate glass ampules that were filled to approximately three-quarters
capacity with aqueous solutions containing TCE at initial concentrations of 100 and 1,000 mg/L. The rate of TCE
degradation and products formed was determined as function of dissolved oxygen concentration, hydroxide ion
concentration, and solids content. There was no significant reduction (>10%) in TCE content of the ampules with initial
concentration of 1,000 mg/L of TCE that were incubated over a 20-day period at 120°C. However, significant changes
in solution pH were observed along with the detection of chloride ions and organic compounds other than TCE. The
concentration of TCE decreased in ampules that initially contained 100 mg/L of TCE and were incubated at 120°C. The
decrease in TCE content was matched with a decrease in ampule pH, an increase in the chloride, formate, and glycolate
content of the aqueous phase, and an increase in the carbon monoxide and carbon dioxide content of the gas phase.
Dichloroacetylene (DCA) was detected in ampules and may represent an intermediate formed during TCE degradation.
DCA is a reactive compound that can interact with the variety of compounds present in soil such as organic carbon.
Thus, the degradation products formed during the in-situ thermal treatment TCE may not be limited to those found in
the ampule experiments since the ampules did not contain organic carbon other than TCE. The rates of TCE degradation
in ampules with anoxic water, both with and without sand, and in oxic water were similar at 120°C. The degradation
rate in ampules with anoxic water and sand was increased by adding 1% (wt) goethite.
The experimental results presented herein represent a first step toward understanding TCE chemical reactivity and
reaction product formation during thermal treatment. Additional experimentation, both at the laboratory and field
scale, is recommended to further elucidate TCE reaction pathways and rates, and to more accurately represent the
complexities inherent in natural subsurface materials and field-scale application of thermal treatment technologies.
-------
-------
1.0
Project Summary
1.1 Introduction
Laboratory studies on the hydrolysis of environmentally
significant halogenated compounds have shown that
trichloroethylene (C2HC13) is extremely recalcitrant in
aqueous environments, with a measured half-life as large
as approximately 100,000 years under neutral conditions
at 25°C (Jeffers and Wolfe, 1996). Other researchers
found TCE to resist hydrolysis at 100°C (Dilling et al.,
1975). However, in-situ aqueous phase degradation of
trichloroethylene (TCE) into carbon dioxide (CO2) and
chloride (Cl") is claimed to occur during the thermal
treatment of contaminated subsurface environments
(Knauss et al., 2000). This claim is based on
experimental results obtained from a completely water-
filled, constant pressure, gold-walled reactor operated in
the temperature range from 70 to 100°C (Knauss et al.,
1999). The only degradation products reported in these
experiments were dissolved carbon dioxide and chloride.
However, no quantitative evaluation of the amounts
of carbon dioxide and chloride recovered with respect
to the initial mass of trichloroethylene was performed.
While limited quantitative data are available on chemical
reaction as a means of destroying contaminants in
thermal remediation (Stegemeier and Vinegar, 2001), in-
situ degradation of TCE into carbon dioxide and chloride
has been observed during thermal conductive heating
at temperatures ranging from 500 to 700°C (Baker and
Kuhlman, 2002).
Subsurface environments are extremely complex
systems, comprised of three phases including mixtures
of solids, liquids, and gases. Subsurface solids are
composed of minerals and organic matter, which may
facilitate the abiotic degradation of TCE into products
other than those found in pure water or gas reaction
environments (Lee and Batchelor, 2003; 2004; Haderlein
and Pecher, 1998). A comprehensive review of TCE
degradation and the degradation products formed in
heated environments in the presence of three phases
(solid, liquid-water, and gas) does not currently exist.
The potential TCE degradation products are not limited
to carbon dioxide and chloride alone, but also include
acutely toxic products such as dichloroacetyl chloride
(CLHC2OC1) and phosgene (COC12) that have been
detected during the gas phase photocatalytic treatment of
TCE (Haag et al., 1996; Amama et al., 2001).
As the use of thermal technologies, including steam
flushing, electrical resistive heating and thermal
conductive heating, to remediate chlorinated solvent
source zones becomes more common, there is a need to
not only determine rates of TCE degradation and thus
how much degradatation can be expected to occur in
situ, but also to elucidate thermal reaction pathways and
degradation products. Because of the strong dependence
of TCE vapor pressure and Henry's Law constant
on temperature, it is essential that such experiments
include a gas phase. For this reason, the experimental
systems used in this work, a quartz tube reactor and
sealed ampules, were specifically selected to provide for
quantification of TCE degradation in presence of three
phases (solid, liquid, gas). Although these experimental
systems do not replicate field conditions, they represent
a significant step forward from previous work, which
considered TCE degradation in single-phase water (e.g.,
Knauss et al., 1999) or gas (e.g., Zhang and Kennedy,
2002) systems.
1.2 Research Objectives
The primary objective of this work was to determine if
significant TCE transformation occurs in three-phase
systems (gas, liquid-water, solid) at temperatures and
conditions typically used for thermal remediation.
Transformation of TCE was confirmed through the
identification of reaction products. Identification of
reaction products also allows some understanding
of the likely dominant reaction mechanisms for the
conditions studied. The research involved a series of
laboratory experiments performed in either a flow-
through quartz tube or sealed glass ampules that were
designed to simulate conditions anticipated to occur
during thermal treatment of porous media contaminated
with TCE. Experimental conditions were varied in order
to systematically evaluate the effects of temperature,
oxygen concentration, hydroxide ion concentration,
water content, and solids content on the rate of TCE
degradation and degradation products formed. The
following conditions (i.e., experimental variables) are
anticipated in the subsurface during thermal treatment:
• The temperature of the subsurface can range from
approximately 50 to 600°C or greater depending
upon the thermal treatment technology employed.
• The subsurface can remain at an elevated
temperature for a period of one month to more than
one year.
• The concentration of oxygen in the gas phase can
range from less than 1 up to 21%. The concentration
of oxygen in the liquid phase can range from less
than 0.1 up to 8 mg/L.
-------
• Within soil pore spaces water and TCE will be
converted from the condensed to vapor phase as the
subsurface temperature increases.
• There may be significant changes in the
concentration of dissolved ionic species as the
subsurface temperature increases.
In addition to changes that occur as a result of heating,
the initial subsurface conditions prior to thermal
treatment can vary depending upon:
• The type and amount of mineral and organic matter
present in the solid phase.
• The initial amount of TCE present in the aqueous
phase, solid phase, and gas phase, and existing as a
separate non-aqueous phase liquid (NAPL).
Based on a detailed review of previous experimental
results described in Chapter 2.3 (see Table 2.4), the
anticipated TCE degradation products under oxidative
conditions include carbon monoxide (CO) and
carbon dioxide (CO2) in the gas phase, and chloride,
dichloroacetic acid, oxoacetic acid, and formic acid
in the aqueous phase when the subsurface is heated to
temperatures of less than 120°C (e.g., steam flushing,
electrical resistive heating) and there are no other
reactive species present. At temperatures greater than
300°C (e.g., thermal conductive heating), the expected
gas phase TCE transformation products include CO,
CO2, phosgene (COC12), and chlorinated hydrocarbons
(see Tables 2.8, 2.11, and 2.12) when there are no other
reactive species present. Therefore, the experimental
systems and analytical methods must be carefully
designed and tested in order to collect and detect a wide
range of degradation products that may occur in the gas,
solid and aqueous phases.
1.3 Experimental Systems
While laboratory-scale apparatus are useful for
investigating simulated subsurface conditions, the
experimental materials must be relatively inert. In
general, borosilicate glass provides thermal stability
(softening point of 820°C), and is primarily composed
of silica (SiO2) which is more similar to subsurface
materials than metal based materials such as stainless-
steel. The collection and analysis of samples from each
phase within a dynamic or closed apparatus is required
to quantify the TCE degradation and reaction product
formation. Calculating the difference between the
amount of carbon and chlorine atoms present before and
after each experiment (i.e., mass balance) allows for
an assessment of whether or not all possible reactants
and products have been measured. The selectivity and
sensitivity of each analytical method must be appropriate
for the expected degradation products. Infrared
spectroscopy is applicable for the analysis of dissolved
carbon dioxide, but requires concentrations of greater
than 0.24 mM (Burt and Rau, 1994) for quantification.
However, most of the carbon dioxide is expected to be
present in the gas phase of three-phase systems, meaning
that the collection and analysis of gas samples will be
more sensitive to the amount of carbon dioxide formed.
Sample preparation methods must also be appropriate.
For example, the presence of haloacetic acids (i.e.,
dichloroacetic acid) may not be detected by traditional
analytical techniques if the proper sample preparation
methods are not used. Determining the presence of
haloacetic acids requires addition of a strong acid to
reach at least pH 2 for spectroscopic detection, and an
additional derivitization step for gas chromatographic
separation.
For the experiments conducted in this project, subsurface
conditions were simulated using two experimental
systems: 1) a 0.5 L flow-though quartz tube apparatus
and 2) 25 and 50 mL borosilicate glass ampules. The
quartz tube was used to study the high temperature
(>120°C), short residence time (<10 minutes) reactions
that are thought to occur within approximately 1 to 3
feet of heater wells used in thermal conductive heating.
The gas-phase effluent from the flow-through tubes
was passed through liquid traps and then captured in
a Tedlar® bag. The liquid-trap fluids, Tedlar®-bag
contents, and solid extracts were analyzed using gas
chromatography (GC), ion chromatography (1C), and
ion selective electrodes. The borosilicate glass ampules
were used to study low temperature range (60 to 120°C)
over longer residence times (>1 day), conditions that
are likely to occur during steam flushing and electrical
resistive heating. No electrical potential was applied to
the ampules, and hence, the system does not directly
simulate electrical resistive heating. The ampules were
destructively sampled at specified time intervals and
samples from each phase were collected for analysis
using gas chromatography (GC), ion chromatography
(1C), and ion selective electrodes. In addition, GC
mass spectrometry (GC/MS) was employed to identify
unknown compounds detected in the aqueous and gas
phase effluent sample and solid phase rinses.
1.4 Conclusions and
Recommendations
1.4.1 Quartz Tube Reactor Experiments
A series of five quartz tube reactor experiments were
completed over a temperature range of 22 to 480°C
(Chapter 3). Specific experiments were conducted
to investigate the effects of temperature, water vapor
content, solids (Ottawa sand), and oxygen content on
TCE degradation and reaction product formation. In
all of the quartz tube experiments, some degree of
TCE degradation was observed, however, the greatest
amount of TCE was transformed (up to 48% of the TCE
introduced) at temperature of 420°C in the presence
-------
of oxygen (air as the carrier gas). The amount of TCE
degraded was dependent on the temperature of the
quartz tube, with more being degraded in the 420°C
experiments than in the 120 or 240°C experiments.
The amount of TCE degraded was also dependent on
the amount of oxygen present in the 240 and 420°C
experiments, with more TCE degraded when air (i.e.,
21% O2) was used as the carrier gas. With nitrogen as the
carrier gas, up to four TCE degradation products were
identified in the liquid-trap fluids and quartz tube rinses,
with no CO or CO2 detected for experiments completed
at 420°C. The amount of TCE recovered for the 420°C
experiments with nitrogen as the carrier gas was greater
than 97%, with up to 3.4% detected as chlorinated
hydrocarbon degradation products. The amount of TCE
recovered as chlorine was greater than 100%, with up to
7% as chlorinated degradation products. The degradation
products detected contained 4 and 6 carbon atoms
with greater than 5 chlorine atoms per molecule. TCE
degradation was proposed to be initiated by thermal
induced unimolecular dissociation of TCE but was also
influenced by chlorine induced degradation. Increasing
the quartz tube water content resulted in an increase
in TCE recovery concurrent with a decrease in TCE
degradation products that was suggested to indicate a
decrease in chlorine induced TCE degradation.
With air as the carrier gas, there was an increase in the
amount of TCE degraded and an increase in the number
of degradation products detected as compared with
experiments completed with nitrogen as the carrier gas.
The average recovery of TCE was greater than 94% with
air as the carrier gas for the 120 and 240°C experiments,
but decreased to approximately 53% in the 420°C
experiments. Carbon-based TCE degradation products
were detected in the 240 and 420°C experiments with
air as the carrier gas. Three degradation products were
identified in the quartz tube rinse from the 240°C
experiments, and up to 13 degradation products were
detected in the liquid-trap fluids and quartz tube rinses in
the 420°C experiments. The degradation products ranged
from single carbon compounds with 3 chlorine atoms
(i.e., chloroform) to compounds with up to 6 carbons
and 6 chlorine atoms (i.e., hexachlorobenzene). Carbon
monoxide (CO), CO2, and phosgene were detected in
the gas phase of the 420°C experiments only with air as
the carrier gas. The amount of TCE recovered as carbon
for the 420°C experiments with air as the carrier gas
ranged from 79.1 to 91.5%, and the amount of chlorine
recovered ranged from 74.6 to 88.8%. TCE degradation
was proposed to be initiated by thermal induced
unimolecular dissociation but was also influenced by
the formation of peroxyl radicals due to the presence
of oxygen. Increasing the quartz tube water content in
the 420°C experiments with air as the carrier gas may
have served to hydrolyze phosgene and remove reactive
chlorine from the gas phase, while not impacting the
amount of TCE degraded.
1.4.2 Sealed Ampule Experiments
Four series of ampule experiments were completed to
investigate the effects of oxygen content, hydroxide
ion content (pH), and solids (Ottawa sand and goethite)
on TCE degradation at temperature ranging from 22 to
120°C for periods of up to 41 days (Chapter 4). The
results of the ampule experiments demonstrate that
TCE was degraded within sealed glass ampules that
contained gas, water, and solids. The rates of TCE
degradation in ampules with anoxic water, both with and
without sand, and in oxic water were similar at 120°C.
The degradation rate in ampules with anoxic water and
sand was increased by adding 1% (wt) goethite, with
a first order half-life on the order of 10 days at 120°C.
The primary TCE degradation products included CO
and CO2 in the gas phase and chloride, hydronium ions,
formate, glycolate in the aqueous phase. Minor amounts
(<1 mg/L) of dichloroacetic acid (DCAA) were detected
in select ampules, most consistently in ampules that that
were stored at 22°C and initially contained 1,000 mg/L
TCE along with oxygen. Dichloroacetylene (DCA) was
detected in minor amounts (i.e., DCA < 1% of TCE)
in ampules that contained TCE and were incubated at
120°C.
Dichlororacetylene, in addition to being a TCE
degradation product, was also thought to represent a
key intermediate. The presence of DCA was proposed
to indicate that the lone hydrogen atom in TCE was
being eliminated by nucleophiles, such as sodium
hydroxide (NaOH), which increased the rate of TCE
degradation and amount of DCA when added to the
ampules as NaOH. Dichloroacetylene was proposed to
be hydrolyzed to form chlorinated organic acids, such as
DCAA, which were then hydrolyzed at 120°C to form
the non-chlorinated organic acids, glycolate and formate.
1.4.3 Implications to Field Applications
The results in this report are important for demonstrating
that transformation reactions can occur during thermal
remediaton. However, extending these laboratory results
toward predicting the rate of TCE degradation during
the in-situ thermal treatment of TCE contaminated
subsurface regions involves a significant degree of
speculation. Laboratory experiments are performed
on simplified systems or with materials that have, in
some way, been altered from their natural state in the
subsurface environment. Most thermal remediation
projects for TCE are focused on physical recovery
through vaporization and vacuum extraction. The
transformation reactions demonstrated here will occur
simultaneously with vaporization, and the relative
rates of the two processes will be controlled by site
specific conditions. A remediation system could not be
-------
designed or operated for one of the processes without the
other process also occurring. In most cases, however,
vaporization rates are likely to be much faster than the
transformation reactions demonstrated here, as thermal
treatment of TCE contaminated sites are often completed
in much shorter times (less than one year) than would
be required to reduce TCE concentrations to a similar
degree by the in situ transformation processes found
here. Thus, in situ thermal remediation systems for TCE
require robust extraction and treatment systems that can
recover and treat the transformation products as well as
the unreacted TCE.
Ampules such as those used in the aqueous phase
experiments here have been used successfully by other
researchers to study transformation rates of volatile
organic compounds. The ampule experiments reported
here extend that work by the addition of Ottawa sand or
1% (wt) goethite, materials collected from the subsurface
and commonly found in soil environments. However,
the Ottawa sand was acid-washed, which could remove
surface coatings, while the goethite was ground prior to
use, potentially creating fresh active surfaces (Papirer et
al., 1993).
The water used in all experiments was deionized,
whereas natural groundwater contains ions. For
example, sodium hydroxide (an anion) was used in the
ampule experiments to simulate a strong nucleophile
and was shown to increase the rate of TCE degradation
at 120°C by an order-of-magnitude compared to the
rate determined for deionized water. The primary
nucleophiles expected in the subsurface environment
include hydrogen bisulfide (HS~), hydroxide (OET),
phosphate (H2PO42-), bicarbonate (HCO30, sulfate
(SO42~), and nitrate (NO3~) in order of decreasing
nucleophilicity with hydrogen bisulfide as the strongest
nucleophile and nitrate as the weakest. Thus the anionic
content of natural groundwater, in addition to the iron
containing minerals present, may strongly influence
the rate of TCE degradation. Therefore, predicting the
potential rate of TCE degradation during in-situ thermal
treatment requires specific information regarding the
geochemistry of the site being treated, keeping in
mind that the increase in temperature can affect the
geochemistry. For example, significant levels of sulfate
(> 1 mM) were formed in ampules incubated at 120°C
from the dissolution of pyrite and marcasite found in the
Ottawa sand.
Free radical reactions may contribute to contaminant
transformation during thermal remedation. Under
ambient conditions free radicals may be created in the
subsurface by the reaction of chlorinated compounds
with naturally occurring iron-containing materials
(Kriegman-King and Reinhard, 1992; 1994). However,
a significant variation from natural conditions in these
laboratory experimental systems is the absence of
naturally occurring organic matter. Although naturally
occurring organic matter has not been found to have an
effect on reactions such as hydrolysis (Haag and Mill,
1988), it has been found to have significant effects on
free radical reactions. Haag and Hoigne (1985) found
that fast consumption of hydroxyl radicals by natural
dissolved organic solutes and bicarbonate ions decreased
the amount of organic pollutants oxidized.
With regard to free radical reactions, TCE has been
found to be virtually unreactive in oxidation reactions
without a radical initiator (Kucher et al., 1990).
Radical-mediated reactions are important for TCE
decomposition in both the aqueous and vapor phase, and
TCE is known to react with hydroxyl radicals (Buxton
et al., 1988). As discussed in Section 2.3.4, the kinetic
data reported by Knauss et al. (1999) are consistent
with a radical chain reaction, but this is essentially an
experimental observation, and the details of the reaction
mechanism are unknown. It is not possible to assess
the importance of this reaction in situ during thermal
remediation. According to Buxton et al. (1988), radicals
can be generated in the laboratory in aqueous solution
by radiolysis of water, photolysis, high frequency
electric discharge, sonolysis, and Fenton-type reactions.
However, none of these types of radical-generating
processes may be present in the subsurface during
thermal remediation.
High temperatures such as that found in an incinerator
flame may generate free radicals in the vapor state
(Taylor et al., 1990), and the possibility exists for this to
occur in the subsurface near heaters or electrodes. The
species of radicals formed depends on the temperature
of the system, the compounds present, and the fuel
to oxygen ratio. At temperatures below 750 - 800°C,
diatomic radicals are sufficiently stable to contribute
to organic reactions, which may form larger molecules
rather than break down chemicals, while at higher
temperatures atomic radicals are prevalent and more
likely lead to the breakdown of compounds (Taylor
etal., 1990). Naturally occurring organic matter is
likely to play a significant role in reaction rates and
products formed in free radical reactions occurring in
the subsurface, but natural organic matter (particularly
humic substances) can be both a source (radical initiator)
and sink (inhibitor, radical termination step) for free
radicals.
Steam flushing and electrical resistive heating are
expected to heat TCE contaminated subsurface regions
to temperatures between 70 to 120°C. Based on the
results for the ampule experiments, the rate of thermally
induced TCE degradation in the 70 to 120°C range could
be significant depending on the mineral species present
in the soil (e.g., goethite) and the anion concentration
-------
of the soil water (e.g., HS"). However, in situ changes
in the phase distribution of TCE as the temperature is
increased must be considered. Prior to heating, the
contaminant (i.e., TCE) will be distributed between the
organic liquid (NAPL) (if present), aqueous, solid, and
gas phases. As the temperature increases, the aqueous
solubility of TCE-NAPL has been shown to increase,
which is likely to result in somewhat higher groundwater
concentrations. However, gas-phase concentration of
TCE would be expected to increase substantially due to
the increase in vapor pressure and Henry's law constant
(Heron et al., 1998). In addition, the gas-phase content
of a soil will increase as water evaporates, increasing
the fraction of contaminant mass in the gas-phase. The
application of a vacuum to the thermally-treated zone is
used to enhance the removal of gas-phase constituents
from the subsurface, but requires continuity of gas flow
pathways to be effective.
Although in some steam injection systems air is
injected with steam to enhance the movement of vapor
phase contaminants to recovery wells or in an effort
to increase oxidation reactions, many contaminated
subsurface environments, such as those found at Cape
Canaveral (Interagency DNAPL Consortium, 2002)
or Fort Lewis's East Gate Disposal Yard (Truex et.
al, 2007) appear to have reductive conditions rather
than oxidative conditions. This may be advantageous,
because despite the known recalcitrance of TCE under
oxidative conditions, laboratory experiments on the
abiotic degradation of TCE have shown reduction may
occur under some anaerobic conditions. Bulter and
Hayes (2001) found TCE transformation to occur with
freshly prepared iron sulfide, but not with an 'aged' iron
sulfide with a slightly more crystalline structure. Su and
Puls (1999) detected TCE degradation with several types
of zero valent iron; however, the reaction rate constant
and activation energy varied significantly. Nevertheless,
at least a one order-of-magnitude increase in reaction
rate was found with each of the zero valent irons as
the temperature was increased from 10 to 55°C. This
process could contribute to TCE degradation in some
ERH remediation systems where iron is used as backfill
around electrodes.
During thermal conductive heating three broad
temperature regimes can be envisioned to emanate
radially from the heater wells: a 700 to 900°C region
located in the immediate vicinity of and within the heater
well, a 500 to 700°C region within a 1 to 3 foot radius
of the heater well, and a 100 to 250°C region located 10
to 20 feet from or between heater wells. Temperatures
at the heating wells for thermal conductive heating are
often in the range that can generate radicals, thus it may
be possible to form free radicals in these systems in the
immediate area around the heater wells. Some of the
free radicals formed could react with naturally occurring
organic matter rather than TCE, forming unknown
products, and effectively quenching the free radical
chain reaction. Thus, the amount of TCE transformed
during actual remediation may be less than what was
found in laboratory experiments. Based on the results
of the quartz tube experiments, TCE is expected to be
transformed at temperature above 400°C into other
chlorinated hydrocarbons, and if sufficient oxygen is
present, into CO and CO2. Carbon monoxide (CO) and
CO2 were only detected when oxygen was present, while
no CO or CO2 was detected in experiments completed
without oxygen. Phosgene (COC12), a toxic gas, will
also form with oxygen present. Phosgene is a gas at
ambient temperature; thus, it would be recovered by
the vacuum extraction system. Phosgene is relatively
stable in incineration environments (Taylor et al., 1990),
and thus may pass through the reaction chamber of a
thermal oxidizer and into the scrubber, where it readily
hydrolyzes to nontoxic products with water and thus
would be anticipated to be removed (Haag et al., 1996).
The more volatile chlorinated-hydrocarbon degradation
products (e.g., chloroform, carbon tetrachloride) can
be recovered by a properly designed and operated
vacuum extraction system. Less volatile degradation
products such as hexachlorobenzene are also likely to
be sufficiently vaporized to be recovered in the vapor
extraction system at the temperatures commonly used
in these systems, although some mass may condense or
be sorbed by soil particles. Chlorinated organic acids
such as dichloroacetic and trichloroacetic acid may
also be formed in the subsurface, although their yields
would be expected to be only a very small percentage
of the TCE (less than 0.1% combined). These organic
acids are water soluble (trichloroacetic acid is a solid
at ambient temperatures), with low vapor pressures and
moderate boiling points (194 and 197°C, respectively;
Verschueren, 2001). A small fraction of their mass
may be recovered by a gas-phase vacuum extraction
system, while mass remaining in the water phase could
hydrolyze in the heated water (see Table 2.6).
The complete transformation of TCE to CO2, CO,
hydrocloric acid (HC1) and water, without the formation
of chlorinated degradation products, has been shown
to require temperatures on the order of 900 to 1,000°C
(Chang and Senkan, 1989; Werner and Cool, 2000).
Since the heater wells normally reach temperatures of
up to 800°C, a fraction of the TCE contaminant mass
may undergo complete oxidation to nontoxic products
within the heater wells prior to extraction to the above
ground treatment system. In the lower temperature
zones outside of the heater wells, the formation of
higher molecular weight chlorinated organic compounds
will be favored if high concentrations of chlorinated
chemicals are present, and some of these, such as
tetrachloroethylene and hexachlorobenzene, are more
-------
difficult to degrade than TCE (Dilling et al, 1975;
Taylor et al., 1990). Thus, a robust vapor extraction
system must be employed as well as an off-gas treatment
system to ensure destruction or removal of chlorinated
hydrocarbons that may exist in the effluent gas stream.
Clearly, the experimental results presented herein
represent only a first step toward understanding TCE
chemical reactivity and reaction product formation
during thermal treatment. Additional experimentation,
both at the laboratory and field scale, is recommended to
further elucidate TCE reaction pathways and rates, and
to more accurately represent the complexities inherent in
natural subsurface materials and field-scale application
of thermal treatment technologies.
1.5 Report Organization
Following this project summary (Chapter 1), background
information (Chapter 2) related to the stability of TCE in
heated systems is presented, followed by a description
of operational conditions for steam flushing, thermal
conductive heating, and electrical resistive heating.
Experimental methods and results for the quartz tube
studies are presented in Chapter 3, followed by the
experimental methods and results for the ampule studies
in Chapter 4, with cited references listed in Chapter 5.
Appendix A describes detailed experimental methods
used for the quartz tube reactor studies, Appendix B
contains dissolved oxygen data from some of the ampule
experiments, and Appendix C contains rate constants
calculated for TCE degradation assuming a first order
reaction rate and a description of the methods used
to compute these rates from the ampule experimental
results.
-------
2.0
Background Information
Trichloroethylene (TCE) is a contaminant commonly
found in the subsurface at industrial and military
installations in the United States and abroad. Improper
disposal or release of liquid or "neat" TCE to the
environment frequently results in the presence of
a separate organic phase contaminant, commonly
referred to as a non-aqueous phase liquid (NAPL),
that can become entrapped within soil pore spaces as
individual droplets and ganglia (Hunt et al., 1988).
These entrapped NAPL droplets and ganglia are
immobile under normal groundwater flow regimes.
If sufficient NAPL is released to the subsurface, the
organic liquid is likely to accumulate in "pools" above
layers of lower permeability media. In general, NAPLs
will not enter a lower permeability layer unless the
entry pressure is exceeded, that is, the pressure exerted
by a continuous NAPL pool must be sufficient to
displace water from the pore space. The presence of
TCE-NAPL in the subsurface often represents a long-
term source of contamination as TCE slowly dissolves
into the groundwater flowing through the "source
zone." Pumping of groundwater and soil gas from
the subsurface followed by above ground treatment is
often used to control the migration of dissolved-phase
TCE plumes, and in some limited cases, to restore the
subsurface. Increasing the subsurface temperature has
been shown to increase the transfer of TCE mass from
the NAPL to the water and gas phases, which increases
the rate and amount of TCE that can be removed from
the subsurface by extraction methods. Thus, subsurface
heating can be employed to dramatically enhance TCE
mass recovery (Davis, 1997) and holds the potential to
transform TCE into nontoxic products via thermally-
induced chemical reactions (e.g., Knauss et al., 1999).
The following sections describe relevant physical and
chemical properties of TCE, selected results from
experiments on the thermal stability of TCE reported in
the literature, and the operational conditions associated
with commonly used in-situ thermal treatment
technologies.
2.1 Trichloroethylene Properties
TCE is an important solvent used for cleaning metal
parts and electrical components, and in the manufacture
of hydrofluorocarbon refrigerants (HSIA, 2001). TCE
is a colorless, sweet smelling, volatile liquid that is
acutely toxic to humans when ingested (Mertens, 1999).
Even though TCE is referred to as a non-flammable
liquid, it should be kept away from open flames and
metal surfaces with temperatures greater than 176°C
due to the flammability of its vapors (Mertens, 1999).
If TCE is exposed to a temperature greater than 420°C
when oxygen is present, it will spontaneously ignite
(Mallinckrodt and Baker, 2003a).
Even though TCE has low solubility in water, TCE
is one of the most commonly found groundwater
contaminants in the United States, and is present at
305 of the 1,236 National Priority List (NPL) sites
(U.S. EPA, 2003). While the long-term health effects of
drinking water contaminated with small amounts of TCE
are not yet known, the U.S. EPA has set the maximum
contaminant level (MCL) for drinking water at 5 ug/L
(ATSDR, 1997). If TCE is found in groundwater
at concentrations greater than 5 ug/L, treatment or
control of the groundwater is usually required. Selected
properties of TCE and water as a function of temperature
are given in Tables 2.1 and 2.2.
Table 2.1 Selected Properties of TCE (McNeill,
1978)
Molecular Weight
(g/mol)
Melting Point (°C)
Boiling Point (°C)
Critical Temperature
Critical Pressure
(MPa)
Koct (mL/g)
Log Kow1
Properties at
Temperature (°C)
TCE Explosive Limit
inAir(Vol%)
TCE Viscosity
(cP= lOOxg/cms)
TCE Liquid Density
(g/mL)
Solubility in Water
(mg/L)
Henry's Law
Constant*
(dimensionless)
Vapor Pressure of
TCE (MPa)
m39
. ~j y
-87.1
86.7
271.0
5.02
160
2.38
20
8 to 10.5
0.58
1.465
1,068
0.3
0.008
60
no
data
0.42
no
data
1,219
1.5
0.042
100
8 to 52
no data
1.325
no data
5.3
(95°C)
0.148
MPa= 1.00 bar = 1.02atm
* Heron et al., 1998; t LaGrega et al., 2001
-------
Table 2.2 Selected Properties of Water (Gebhart
etal., 1988)
Properties at
Temperature (°C)
Water Viscosity
(cP= lOOxg/cms)
Water Density (g/mL)
Water Solubility in TCE
(mg/L)
Vapor Pressure of Water
(MPa)
20
0.99
0.998
330
0.002
60
0.46
0.983
1,090
0.020
100
0.28
0.958
no
data
0.101
MPa= 1.00 bar = 1.02atm
100
Temperature (°C)
Figure 2.1 Water and TCE-NAPL vapor pressure as a
function of solution temperature at 1 bar or
1.02 atm of total gas phase pressure.
2.2 TCE-Water Phase Behavior
Boiling of a water and TCE-NAPL mixture occurs
at 73.4°C, which is below the boiling point of either
water (100°C) or TCE-NAPL (86.7°C). Boiling occurs
when the vapor pressure of a liquid mixture exceeds
the surrounding gas phase pressure. For a mixture
consisting of two immiscible liquids, such as water
and TCE-NAPL, the total vapor pressure is equal to
the sum of the vapor pressures of each pure constituent
(Dalton's Law: P = P° t + P° ), and the mixture will
v t water TCE"
boil when the total vapor pressure is equal to the local
gas phase pressure (Pt = Patm) (Atkins, 1998). As long
as TCE-NAPL is present, the mixture will boil at the
lower temperature (73.4°C), and since TCE-NAPL has
a greater vapor pressure than water, the composition of
the boiling vapor is 93% TCE and 7% water by weight
(Horvath, 1982). This phenomenon serves as the basis
for steam distillation, in which TCE is separated from
water at temperatures below its normal (pure) boiling
point.
The diagram in Figure 2.1 shows the pure phase
vapor pressures calculated using the Antoine equation
[P° = 10(A-
-------
In the presence of pure oxygen, the amount of TCE
degraded increased with increasing temperature
(Table 2.3) as indicated by the increase in the amount of
NaOH required to titrate the phenolphthalein indicator
to pink. The minimum amount of TCE degraded in
24 hours was 0.001% (mole basis) at 24°C, while
the maximum of 2.52% (mole basis) occurred at
a temperature of 130°C. The maximum first-order
half-life for the disappearance of TCE with pure oxygen
present was 475 years at 24°C, and 99 days at 130°C
based on the %TCE degraded in 24 hours. A lower
TCE degradation rate was observed with air present,
presumably due to the decrease in oxygen content. The
trend of increasing TCE degradation with temperature,
as observed with pure oxygen, was not observed above
90°C with air present. Carlisle and Levine (1932)
suggested that a shift to non-acid degradation products
occurred above 90°C when air was present since the
acid titration analysis method was only sensitive to
hydronium ions.
Mugdan and Wimmer (1934) quantified the degradation
products in the gas phase and NAPL after passing
oxygen gas through TCE-NAPL heated to temperatures
between 50 and 70°C. The gas-phase products included
hydrochloric acid (HC1), carbon monoxide (CO), and
phosgene (COC12), while the only product found in
the NAPL was dichloroacetyl chloride (CLHC2OC1).
After passing oxygen gas through TCE-NAPL heated
to 60°C, Kirkbride (1942) observed the formation of
additional products in the NAPL, including TCE epoxide
(C12COCHC1) and hexachlorobutene (C4H2C16).
McKinney et al. (1955) found that TCE-NAPL was
completely degraded to 15% gas-phase products and
85% NAPL products, on a molar basis, after passing
oxygen through TCE-NAPL at 70°C. The gas-phase
products included HC1, CO, and phosgene (COC12),
while the NAPL reaction product was water-soluble, had
a density of 1.545 g/mL at 20°C, and was a nearly equal
mixture of dichloroacetyl chloride (C12HC2OC1) and
TCE epoxide (C12COCHC1) (Table 2.4). McKinney et al.
(1955) completed experiments using 1) TCE stabilized
with triethylamine, 2) unstabilized TCE, and 3) TCE
that had been used for extracting oil from soybeans. The
triethylamine stabilizer was removed from TCE-NAPL
by the soybean oil extraction process meaning that this
used, or waste TCE, which was historically released
into the environment, was no longer stabilized against
reacting with oxygen. The complete degradation of
TCE-NAPL was reported after 193 hours of bubbling
oxygen through TCE-NAPL at 70°C. However, the
Table 2.3 Selected TCE Stability Test Results 24 hour test (Carlisle and Levine, 1932)
Temperature (°C)
24
40
50
70
90
110
130
TCE NAPL and Pure Oxygen (100% O2)
mL of NaOH added to 25
mL of water
H+ formed (mole)*
%TCE degraded (mole
basis)f
0.4
4.0xlO-6
0.00
416
4.2 xlO'3
1.50
440
4.4 xlO'3
1.59
500
5.0 xlO-3
1.80
520
5.2 xlO'3
1.87
590
5.9 xlO'3
2.13
700
7.0 xlO'3
2.52
TCE NAPL and Air (21% O2, 79% N2)
mL of NaOH added to 25
mL of water
H+ formed (mole)*
%TCE degraded (mole
basis)f
2.2
2.2xlO'5
0.01
~
2.5
2.5 xlO'5
0.01
192
1.9 xlO'3
0.69
191
1.9 xlO'3
0.69
119
1.2 xlO'3
0.43
130
1.3 xlO'3
0.47
* Calculated based on the results of Carlisle and Levine assuming 1 mole of OH: was equal to 1 mole H+.
f Assuming one Cl atom lost per TCE molecule (HC2C12 hydrolysis product).
-------
reaction rate was determined after addition of benzoyl
peroxide to TCE, where benzoyl peroxide is known to
generate peroxyl radical initiator compounds above 70°C
(Fossey et al, 1995). McKinney et al. (1955) also found
that partially-oxidized TCE influenced the rate of TCE
degradation.
contained NaOH at 10 and 30°C. McKinney et al. (1955)
speculated that dichloroacetic acid (DCAA) was the
hydrolysis product of dichloroacetyl chloride, while
the non-chlorinate organic acids were formed from the
hydrolysis of TCE epoxide.
Table 2.4 Oxygen and TCE Gas-Phase and NAPL
Degradation Products (McKinney et al.,
1955)
Reaction Product
hydrochloric acid (HC1)
carbon monoxide (CO)
phosgene (COC12)
dichloroacetyl chloride
(C12HC2OC1)
TCE epoxide
(C12COCHC1)
Phase
gas
gas
gas
NAPL
NAPL
Approximate
Amount
(% mole basis)
5
5
5
47
38
The reaction between oxygen and TCE-NAPL is
thought to involve a radical chain reaction mechanism
(Kaberdin and Potkin, 1994). Kucher et al. (1990) used
azo-bis-isobutyronitrile, a known temperature induced
radical chain initiator, to study the oxidation of TCE
in acetonitrile at 75°C. The reaction products included
TCE epoxide (C12COCHC1) and dichloroacetyl chloride
(C12HC2OC1) in a ratio of 3:1 after 1 hour of reaction
time.
In summary, exposing TCE-NAPL to oxygen in the
temperature range from 50 to 75°C, in the absence of
water, resulted in the formation of gas-phase and NAPL
reaction products that were thought to result from a
radical chain reaction mechanism. The next section
covers past experiments performed to examine the
compounds formed after exposing the TCE-NAPL and
oxygen degradation products to water.
2.3.2 Hydrolysis of TCE-NAPL Degradation
Products
Placing the TCE-NAPL degradation products
dichloroacetyl chloride and TCE epoxide in water
at 27 and 75°C resulted in the formation of gas- and
aqueous-phase products (Table 2.5). The gas-phase
products included CO and CO2, and the aqueous-phase
products included chloride ions, dicholoroacetic acid
(HC12C2OOH) along with oxoacetic (HOC2OOH)
and formic acid (HCOOH). The temperature of the
reaction (27 or 75°C) appeared to have little effect on
the distribution of hydrolysis products. Similar reaction
product distributions were also noted in water that
Table 2.5 Hydrolysis of TCE NAPL Degradation
Products (McKinney et al., 1955)
Reactor Temperature (°C)
Reaction Product
carbon monoxide
(CO)
carbon dioxide (CO2)
chloride ion (Cl")
dichloroacetic acid
(HC12C2OOH)
oxoacetic acid
(HOC2OOH)
formic acid
(HCOOH)
Phase
gas
gas
aqueous
aqueous
aqueous
aqueous
27
75
moles formed per
147.4 g of liquid
reaction products
0.087
0.005
1.410
0.740
0.180
0.033
0.100
na
1.200
0.850
0.057
0.060
na - not analyzed
The degradation of TCE epoxide was determined in a
separate experiment by Kline et al. (1978) that involved
injecting TCE epoxide into a solution of acetone
(0.2 mL) which contained 1.5 mL of 0.5 M sodium
phosphate buffer. Dichloroacetic acid (DCAA) was
the only reported degradation product after 4 minutes
at 37°C. Cai and Guengerich (1999) prepared TCE
epoxide from TCE using m-chloroperbenzoic acid, a
known radical initiator compound, and then placed TCE
epoxide in water at 0°C, and the degradation products
were measured as a function of pH. The products formed
included CO in the gas phase, formic acid, oxoacetic
acid, and DCAA in the aqueous phase over a pH range
from 0 to 14. The amount of CO and formic acid
formed increased with pH, the amount of oxoacetic acid
decreased with pH, and the amount of DCAA formed
was independent of pH.
Based on the work described above, TCE can be
transformed into dichloroacetyl chloride and TCE
epoxide after exposure to oxygen with TCE epoxide,
and potentially dichloroacetyl chloride, transformed
into DCAA upon exposure to water. Thus, DCAA is
one of the TCE degradation products anticipated to
-------
form during the thermal treatment of TCE contaminated
subsurface environments. DCAA is a colorless liquid at
room temperature (25°C) with a density of 1.57 g/mL,
a melting point of between 9 and 11°C, and a boiling
point of 197°C (Mallinckrodt and Baker, 2003b). DCAA
has been classified as a probable human carcinogen with
the maximum contaminant level goal of 0 ug/L (U.S.
EPA, 1998). DCAA is soluble in water with a practical
drinking water treatment level of 6 ug/L.
Haag et al. (1996) measured the rate of DCAA
disappearance from water heated to between 88 and
180°C as a function of NaOH concentration. The authors
obtained a half-life of 1.71 days for the hydrolysis of
DCAA at 103°C and pH 7, and a half-life of 1.27 hours
with 0.96 M of NaOH present. Prager et al. (2001)
showed that DCAA was hydrolyzed to chloride and
oxoacetic acid in heated water and that a temperature of
180°C was required to achieve complete degradation of
DCAA in 8 minutes.
The expected half-life for DCAA in water at pH 7
calculated from the Arrhenius parameters determined
by Prager et al. (2001) are given in Table 2.6. Thus
DCAA is expected to accumulate in water during the
degradation of TCE at temperatures less that 70°C while
DCAA is expected to be degraded into oxoacetic acid
within a few days at temperatures greater than 90°C
based on the half-lives given in Table 2.6.
Table 2.6 Rate of Dichloroacetic Acid (DCAA)
Disappearance from Heated Water
(Prager etal., 2001)
Temperature (°C)
60
70
80
90
100
120
First Order
Disappearance
Rate (I/day)
9.3xlO-4
4.8xlO'3
2.3xlO'2
9.8xlO-2
39.3xlO'2
5.1
Half-Life
(day)
742
143
30.5
7.0
1.8
0.14
2.3.3 Degradation of TCE Dissolved in Water at
Elevated Temperatures
Carlisle and Levine (1932) placed approximately 25 mL
of TCE-NAPL and 25 mL of water into 80 mL glass
vials with nitrogen gas in the headspace to determine if
TCE was degraded by water at elevated temperatures.
The vials were sealed and heated to fixed temperatures
between 50 and 150°C for 24 hours. After cooling the
vials to room temperature, the acid content of the water,
an indicator of TCE degradation via the formation of
acidic compounds, was determined by titrating with
a 0.01 M NaOH solution until the phenolphthalein
indicator turned pink. Less than 0.35% by weight of the
TCE-NAPL was lost assuming that one chlorine atom
was removed per TCE molecule, which led Carlisle and
Levine (1932) to conclude that TCE does not readily
hydrolyze in water. The observed reduction in TCE
content was attributed to the small amount of oxygen
within the vials at the beginning of each experiment.
Assuming that oxygen-saturated water (8 mg/L O2)
was present at the start of each experiment, then
approximately 6.25xlO~5 moles of oxygen were available
to react with TCE. Although this is a sufficient amount
of oxygen to account for the acid formed in vials heated
to 50°C, it is insufficient by 2 to 15 times to account
for the acid formed at temperatures greater than 50°C.
Carlisle and Levine (1932) stated that the thermal
decomposition of TCE at higher temperatures had
probably occurred, although no reaction mechanism or
reaction products were proposed or measured.
Dilling et al. (1975) completed a year-long experiment
at ambient conditions to measure the persistence of TCE
dissolved in water. Oxygen-saturated water (8 mg/L
O2) containing 1.0 mg/L of TCE was loaded into each
of three ice-cooled Pyrex tubes so that approximately
one-half of the tube volume was filled with solution
(i.e., gas phase was present) and then the tubes were
flame sealed. The sealed tubes were placed in a dark
container and stored at approximately 25°C. One tube
was destructively sampled after 6 months (182 days),
and the remaining two tubes were destructively sampled
after one year (365 days). Only aqueous samples
were collected and they were only analyzed for TCE
content. The reported first-order disappearance rate was
2.1xlO~3 day1, corresponding to a first-order half-life
of 326 days at 25°C. In a separate experiment, Pearson
and McConnell (1975) measured the persistence of
TCE in water using sealed glass bottles and reported
an estimated half-life of 2.5 years (912 days) for the
disappearance of TCE from water at 25°C.
Jeffers and Wolfe (1996) studied the disappearance
of TCE dissolved in water by placing approximately
0.3 mL of TCE contaminated water in glass tubes
and flame sealing both ends to create a sealed bulb
with approximately 0.02 mL of headspace. The TCE
contaminated water was prepared by mixing water
with TCE-NAPL for 2 minutes at room temperature to
yield an initial concentration estimated to be 10% of the
solubility limit for TCE (i.e., 110 mg/L for TCE) (Jeffers
et al., 1989). Experiments were completed in water with
0.01 M HC1 adjusted to pH 7 (Jeffers and Wolfe, 1996)
and in alkaline water containing from 0.1 to 0.001 M
NaOH (Jeffers et al., 1989; Jeffers and Wolfe, 1996).
-------
The water used was deionized, distilled, and boiled prior
to use, which probably resulted in low dissolved oxygen
content, however, no dissolved oxygen measurements
were reported. The bulbs were heated to temperatures
between 60 and 190°C for an unspecified period of time.
The bulbs were then cooled to room temperature and
the liquid content was analyzed by gas chromatography
for TCE content only. The only data reported were the
activation energy (120 kJ/mol) and pre-exponential
factor (5.Ox 109 I/minute) for the Arrhenius equation
[k = Axexp(-E/RT)] which was used to estimate a
first-order rate constant of 4.5*10'12 (1/mintues) along
with an estimated half-life of greater than 100,000 years
for the disappearance of TCE from water at ambient
temperature (25°C). The calculated first-order rate
constant at 90°C was 2.7x 10'8 (1/mintues) with a half-
life of approximately 49 years based on the Arrhenius
parameters reported by Jeffers and Wolfe (1996).
Gu and Siegrist (1997) increased the rate of TCE
disappearance from water by adding sodium hydroxide
(NaOH). They reported the complete disappearance of
TCE after 300 minutes from an aqueous solution that
had an initial TCE concentration of 630 mg/L after
amending with 2 M of NaOH and heating to greater than
60°C. The primary reaction products included chloride
and glycolic acid (HOCH2COOH), with intermediate
products including DCAA and monochloroacetic acid
(H2C1C2OOH). Nearly all the chlorine atoms originally
present as TCE were recovered as chloride in the reactor
effluent at 80°C, however, only 60% of the carbon
atoms introduced were recovered as organic acids. Gu
and Siegrist (1997) suggest that the unaccounted for
carbon may have been lost to gas phase degradation
products (i.e., CO2) that were not captured for analysis,
however, the alkaline solution would be expected to
serve as a trap for CO2. The authors also acknowledged
that the organic-acid detection limit (50 mg/L) for the
high pressure liquid chromatography (HPLC) analysis
method made it difficult to account for all the organic
acid degradation products.
Atwater et al. (1996) demonstrated the removal of TCE
from water using a flow-through reactor that contained
ruthenium and platinum on activated carbon granular
solids heated to between 90 and 120°C. The water
contained TCE at 15 mg/L and dissolved oxygen in
stoichiometric excess. When operated at 120°C, the
reactor was capable of removing 91% of the influent
TCE with a residence time of 12 seconds. However, the
appearance of chloroform (CHC13) in the reactor effluent
led Atwater et al. (1996) to increase the residence time to
5 minutes in order to achieve the complete degradation
of TCE without forming the unwanted chloroform
degradation product.
In summary, TCE dissolved in water is degraded with
a half-life ranging from approximately 1 year (Billing
et al., 1975) to greater than 100,000 years (Jeffers and
Wolfe, 1996) at room temperature (25°C). The rate of
TCE degradation can be increased by heating with the
half-life reduced to 49 years at 90°C based on results by
Jeffers and Wolfe (1996). The rate of TCE degradation
can be further increased by adding sodium hydroxide or
solid catalysts with the completed degradation of TCE
after 300 minutes at 60°C when amended with 2 M
NaOH and after 5 minutes at 120°C with the ruthenium
catalyst.
2.3.4 Thermal Degradation of TCE in a Water-
Filled Reactor
Knauss et al. (1999) measured the disappearance of TCE
from a water-filled reactor in an effort to demonstrate
that dissolved-phase TCE could be degraded in-situ
during thermal treatment of TCE contaminated aquifers.
The reactor consisted of a gold-walled cylinder with
a wall thickness of 0.01 inch and an outside diameter
of 1.75 inches by 7 inches long for a total volume of
approximately 250 mL (Seyfried et al., 1979). The
gold cylinder was sealed with a titanium head piece
that contained a single gold capillary tube for sample
collection. The gold cylinder and titanium seal were held
within a steel housing that was pressurized to between
0.1 to 3.4 mPa (1 to 340 bar) and heated to between
70 and 100°C. Pressurizing the gold cylinder caused
all reaction products to remain dissolved in water and
allowed small liquid samples to be forced from the
reactor through the gold capillary tube. Seyfried et al.
(1987) recommended rinsing the titanium head with
dilute HC1 solution followed by concentrated nitric acid
(HNO3) solution to remove any potential sources of
contamination. They also recommended heat treating the
titanium head at 300°C in air to develop an inert surface
oxide layer. For example, McCollom and Seewald
(2003) reported heating their titanium fittings in air
for 24 hours at 400°C prior to use in experiments on
the hydrothermal stability of formic acid. Knauss et al.
(1999) did not discuss procedures used to prepare their
reactor.
Knauss et al. (1999) reported results obtained for nine
separate experimental runs (Table 2.7). Each experiment
was completed with air-saturated water (8 mg/L O2) that
contained 150 mg/L of phosphate buffer (pH 7.2). Water
solutions with initial TCE concentrations between 0.3
and 21 mg/L were placed into the gold-walled reactor
with no headspace and heated to a fixed temperature
between 70 and 100°C at a constant pressure of 1 MPa
(10 bar) for an extended time period. Aqueous samples
were collected from the reactor periodically through
the gold capillary tube into 1 mL gas tight syringes.
Analysis for inorganic ions, including chloride, was
completed using a HPLC (HP 1090) connected to a
-------
Table 2.7 Summary of Knauss et al. (1999) Experimental Results
Experiment
TCE-35
TCE-37
TCE-39
TCE-40
TCE-41
TCE-42
TCE-43
TCE-51
TCE-53
Duration
(days)
6.11
19.2
11.1
43.3
4.24
7.28
2.23
7.2
3.31
Temperature
(°C)
100
81
90
70
90
90
90
90
90
Initial TCE
(mg/L)
5.96
5.87
21.30
5.50
1.45
2.87
1.62
6.09
5.15
Final Cl'
(mM)
0.151
0.114
0.400
0.145
0.035
0.016
0.049
5.563*
3.809*
Cl found/
Cl feed (%)
111
100
92
118
105
244
143
4012
3517
Final CO2
(mM)
0.157
0.116
0.310
0.182
0.070
N/A
0.056
0.120
0.104
CO2 found/
C02 feed (%)
173
152
107
224
315
N/A
246
130
144
* Data presented by Knauss et al. (1999) Table 1, but appears to be incorrectly reported.
Reported analytical detection limits: TCE = 0.0002 mM, Cr = 0.003 mM, CO2 = 0.068 mM.
conductivity detector. The aqueous phase TCE content
was determined using purge and trap separation with
analysis by a gas chromatograph connected to a flame
ionization detector.
Knauss et al. (1999) reported that chloride, hydronum
ions (H+), and dissolved CO2 were the only degradation
products detected during preliminary experiments
designed to look for intermediates. However, no analysis
of the experimental results was provided to demonstrate
that the initial amount of TCE in the reactor was
accounted for by the degradation products detected at the
end of the experiment (i.e., mass balance). Based on the
data presented by Knauss et al. (1999) for the amount of
chloride and CO2 detected, the carbon and chloride mass
balances were calculated and are provided in Table 2.7.
The final amount of chloride was within 11% of the
initial amount introduced as TCE (moles Cl" = 3 xmoles
TCE) for experiments TCE-35 through -41 but was
greater than the initial amount for experiments TCE-42
through -53. For example, the amount of chloride
reported in experiment TCE-42 was 244% of the initial
amount of TCE present in the reactor. The amount of
chlorine formed during experiments TCE-51 and -53
must have been reported incorrectly since these values
are orders-of-magnitude in excess of the amount of
chlorine initially present in the reactor as TCE.
There was greater variability in the carbon mass balance
shown in Table 2.7 as compared to the chloride balance,
which may have been due to the difficulty in measuring
dissolved phase CO2 at these low concentrations. Knauss
et al. (1999) determined the amount of dissolved total
CO2 formed, stated as the sum of carbonic acid (H2CO3),
bicarbonate (HCO3"), and carbonate (CO3"2), using
direct infrared (IR) spectroscopy. No description of the
IR analysis method (e.g., sorption bands used or scan
time) was provided, although the reported detection
limit was 0.068 mM. Falk and Miller (1992) studied
fourier-transform infrared (FTIR) spectroscopy as an
analytical method for determining the aqueous phase
concentration of total CO2 using the co-added signals
from 400 interfere grams (5-minute scan time) with 4
cm'1 band resolution. Falk and Miller (1992) concluded
that this was not a feasible analysis technique for HCO3"
or CO3"2 because the adsorption bands (1385 and 1360
cm'1, respectively) overlapped and were within the water
vapor region. Analysis of dissolved CO2 was found to
be feasible at the 2342.9 cm'1 adsorption band, with an
estimated detection limit of 0.4 mM. Falk and Miller
(1992) stated that increasing the scan time could have
decreased the detection limit. Burt and Rau (1994)
reported a dissolved CO2 detection limit of 0.24 mM.
Hence, the detection limit reported by Knauss et al.
(1999) is 3.5 to 6 times lower than those reported by
Burt and Rau (1994) and Falk and Miller (1992). The
ratio of the CO2 found to CO2 (as TCE) in the feed,
as reported in Table 2.7, was consistently greater than
one, which may indicate that the IR analysis method
employed by Knauss et al. (1999) was not sensitive to
the low CO2 concentrations because the signal to noise
ratio was too small to accurately resolve the 2342.9 cm'1
adsorption band.
Knauss et al. (1999) provided the following expression
for the rate of TCE disappearance based on the
experiments completed at 90°C when dissolved oxygen
-------
was in excess:
dCTI
dt
- = -5.77 + 1.06X1CTV
(2.1)
where Cg is the initial TCE concentration (mol/kg ~
molality). Although Equation 2.1 fit the experimental
data, analyzing the data reported by Knauss et al. (1999)
using traditional kinetic reaction modeling techniques
provides additional detail regarding the mechanism
of TCE disappearance. Figure 2.2 contains the
concentration of TCE with time, as measured by Knauss
et al. (1999), for four of the experiments completed at
90°C. Also shown in Figure 2.2 is the predicted TCE
concentration with time assuming a zero-order reaction
model described by:
dCT,
dt
- = -kn
(2.2)
2-M mg/L fJCE-39)
0 5.7 mg/L fJCE-51)
r 1.5 mg/L fJCE-41)
1, 1.6 mg/L CTCE-43)
Zero Order Kinetic Model
Figure 2.2 Measured concentration of TCE vs. time
along with the zero-order reaction model
fit.
The disappearance of TCE appears to follow the zero-
order reaction model over the initial two days, however,
the rate of TCE disappearance increased relative to
the zero-order rate after two days for Experiment
TCE-39 and decreased relative to the zero-order rate for
Experiments TCE-51, -41, and -43.
Figure 2.3 contains the same data shown in Figure 2.2
plotted as the natural log of the TCE concentration
normalized by the initial TCE concentration. Also
shown in Figure 2.3 is the change in normalized TCE
concentration as predicted according to a first-order
reaction model described by:
dt
- = -klCTCE or In
= -klt
(2.3)
First Order (TCE-39)
-•- 21.4 mg/L (TCE-39)
. O 5.7mg/L(TCE-51)
-T- 1.5mg/L(TCE-41)
-«f. 1.6mg/L(TCE-43)
1st Order Kinetic Model
Time (days)
Figure 2.3 Natural log of the measured TCE
concentration normalized by the initial
TCE concentration vs. time along with the
first-order reaction model fit.
The disappearance of TCE from the gold-walled reactor
did not follow the first-order reaction model in that the
rate of TCE disappearance was less than predicted by
the first-order model during the initial two days of each
experiment followed by an increase in the rate of TCE
disappearance relative to that predicted by the first-order
model. The rate of TCE disappearance also appears
to depend on the initial concentration of TCE with a
decrease in the rate of TCE disappearance corresponding
to an increase in the initial TCE concentration
(Figure 2.3).
Figure 2.4 contains the same data shown in Figures 2.2
and 2.3 but plotted as the reciprocal of the TCE
concentration vs. time, consistent with a second-order
reaction model describe by:
dCT.
1
(2.4)
dt
The disappearance of TCE during Experiment TCE-39
appears to follow the second-order reaction model
(Equation 2.4) over a period of four days but then the
rate of TCE disappearance deviates from that predicted
by the second-order model.
While the disappearance of TCE from the gold-
walled reactor operated at 90°C followed the zero-
order reaction model over the initial two days of each
experiment (Figure 2.2), the disappearance of TCE
was not described by the zero-, first-, or second-order
reaction models over the entire experimental period.
An alternative reaction model involves a radical chain
reaction mechanism which incorporates the following
reaction steps:
-------
-»-21.4mg/Ln'CE-39)
..0"5.7mg/LCTCE-51)
^T.- 1.5mg/LCTCE-41)
-v- 1.6mg/LCTCE-43)
2nd Order Kinetic Model
O 0.02 -
Time (days)
20 30
Time (day)
- 70
o"
o z_^
§
-so 2
a
-40 £
Figure 2.4 Reciprocal of the measured concentration
of TCE vs. time along with the second-
order reaction model fit.
Initiation: TCE + M* ->
Peroxyl Radical: TCE* + O
TCE* k*
->TCE-O* k*
Propagation: TCE + TCE-O * -> TCE epoxide +
dichloroacetylchloride k*
(2.5)
(2.6)
(2.7)
The symbol M* represents some radical initiator such
as the gold or titanium surface within the reactor or
chlorine radicals which transfer a single electron to TCE
and results in the formation of the TCE radical species
(TCE*). This three step TCE disappearance mechanism
was based on work by Kucher et al. (1990) and was used
to fit the results for the Knauss et al. (1999) experiment
completed at 70°C (TCE-40). The 70°C experiment
was chosen because the rate of TCE disappearance was
slower as compared to the 90°C experiments and thus
the features that indicate a radical chain mechanism,
including a delayed reaction rate during the initial three
days (reactor heat-up was less than one day) as the
concentration of the peroxyl radicals increased followed
by an increase in the TCE disappearance rate between
day 10 and 40 (Figure 2.5), were more pronounced.
Figure 2.5 shows the concentration of TCE vs. time data
as reported by Knauss et al. (1999), along with the best
fit using the reaction model described by Equations 2.5
through 2.7 determined using finite difference analysis
with time steps of 0.1 days. The disappearance of
TCE followed the radical chain model over the 43 day
experimental period with the reaction rate coefficients
for the initiation and peroxyl radical formation (k* and
k*) equal to 1.62 and 1.80 mmolal"1 day1 respectively,
while the rate coefficient for the peroxyl radical attack
on TCE (k*) was equal to 132.6 mmolal'1 day1.
The close agreement between the radical chain model
and the measured TCE disappearance for Experiment
TCE-40 does not necessarily validate this model.
Figure 2.5 Measured concentration of TCE vs. time
for Experiment TCE-40 and radical chain
reaction model fit.
However, the radical chain model is consistent with
two of the key observations made by Knauss et al.
(1999), namely that the rate of TCE disappearance
was dependent on the initial TCE concentration and
that the rate of TCE disappearance was independent of
the dissolved oxygen concentration as long as it was
in excess of the initial amount of TCE present. The
dependence on the initial TCE concentration is due to
the slow rate of radical initiation (k* and k*) compared
to the fast rate of the peroxyl-radical TCE reaction (k*).
That is, the formation of peroxyl radicals is the rate
limiting step (k*lk^* = 0.01).
Knauss et al. (1999) found that dissolved-phase TCE
could be degraded in a heated reactor with CO2 and
chloride as the only detected degradation products.
The time for one-half of the initial amount of TCE
to be degraded at 90°C ranged from approximately 1
to 5 days depending on the initial TCE concentration
(Figure 2.2). Knauss et al. (1999) reported that the rate
of TCE disappearance from the heated reactor was best
described using a pseudo first-order reaction model
(Equation 2.1). Analysis of the Knauss et al. (1999) data
provided herein suggests that a radical chain reaction
mechanism provided the best fit for the disappearance of
TCE over the entire experimental period.
2.3.5 Degradation of Gas-Phase TCE within
Heated Quartz Tubes
One method of treating unwanted waste TCE is
by feeding the waste into incinerators operated at
temperatures greater than 1,000°C. The degradation of
TCE and the products formed during the incineration
process have been studied by passing gas-phase TCE
through heated quartz tubes, trapping the effluent leaving
the quartz tubes, and analyzing the traps to determine
the TCE degradation products formed. The following
section provides details of past quartz tube experimental
-------
results with the goal of anticipating the degradation
products that might form during the high-temperature
treatment of subsurface environments contaminated with
TCE.
Pyrolysis is a general term used to describe organic
chemical reactions that occur at elevated temperatures
(Brown, 1980; Moss, 1994). Pyrolysis has also been
used to indicate high temperature gas-phase reactions
that occur in the absence of oxygen (Mulholland et
al., 1992), whereas pyrolysis has been used by others
to describe high temperature gas-phase reactions that
occur with oxygen present (Yasuhara and Morita, 1990).
The term pyrolysis is avoided in the following sections
because of the ambiguity with regard to the presence of
oxygen, instead of using this terminology, the oxygen
content of the gas phase will be stated when appropriate.
Graham et al. (1986) measured the amount of TCE
degraded after injecting TCE-NAPL into a heated quartz
tube (2 second residence time) as a function of quartz
tube temperature and oxygen concentration (Table 2.8).
The amount of TCE degraded increased with quartz tube
temperature and oxygen content. TCE degradation was
initiated at 600°C when the amount of oxygen present
was equal to the stoichiometric amount required for
the complete combustion of TCE (2.5 moles of O2 per
mole TCE) and decreased to 500°C when the amount
of oxygen present was in excess to the stoichiometric
amount. Temperatures greater than 800°C were required
to degrade 99% of the TCE introduced into the quartz
tube, independent of oxygen content. Graham et al.
(1986) detected the greatest number of TCE degradation
products at 750°C with some products detected at
1000°C after 99.9% of the parent TCE had been
degraded; however, the exact identity and distribution of
products was not reported.
Increasing the residence time within a quartz tube has
been shown to decrease the temperature at which TCE
degradation is initiated. Yasuhara and Morita (1990)
passed air (80% N2 and 20% O2) at 50 mL/min through
chilled TCE and into a quartz tube that was maintained
at a temperature between 300 and 800°C. The amount
of oxygen present, approximately 1.6 moles O2 per mole
TCE, was less than the stoichiometric amount required
for complete combustion. The residence time within the
quartz tube ranged from 23 to 43 seconds (Table 2.9),
and the degradation of TCE was initiated at a
temperature of less than 300°C with approximately 99%
of the TCE degraded at 500°C. Therefore, an increase in
the quartz tube residence time to greater than 20 seconds
resulted in 200 and 300°C reduction in the temperature
required for the initiation of TCE degradation and
for 99% destruction of TCE, respectively. Zhang
and Kennedy (2002) found that TCE degradation
within a surface boundary layer with residence time
of approximately 0.04 seconds did not occur until the
temperature reached 1000°C.
Yasuhara and Morita (1990) also quantified
condensable TCE degradation products by passing the
effluent gas stream leaving the quartz tube through
a dichloromethane (CH2C12) filled trap. The greatest
number of reaction products (23 compounds) was
identified in the dichloromethane trap fluid after 1
hour of feeding TCE into a quartz tube maintained
at 400°C. The most prevalent compounds found at
400°C included TCE, tetrachloroethylene (C2C14),
carbon tetrachloride (CC14), hexachloroethane (C2C16),
hexachlorobutadiene (C4C16), and hexachlorobenzene
(C6C16). TCE was not present (i.e., > 99% destruction) in
the dichloromethane trap fluid when the quartz tube was
maintained at temperatures greater than 600°C, while
tetrachloroethylene (C2C14) and carbon tetrachloride
(CC14) were detected at all temperatures between 300
and 800°C (Table 2.9).
Froese and Hutziner (1994) determined the amount of
chlorinated benzenes and phenols formed after passing
TCE and air (0.9 to 1.5 second residence time) through
a heated quartz tube that contained 0.5 grams of solids.
Table 2.8 Amount of TCE Degraded after Passing Through a Heated Quartz Tube
Residence Time of 2 Seconds (Graham et al., 1986)
Quartz Tube Temperature (°C)
Oxygen Content
None*
Stoichiometric
Excess
500
600
650
700
800
950
1000
Amount TCE degraded (wt%)
NM
NM
0
0
0
20
10
NM
70
40
85
90
75
98.5
98.5
99.9
NM
NM
NM
NM
NM
* Measurements completed by injecting a mixture of chlorobenzene, carbon tetrachloride, TCE, trichloro-trifluoroethane
and toluene. Pure TCE-NAPL was only used in the Stoichiometric and Excess experiments.
NM - not measured
-------
Table 2.9 Selected Degradation Products after Passing TCE through a Heated Quartz Tube (Yasuhara and
Morita, 1990)
Quartz Tube Temperature (°C)
Residence Time (seconds)
Selected Reaction Products
trichloroethylene (C2HC13)
tetrachloroethylene (C2C14)
carbon tetrachloride (CC14)
hexachloroethane (C2C16)
hexachlorobutadiene (C4C16)
hexachlorobenzene (C6C16)
Total
300
43
400
36
500
32
600
28
700
25
800
23
Amount present in liquid trap after 1 hour (% of Carbon in Feed)
65.51
0.30
0.02
0.18
0.21
0.00
66.21
31.02
6.95
0.76
1.94
0.94
0.24
41.84
0.17
11.78
3.58
0.27
0.91
0.43
17.13
0.00
13.92
6.81
0.06
0.18
0.09
21.06
0.00
4.68
4.48
0.04
0.00
0.01
9.21
0.00
0.03
2.87
0.00
0.00
0.00
2.90
The solids included fly ash collected from an incinerator
and a series of silica gel (SiO2) solids that were amended
with aluminum oxide (A12O3) (10 wt% Al), hematite
(Fe2O3) (10 wt% Fe), and copper oxide (CuO) (1 wt%
Cu) (Table 2.10). The effluent from the quartz tube
reactor was passed through a tube filled with activated
carbon (Carbotrap) to collect condensable degradation
products. The Carbotrap and quartz tube reactor
were extracted with dichloromethane, toluene, and a
1:1 hexane/dichloromethane mixture to determine the
amount of chlorinated benzenes and phenols formed.
The quartz tube and solids was extracted with toluene,
methanol, and the 1:1 hexane/dichloromethane mixture.
Lower molecular weight compounds such as carbon
tetrachloride (CC14) or tetrachloroethylene (C2C14) were
not analyzed.
The amount of chlorinated benzenes and phenols formed
as a function of temperature at 400, 500, and 600°C
was determined only with fly ash as the solid phase
(Table 2.10). A number of di-, tri-, tetra-, penta-, and
hexa- chlorinated benzene and phenol compounds were
detected, primarily condensed within the quartz tube and
on the solids. However, the penta- and hexachlorinated
compounds were formed in the greatest abundance. The
greatest amount of chlorinated compounds was formed
at 600°C, with hexachlorobenzene reported as the
predominant TCE degradation product (Table 2.10).
Table 2.10 Selected Degradation Products after
Passing TCE through a Heated Quartz
Tube Containing Fly Ash, 0.9 to
1.5 second Residence Time (Froese and
Hutzinger, 1994)
Quartz Tube
Temperature (°C)
Selected Reaction
Products
pentachlorobenzene
(C6HC15)
hexachlorobenzene
(C6C16)
pentachlorophenol
(C6C15OH)
400
500
600
wt% of TCE in feed
nd
nd
5xlO-6
IxlO'2
IxlO'2
3xlO-5
40.0xlO-3
SOO.OxlO-3
1.3xlO-3
Values estimated from graphs found in Froese and
Hutzinger (1994).
nd - amount not evident in graph
-------
The effect of solids on the amount of chlorinated
benzenes and phenols formed in the quartz tube was
evaluated at 600°C (Table 2.11). Only 11% of the quartz
tube volume was filled with solids, which were located
near the effluent end of the tube. The presence of the
silica gel (SiO2) was shown to have no impact on the
formation of chlorinated benzenes and phenols, while
the presence of fly ash and aluminum oxide increased the
formation of these products and the presence of hematite
(Fe2O3) and copper oxide (CuO) resulted in a decrease in
the amount of chlorinated benzenes and phenols formed.
Mulholland et al. (1992) determined the condensed-
phase products formed after passing TCE and nitrogen
(no oxygen present) through a heated quartz tube
(1.5 second residence time) at temperatures ranging
from 800 and 1200°C. The solids produced by the
degradation of TCE were collected on a filter, which
was subsequently rinsed with dichloromethane
(CH2C12) to determine the tar and soot fractions,
where soot was defined as the fraction that is
insoluble in dichloromethane. The chemical species
present in the tar fraction were identified using mass
spectrometry (MS), liquid chromatography, and IR
analysis. Approximately 10% (wt) of the TCE that
passed through the quartz tube heated to 800°C
was converted into tar. Hexachlorobenzene (C6C16),
hexachlorophenylacetylene (C8C16), octachlorostyrene
(C8C18), and octachloronaphthalene (C10C18) were the
most abundant compounds found in the tar based on MS
response.
Mulholland et al. (1992) suggested that
dichloroacetylene (C2C12) was a key intermediate
that led to the formation of the higher molecular
weight compounds. Wu and Lin (2004) detected
dichloroacetylene (C2C12) as one of the primary
TCE degradation products after passing TCE and a
stoichiometric amount of oxygen through a quartz tube
(residence time between 0. to 1.5 seconds) heated to
between 575 to 850°C.
Chang and Senkan (1989) measured the intermediates
and products that formed after burning a mixture of
TCE (22.6%), oxygen (33.1%), and argon (Ar; 44.3%),
where oxygen was in excess of the stoichiometric
requirement for complete combustion. The mixture
burned as a two-stage flame, with the initial stage at
approximately 1,000°C and the final stage at 1,500°C.
The final degradation products included (in order of
abundance) CO, HC1, chlorine gas (C12), and CO2.
Intermediates identified in the initial flame stage
included phosgene (COC12), tetrachloroethylene (C2C14),
carbon tetrachloride (CC14), dichloroacetylene (C2C12),
dichloroacetyl chloride (C12HC2OC1), and trichloroacetyl
chloride (C13C2OC1) among others (Table 2.12).
The data reported by Chang and Senkan (1989)
demonstrate that passing TCE through a temperature
gradient from 600 to 1,000°C, with oxygen present,
produced a variety of chlorinated compounds. These
chlorinated compounds were then transformed into non-
chlorinated carbon compounds (e.g., CO2), but only at
temperatures in excess of 1,000°C in the final stage of
the flame.
In summary, passing gas-phase TCE through quartz
tubes heated between 300 and 800°C resulted in the
formation of a wide variety of compounds from carbon
tetrachloride (CC14) and tetrachloroethylene (C2C14) to
hexachlorobenzene (C6C16). Thus, the possibility exists
that these compounds could be formed during the in-situ
thermal treatment of regions contaminated with TCE
where temperatures exceed 300°C. Based on the work
by Chang and Senkan (1989), temperatures in excess of
1,000°C would be required to destroy these compounds.
Reducing the amount of chlorinated degradation
Table 2.11 Selected Degradation Products at 600°C as a Function of Quartz Tube Solids Content, 0.9 to 1.5
second Residence Time (Froese and Hutzinger, 1994)
Quartz Tube Contents
(solids were 11% of
tube volume)
Empty
SiO2
Si02/Al203
SiO2/Fe2O3
SiO2/CuO
SiO2/Flyash
Products Formed wt% of TCE in feed
pentachlorobenzene
hexachlorobenzene
pentachlorophenol
0.0058
0.0025
nr
0.0035
0.0058
1.4xlO-4
0.012
0.080
8.2xlO-4
1.2xlO-4
2.2xlO-4
3.0xlO-4
3xlO'4
5xlO-4
0.4xlO-4
0.04
0.30
1.3xlO-3
Values estimated from graphs found in Froese and Hutzinger (1994).
nr - not reported
-------
Table 2.12 Selected Compounds in a TCE Flame with C1:H Ratio of 3 (Chang and Senkan, 1989)
Species in Flame
carbon monoxide (CO)
carbon dioxide (CO2)
hydrochloric acid gas (HC1)
chlorine gas (C12)
tetrachloroethylene (C2C14)
phosgene (COC12)
carbon tetrachloride (CC14)
dichloroacetyl chloride (C12HC2OC1)
trichloroacetyl chloride (C13C2OC1)
dichloroacetylene (C2C12)
hexachloropropene (C3C16)
hexachlorobutadiene (C4C16)
hexachloroethane (C2C16)
Initial Stage
(600-1000°C)
(% mole basis)
15.00
4.00
10.00
9.00
2.00
1.80
1.50
1.00
0.80
0.80
0.50
0.35
0.15
Final Stage (1500°C)
(% mole basis)
1
20
10
9
nd
nd
nd
nd
nd
nd
nd
nd
nd
Maximum measured values estimated from graphs found in Chang and Senkan (1989).
nd - below analysis detection limit
products and increasing the non-chlorinated products is
thought to be dependent on the amount of chlorine and
hydrogen in the high-temperature region. The following
section provides a discussion of experiments performed
to investigate changes in TCE degradation product
distribution as a function of the chlorine to hydrogen
(C1:H) ratio.
2.3.6 TCE Degradation Products as a Function
of the CI:H Ratio
Mulholland et al. (1992) suggested that the ratio of the
chlorine to hydrogen (C1:H) present in the quartz tube
would affect the type of degradation products formed.
With a C1:H ratio of less than one (C1:H < 1), chlorine
would preferentially react with hydrogen to form
HC1, and with a C1:H ratio of greater than 1 (C1:H>1),
chlorine was predicted to react with carbon to form
chlorinated hydrocarbons. The experiments completed
by Chang and Senkan (1989) and Mulholland et al.
(1992) represent results for TCE degradation with a C1:H
ratio of 3, thus the observed chlorinated hydrocarbons
were the expected TCE degradation products.
In contrast, Werner and Cool (2000) measured the
products formed during combustion of TCE using a
chlorine to hydrogen ratio of less than 1 (Table 2.13).
Here, the authors introduced TCE into a methane
(CH4) flame that consisted of 17% CH4, 35% O2,
46% Ar, and 2% TCE by volume for a C1:H ratio of
approximately 0.09. A two-stage flame was not observed
in contrast to the high C1:H ratio experiment by Chang
and Senkan (1989). The final combustion products
included (in order of abundance) H2O, CO2, CO, HC1,
methane (CH4), and O2. Intermediates identified in the
flame adjacent to the burner surface (200 to 1000°C)
included dichloroethylene (C2H2C12), vinyl chloride
(C2H3C1), ethylene (C2H4), dichloroethenol (C12C2HOH),
dichloroketene (C12C2O), chloroketene (C2HC1O), and
ketene (C2H2O), indicating that the oxidation state of the
TCE carbon atoms was being reduced within the flame
with C1:H ratio of less than one. The observation that
the TCE carbons were reduced in low C1:H ratio flames
is also supported by the results of Yang and Kennedy
(1993) who found acetylene, ethylene, and ethane were
the primary intermediates after introducing TCE into a
methane flame with C1:H ratio of 0.14.
-------
Table 2.13 Selected Compounds after Passing TCE
through a Flame with C1:H Ratio of 0.09
(Werner and Cool, 2000)
Species in Flame
carbon monoxide
(CO)
carbon dioxide
(C02)
hydrochloric acid
gas (HC1)
water (H2O)
methane (CH4)*
oxygen (O2)*
dichloroehtylene
(C2H2C12)
ethylene (C2H4)
vinyl chloride
(C2H3C1)
ketene
dichloroethenol
dichloroketene
chloroketene
Initial Flame
(200-1000°C)
(% mole basis)
6.00
3.00
2.50
17.00
3.00
12.00
0.50
0.40
0.07
0.05
0.01
0.01
0.01
Final Flame
(1500°C)
(% mole
basis)
6
11
2.5
19
0.5
0.5
nd
nd
nd
nd
nd
nd
nd
Maximum measured values estimated from graphs found
in Werner and Cool (2000).
* Present in feed
nd - below analysis detection limit
Zhang and Kennedy (2002) used methane (CH4),
dimethyl ether (C2H6O), and propane (C3H8) to study
the effect of decreasing the C1:H ratio on the destruction
of TCE flowing over a heated ceramic surface. There
was no change in the amount of TCE (0.5% TCE
and 99.5% N2) between the influent and effluent after
passing TCE past (residence time of 0.04 seconds) a
heated ceramic surface up to the temperature of 1000°C.
Adding methane (4% CH4, 0.5% TCE, and 95.5% N2)
to the TCE gas stream flowing past the heated ceramic
surface did not yield any measurable TCE destruction.
The addition of propane (4.4% C3H8, 0.5% TCE, and
95.1% N2) did cause some TCE degradation (the exact
amount was unspecified) whereas adding dimethyl ether
(4% CH4, 7 % C2H6O, 0.5% TCE, 88.5% N2) resulted in
the complete destruction of TCE. Zhang and Kennedy
(2002) speculated that the methyl radical (CH3') was
the primary species and the hydrogen radical (FT)
the secondary species involved in the destruction of
TCE based on the calculated concentrations of these
constituents at 1000°C.
Chuang and Bozzelli (1986) performed an experiment
using hydrogen gas (H2) and water as the hydrogen
sources for the transformation of chloroform (CHC13)
to HC1 within a heated quartz tube operated over a
temperature range of 550 to 1000°C. The residence times
were between 0.02 and 2 seconds, and the C1:H ratio
was approximately 0.14. Several intermediate products
were formed in the presence of hydrogen gas, including
dichloromethane, monochloromethane, and methane,
which indicated that the chloroform carbon oxidation
state had been reduced. The products formed when water
was used as the hydrogen source at temperatures below
950°C included PCE and TCE, indicating an increase
in the number of chlorine atoms per carbon or that the
chloroform carbon had been oxidized. Although the
complete destruction of chloroform was observed in the
presence of both hydrogen and water, hydrogen was able
to reduce chloroform beginning at 600°C, consistent
with the fact that water is more stable at elevated
temperatures than hydrogen. The ratio of chlorine
to hydrogen may affect the type of TCE degradation
products formed. With a C1:H ratio of greater than one,
chlorine produced from the degradation of TCE may
react with the remaining TCE and TCE degradation
products to form chlorinated compounds. With a C1:H
ratio of less than one, chlorine may react with hydrogen
atoms to yield HC1 and prevent the formation of
unwanted chlorinated hydrocarbons.
Taylor et. al (1990) developed a Thermal Stability
Ranking under fuel-rich, low-Cl conditions for the
hazardous organic compounds listed in Appendix VIII
of 40 CFR Part 261.3. Of the 320 compounds on this
list, TCE is classified as one of the 77 compounds most
resistant to decomposition. The decomposition of
these resistant compounds is believed to be dominated
by bimolecular decomposition processes. At high
temperatures (800 - 1000°C), decomposition is
mainly affected by H atom methathesis and Cl atom
displacement reactions, while at lower temperatures
reactive organic and inorganic radicals may be present,
creating increased molecular weight compounds.
Chang and Senkan (1989) found that the major reaction
pathway for TCE in a flame (greater than 750°C) was
by chlorine radical (Cl') attack, with the formation of
C2C13' and HC1. In cooler parts of the flame, addition
of C2C13' and CIO' to TCE is also important, with the
formation of higher molecular weight species such as
hexachlorobenzene.
-------
2.4 Operational Conditions of In Situ
Thermal Treatment Technologies
The following sections provide a brief review of
steam flushing, thermal conductive heating, and
electrical resistive heating, three thermal remediation
techniques commonly used to treat chlorinated solvent
contaminated aquifers (U.S. EPA, 2004). The purpose
of these sections is to discuss subsurface conditions
(e.g., temperature, residence time) that might exist
during application of these thermal technologies. The
increase in subsurface temperature resulting from
thermal treatments will lead to substantial changes in the
distribution of volatile organic contaminants between
the solid, liquid, and gas phases. For example, the
vapor pressures and Henry's Law constants of TCE and
PCE increase markedly with temperature (Heron et al.,
1998; Sleep and Ma, 1997), indicating that a substantial
fraction of the contaminant mass will exist in the gas
phase during thermal remediation. As a result, thermal
treatment systems incorporate a vacuum extraction
system to recover gas-phase contaminants from the
subsurface. The following sections are not intended to
provide a comprehensive review of thermal treatment
methods or their application, but rather to provide the
reader with a brief summary of the basic approaches and
principles.
2.4.1 Steam Flushing
Injecting steam into the subsurface through wells has
been shown to be effective for mobilizing fluids, heating
the subsurface, and removing TCE (Udell, 1997).
When steam is injected into the subsurface, it initially
condenses, releasing the latent heat of vaporization
which heats the soil and interstitial fluids. With
continued injection, three distinct zones develop: a
nearly isothermal steam zone at steam temperature
surrounding the injection point, a relatively narrow
variable temperature zone, and an isothermal zone at
ambient temperature. The temperature of the injected
steam is limited by the injection depth, as the injection
pressure must remain below the overburden pressure
to avoid steam breakthrough at the ground surface,
and specifying the steam pressure fixes the steam
temperature. In general, pressures of less than 0.5 psi
ft of overburden are employed in unconsolidated media
to avoid breakthrough of the steam front at the surface
(Davis, 1997; Udell, 1997). Thus, steam temperatures
are normally in the range of 120 to 140°C for injection
depths between 40 and 120 feet below ground surface.
Interstital fluids are displaced in front of the steam
zone, and residual liquids held in the pore space and
adsorbed contaminants are vaporized when the steam
zone reaches them, and are then transported to the
steam front. Within the ambient temperature zone,
these vapors can condense, forming a contaminant
bank. When the contaminant is a dense, non-aqueous
phase liquid (DNAPL) such as TCE, this can lead to
downward mobilization of the DNAPL (Schmidt et
al., 1998; Kaslusky and Udell, 2002). Thus, when
treating DNAPLs such as TCE, air co-injection is now
commonly used. Co-injecting air with the steam has the
effect of carrying part of the heat as well as contaminant
vapors to the extraction wells more quickly, creating a
much wider variable temperature zone and somewhat
reducing the temperatures within the steam zone
(Kaslusky and Udell, 2002).
The depth of steam injection sets a maximum injection
pressure that can be used, but some field applications of
steam injection have chosen to employ lower pressures
and temperatures. Thus, a wide range of temperatures
have been used in field applications of steam injection
remediation. The following two case studies describe
steam injection demonstration projects where the
purpose of the demonstration was to document in situ
oxidation of TCE during steam injection.
Steam-air coinjection was used for the steam injection
demonstrataion at Launch Complex 34, Cape Canaveral,
Florida to recover TCE from the surficial aquifer. The
vendor claimed that hydrous pyrolysis/oxidation (HPO)
of TCE would occur during the demonstration. Air
was injected into the deep wells at rates ranging from
approximately 3,000 - 8,000 Ibs/day. Steam was
injected at rates generally in the range of 15,000 -
80,000 Ibs/day, which lead to subsurface temperatures
in the steam zone ranging from about 90 to 150°C
(Integrated Water Resources, 2003). Groundwater
dissolved oxygen concentrations prior to steam
injection were less than 1 mg/1, the redox potential
was negative, and cis-l,2-DCE was detected in most
groundwater samples with concentrations as high as
260 mg/1. Vinyl chloride was not detected prior to steam
injection (detection limits were as high as 83 mg/1).
After steam and air co-injection, dissolved oxygen
remained low (oxygen is not readily soluble in hot
water), redox potential ranged from negative to positive
values, c/5-l,2-DCE was detected at concentrations as
high as 52 mg/1, and vinyl chloride was now detected
at concentrations as high as 0.128 mg/1. Chloride
concentrations and alkalinity, which would have been
expected to increase if significant TCE oxidation
had occurred, instead showed decreases (Interagency
DNAPL Consortium, 2002). However, the inside-out
design of this remediation system caused groundwater
from outside the steam treatment area to be continuously
pulled towards the treatment area, and this may have
masked changes in groundwater concentrations that
occurred within the heated zone.
A steam enhanced extraction demonstration using a
single injection and extraction well was carried out
at Beale Air Force Base (Carroll et al., 2004). The
purpose of the demonstration was to produce conditions
in the subsurface for HPO of TCE to be optimized,
-------
thereby minimizing the need to extract volatilized
contaminants. Oxygen was co-injected with the steam
into a single well for 48 hours, then the well was
shut in to allow reaction to occur before dual phase
extraction was initiated and a volume of water in
excess of the amount injected as steam was extracted.
Measured temperatures reached as high as 113°C. Prior
to the demonstration the groundwater was essentially
anoxic and dechlorination of the TCE was occurring;
however, during the demonstration dissolved oxygen
levels increased to 4 - 5.5 mg/1 within the heated
zone. Thus it appears that some groundwater samples
during steam and air injection were supersaturated with
oxygen. Trends in dissolved oxygen, alkalinity and ion
concentrations suggest that HPO occurred; however, the
authors acknowledge that, "Distinguishing between the
relative importance of HPO, Steam Enhanced Extraction
(SEE) and other potential mechanisms for destruction
or removal of chlorinated VOCs has proved to be
problematic with the available data."
Thus, during steam injection remediation, volatile
contaminants are found dissolved in the aqueous
phase, as a separate phase (NAPL), and in the gas
phase. Contaminants are exposed to soil at elevated
temperatures, oxygen, and steam during their migration
toward vacuum extraction wells. Steam injection
systems are normally designed to allow for a couple
weeks to a month for the steam zone to reach the
extraction wells. Establishing steam zones throughout
the area to be heated, allowing heat conduction into
low permeability zones that do not readily take steam,
and pressure cycling to aid in contaminant recovery
usually requires several months of time for a full scale
steam remediation. Residence times for vapors in the
subsurface will depend on the air and steam injection
rates and the vacuum pressure at the recovery wells,
but can be expected to be relatively short, on the order
of a day. However, dissolved and adsorbed phase
contaminants may be exposed to high temperatures
for considerably longer time periods, on the order of
months.
The residence time of gas-phase contaminants in the
heated subsurface is difficult to anticipate since each
steam drive application is tailored to specific subsurface
conditions. The residence time at the Visalia Superfund
site was determined by measuring the time to recover
xenon and helium gas tracers. The initial displacement
stage residence time was 10 hours between an injection
and extraction well that were 24 meters apart based
on the xenon tracer (Newmark et al., 1998). For most
steam flushing applications, the residence time of gas-
phase contaminants flowing through heated soil during
transport to vapor recovery wells is expected to be less
than one day.
2.4.2 Thermal Conductive Heating
Steel wells containing resistive heating elements can be
used to heat subsurface regions contaminated with TCE
via thermal conduction heating (TCH) with recovery of
the volatilized TCE accomplished by applying vacuum
extraction through the heated steel well screens. Heating
elements within the heater wells typically operate at
temperatures between 650 and 800°C, and the heat
is conducted radially into the soil. Vapors that are
generated move countercurrent to the direction of the
heat as they are extracted at the heater wells. When
volatile organic compunds (VOCs) are to be remediated
using thermal conductive heating, the target treatment
temperature for the midpoint between wells is generally
100°C. Steep temperature gradients are formed, drying
out the soil nearest the heater wells. If there is a water
table present and the soil has sufficient permeability,
much of the area between heater wells will remain
saturated and thus at temperatures below 100°C. For
treatment zones above the water table, greater drying
of the soil will occur, and the high temperature zones
will extend further from the heater wells. Heater well
spacings on the order of 12 to 18 feet are commonly
used when treating VOCs, with treatment times on the
order of 4 to 6 months.
In one of the earliest applications of TCH for the
remediation of VOCs, Vinegar et al. (1999) reported
using heater well temperatures between 745 and 900°C
to remediate a site located in Portland, Indiana, that was
contaminated with TCE and PCE. The heater wells were
located every 7.5 feet resulting in one heater well every
50 square feet with the soil temperatures between heater
wells ranging from 100 up to 250°C after heating for 5
months. Stegemeier and Vinegar (2001) speculate that
the high temperature soil region (500 to 700°C) extends
approximately 1 foot radially from each heater well.
The residence time of TCE within this 1-foot region is
controlled by the rate of gas extraction. The treatment
zone area was 7,500 square feet to a depth of 18 feet,
which represents a treatment volume of approximately
40,500 cubic feet, assuming a gas filled porosity of
0.3. A single 1,800 cfm blower was used to extract gas
from 130 heater/vacuum wells. Using the reported
blower capacity and the estimated treatment volume
yields an estimate for the overall gas residence time of
22.5 minutes (40,500 ft3 + 1,800 ftVmin = 22.5 min).
A first-cut estimate of the gas residence time within the
1-foot high-temperature zone that is adjacent to each
heater well is 0.5 minutes according to:
71 I2 ft2
18ft
130 wells
0.3 ft3 void
ft3 soil
min
4,669 ft3
— 0.5 mm
(2.8)
-------
1,800 ft3
min
773 K
298 K
which represents the circular area around the well, the
length of well screen, the number of wells, and the
porosity of soil divided by flow rate. Here, the gas flow
rate was corrected to 500°C using the ideal gas law
according to:
= 4,669 fWmin at 500°C
(2.9)
A second application of thermal conductive heating
used a 3,000 cfm blower to extract gas from 761 heater/
vacuum wells at a TCE-contaminated site in Eugene, OR
(Stegemeier and Vinegar, 2001). The estimated residence
time in the 1-foot high temperature zone for this case
was 1.1 minutes, calculated following the approach
described above. These residence time estimates assume
that gas is uniformly removed from each well and that
no preferential flow channels exist. In reality, the gas
flow through each 1-foot high temperature zone could
range from seconds to days depending on the soil gas
permeability and pressure distribution within the vacuum
manifold system.
In conductive heating remedial systems, gas from
uncontaminated subsurface regions flows into the
contaminated treatment zone that have been heated
to temperatures between 100 to 250°C, and becomes
saturated with the volatile contaminants (e.g., TCE).
Vapors entering the heated treatment zone are not
likely to be atmospheric air with 21% oxygen as vapor
barriers are commonly used if the treatment area is close
to the ground surface. Thus, the vapors more likely
come from other subsurface regions that have lower
oxygen content due to microbiologic consumption. The
contaminant-saturated gas then travels through a high
temperature region located adjacent to each heater/
vacuum extraction well. Baker and Kuhlman (2002)
suggest that TCE degradation occurs as vapors migrate
through the soil region adjacent to the heater/vacuum
well, which may reach temperatures of 500 to 700°C.
This high temperature zone is claimed to function as a
"packed-bed reactor that is hot enough to accomplish
rapid decomposition by either pyrolysis, if oxygen is
deficient, or by oxidation, if oxygen is available" (Baker
and Kuhlman, 2002).
According to data from Kim and Choo (1983) on
pyrolysis of TCE via a dehydrochlorination reaction
pathway which forms dichloroacetylene (presented by
Baker and Kuhlman, 2002), the destruction of 99%
of the TCE entering the 500 to 700°C region would
require a residence time of approximately 7 days at
500°C and 7 seconds at 700°C. The thought is that TCE
enters into the high temperature zone at 500°C and is
transformed into intermediate products. The intermediate
products formed from TCE degradation at 500°C then
undergo further transformations as they encounter
temperatures near 700°C closer to the heater/vacuum
wells. As discussed in Section 2.3.5, dichloroacetylene
is thought to lead to the formation of higher molecular
weight compounds. To date, there is no data available to
determine the amount of TCE destruction or the reaction
byproducts that may be expected from conductive
heating remediation of TCE.
Based on the examples discussed above, thermal
conductive heating may result in gas phase TCE being
exposed to temperatures ranging from 100 to 250°C for a
period of time greater than a month, and to temperatures
ranging from 500 to 700°C for a period from days to
seconds. Vapors in the treatment area are likely to be
low in oxygen content. Water vapor may or may not be
present depending on whether the target treatment area is
above or below the water table. To date, well-controlled
experiments have not been conducted to confirm or
refute TCE reactivity and byproduct formation under
these conditions.
2.4.3 Electrical Resistive Heating
Electrical resistance heating (ERH) passes an electrical
current through the subsurface zone targeted for
treatment. The current is actually conducted by water
within the pores of the soil, and the resistance of soils
to carrying current results in resistive heating (Beyke
and Fleming, 2005). The electrical current is delivered
through steel rods (electrodes) installed into the
contaminated soil. If the system is above the water table
or in low permeability soils, water is injected into the
annular space between the soil and electrodes during
heating to prevent the soil adjacent to the electrodes
from drying out. The temperature of the subsurface
increases relatively slowly, and reaches 100 to 120°C,
the boiling point of water at the local pressure, generally
after about two to three months of heating time. Because
water is needed in the pore spaces to carry current
between electrodes, the temperature cannot go higher
than the boiling point of water at the local pressure,
and the soil cannot be allowed to dry out. Volatile
organic compounds and water within the heated soil are
subsequently recovered by vacuum extraction through
the electrodes or through extraction wells located within
the heated soil region. For example, a TCE-contaminated
aquifer was heated to a temperature of 100°C for 6
months using ERH (Beyke and Fleming, 2005), and
approximately 30,000 pounds of TCE was recovered
through vapor extraction wells located within the heated
aquifer formation. Heron et al. (1998) demonstrated
electrical resistive heating in a controlled laboratory-
scale box filled with TCE-contaminated water. An
average temperature of 90°C was maintained within
the box for a period of 25 days, and a single centrally-
located extraction well was used to recover gas-phase
TCE.
-------
Between 2003 and 2007, three areas containing spent
chlorinated solvent NAPL at the Fort Lewis Army
Logistics Center's East Gate Disposal Yard (EGDY) near
Tacoma, Washington, were remediated using ERH. The
main objective of the remediation was to recover TCE
DNAPL, which was the source for a plume more than
a mile long downgradient of the site. Pre-remediation
site characterization activities showed that there were
considerable petroleum hydrocarbons present as well
as cis-l,2-DCE from the reductive dechlorination of
TCE. During the remediation of the third NAPL area,
groundwater and vapor samples were collected and
analyzed to determine if dechlorination (either biotic
or abiotic) continued as the temperature of the site was
increased. Vapor and dissolved phase consitutents that
were analyzed for included methane, ethane, ethene, and
acetylene, as well as cis- and trans-l,2-DCE, and vinyl
chloride. Although cis-l,2-DCE concentrations were
high in some areas of EGDY prior to the remediation,
very little vinyl chloride was found. Baseline samples
for gases analyzed prior to the initiation of heating
showed significant methane concentrations, but ethane,
ethene, and acetylene were below detection limits.
Within a week after the initiation of heating, at which
time two small areas of the site had reached temperatures
as high as 70°C while most of the site was between
10 and 15°C, the ethene concentration in the effluent
vapors was 30 percent of the TCE concentration. During
the second week of operations, ethane concentrations
also reached approximately 30 percent of the TCE
concentration. However, as heating of the site
continued, the concentrations of ethane and ethene both
dropped significantly in the effluent vapors. Acetylene
also appeared in the effluent stream soon after heating
began, but its concentration remained low. Methane
concentrations ranged from 100 percent to 2 percent
of the TCE concentrations, with an average of about
20 percent. Once the average temperatures of the
site reached approximately 90°C, ethane, ethene, and
acetylene concentrations again fell to below detection
limits, while methane concentrations remained as
significant portion of the effluent vapors (Davis, 2007).
In summary, electrical resistive heating will result in the
exposure of liquid contaminants to slowly increasing
temperatures, which eventually can reach up to 120°C.
Once a volatile contaminant is vaporized, it likely will
not remain in the heated subsurface for a period of time
greater than one day. There will always be water vapor
present in the effluent stream, however, the vapors are
likely to be low in oxygen content. Ambient air will
not likely enter the heated subsurface region; if the
treatment zone is near the ground surface a vapor barrier
is commonly used to prevent atmospheric air from
being pulled into the system. The residence time of gas
phase contaminants that pass through heated soil during
transport to vapor recovery wells is expected to be less
than one day.
2.4.4 Hybrid Thermal Technologies
Steam flushing and electrical resistive heating may
be implemented simultaneously. For example, the
thermal treatment design for the Young-Rainey Science,
Technology, and Research Center located in Largo,
Florida, involved a combination of electrical resistance
heating to initially heat surrounding and underlying soils,
followed by steam drive to flush contaminants from soils
within the preheated region to extraction wells (U.S.
DOE, 2003). In practice, this would mean potentially
exposing TCE to temperatures approaching 120°C
for more than a day before driving the TCE from high
permeability soils via steam flushing.
-------
3.0
TCE Degradation in Flow-Through Quartz
Tube Reactors
3.1 Introduction
A series of controlled experiments were conducted to
investigate TCE reactivity and degradation product
formation when passing gas-phase TCE through a quartz
tube reactor that was heated to temperatures ranging
from 60 and 480°C. The experiments were intended
to approximate conditions that could occur during the
extraction of gas-phase TCE either from or through a
heated subsurface region, and to explore the potential
effects of oxygen content, water content and solids
(Ottawa sand) on TCE reactivity under these conditions.
However, these experiments were not intended to
precisely replicate all of the conditions and variables that
could be encountered in the field. The first experimental
series was performed to determine the minimum
temperature at which TCE degradation products could be
detected when the quartz tube reactor was partially filled
with Ottawa sand. The second set of experiments was
performed with the reactor tube empty at temperatures
up to 480°C in order to obtain experimental data that
could be directly compared to results reported by
Yasuhara and Morita (1990) for a similar empty reactor
system. The third experimental series focused on
determining TCE degradation products from quartz
tube reactors that were either partially- or completely-
filled with Ottawa sand, and operated at 420°C. The
fourth experimental series was designed to investigate
the effect of water, introduced as vapor into the TCE
saturated carrier gas, on the degradation products formed
within an empty quartz tube reactor operated at 420°C.
Results obtained from the first four experiments were
used to refine methods and design the fifth experimental
series, which was conducted to determine the amounts
of TCE degradation products formed as a function of
three experimental variables; 1) quartz tube temperature,
2) oxygen content, and 3) water vapor content. A
summary of the quartz tube reactor experiments is
presented in Table 3.1.
A description of the quartz tube reactor apparatus and
the methods used to collect and measure chemical
compounds found in the gas-phase effluent exiting the
quartz tube are presented in Section 3.2. Experimental
methods specific to each of the five experimental
series are given in Section 3.2.3. Results of all five
experimental series are presented in Section 3.3.
Detailed experimental procedures related to quartz tube
preparation, temperature profile, TCE introduction,
and effluent trapping are provided in Appendix A. The
final two sections of the chapter provide a discussion
(Section 3.4) of the quartz tube results in terms of
potential chemical reaction mechanisms and a summary
(Section 3.5) of quartz tube experimental results and
conclusions.
Table 3.1 Summary of Flow Through Quartz Tube Experiments
Experimental
Series
1
2
o
3
4
5
Quartz Tube Contents
100 grams sand
Empty
100 grams sand and
completely sand filled
Empty
Empty
Temperature Range
(°C)
24 to 420
22 to 480
420
420
120 to 420
Purpose
Identify degradation products
Compare with literature results
Partially vs. completely sand filled quartz
tube
Evaluate effect of water
Degradation products as a function of oxygen
and water vapor content
-------
3.2 Experimental Materials and
Methods
3.2.1 Materials
Ottawa sand (ASTM 20-30 mesh) was obtained from
U.S. Silica Co. (Berkeley Springs, WV). Prior to use,
2,000 grams of sand was placed into a 3 L capacity
Pyrex glass drying tray and a 1 N solution of nitric acid
was added to completely cover the sand. The sand
was then allowed to soak in the nitric acid solution for
30 minutes before draining the excess liquid. The 1 N
nitric acid soaking process was repeated, and then the
sand was rinsed in DI water. The wet sand was placed
into a drying oven and heated to 130°C for 3 hours to
remove excess moisture, and then heated to 200°C for
and additional 2 hours to complete the drying process.
A 2 L bottle of 99.5% American Chemical Society
(ACS) reagent-grade TCE was obtained from Sigma-
Aldrich, Inc. (Milwaukee, WI). The TCE was not
stabilized with an anti-oxidant. TCE from the 2 L bottle
was used in all experiments described herein, and for the
preparation of calibration solutions. A dedicated 40 mL
vial with Teflon lined septum affixed with a screw cap
was periodically filled with TCE from the 2 L bottle.
TCE used in each experiment was dispensed directly
from the 40 mL vial. The 40 mL vial and 2 L bottle of
TCE were stored in the flammable storage locker at
room temperature.
3.2.2 Quartz Tube Apparatus
The quartz tube experimental system consisted of a
quartz-glass tube, a quartz-glass pre-mix chamber,
and a quartz-glass effluent transition (Figure 3.1). The
quartz tube was General Electric Type 124 fused quartz
(Technical Glass Products, Mentor, OH), with an outer
diameter (OD) of 38 mm, wall thickness of 2 mm, and
a length of 53 cm. Two quartz tubes were used in the
experiments: the first was customized (Lillie Glassware,
Marietta, GA) by installing a slotted quartz-glass shelf
located at the midpoint of the quartz tube (quartz tube
#1) to hold sand, and the second section of quartz
tube was completely open (quartz tube #2). The pre-
mix chamber was manufactured by Lillie Glassware
(Marietta, GA) to provide an approximate 70 mL
chamber in which influent gas and TCE could mix
before entering the quartz tube. The effluent connection
was custom made by Lillie Glassware (Marietta, GA)
from quartz glass to transition the gas flow from the 38
mm OD tube down to an 8 mm OD tube. The pre-mix
chamber and effluent transition were connected to the
quartz tube using custom made 38 mm ID, 316 grade
stainless steel (316-SS) adapters (Swagelok Co., Salon,
OH) fitted with Viton® o-rings.
Pre-Mix Chamber
Quartz Tube •/
Carrier Gas Direction
316-SS Adaptors
Effluent Transition
To Effluent Traps
Figure 3.1 Quartz tube experimental apparatus.
The flow rate of the TCE-free carrier gas at 22°C
entering into the experimental apparatus was determined
using a mass flow meter (Model 179A, MKS
Instruments, Andover, MA). The mass flow meter was
calibrated using an ADM 2000 gas flow meter (J&W
Scientific, Folsom, CA) that had been calibrated by
California Integrated Coordinators (Placerville, CA).
The pressure within the reaction system was determined
using a pressure transducer (Honeywell, Freeport, IL),
which was calibrated using water- (0.02 to 0.05 bar)
and mercury- (0.07 to 0.7 bar) filled manometers. Both
the mass flow meter and pressure transducer were
connected to a data logger (CR23X, Campbell Scientific,
Logan UT) to automatically record the gas flow rate and
pressure within the quartz tube system at one second
intervals during each isothermal experiment.
The effluent end of the quartz tube was connected to
a 40 mL screw-thread vial via a 10 cm long section of
1/16 inch OD poly ether-ether ketone (PEEK) tubing.
The PEEK tubing was affixed to the 38 to 8 mm effluent
transition by a 316-SS Swagelok (Solon, OH) union with
Teflon ferules and was inserted through a pre-drilled
hole in a Teflon lined septum affixed to the 40 mL vial
with an open-hole screw cap. The 40 mL vial contained
approximately 30 mL of dichloromethane (DCM) and
-------
was located in a 500 mL beaker filled with crashed ice.
The purpose of the DCM filled 40 mL vial was to trap
all compounds with greater than two carbons (e.g., PCE)
exiting the quartz tube. After passing the TCE saturated
carrier gas through the heated quartz tube, the DCM
trap was removed, sealed with a Teflon lined septum
without holes, and weighed using an analytical balance
(Model* AG245, Mettler-Toledo, Columbus, OH). The
weight of the DCM filled trap was used to estimate the
volume of DCM in the 40 mL vial assuming a DCM
density of 1.325 g/mL.
Post-experiment sand samples (5 grams each) were
collected and placed into separate 25 mL test tubes along
with 2 mL of DI water. Each test tube was sealed with a
Teflon lined septum affixed with an aluminum crimp and
then placed in a freezer. The frozen sand samples were
then processed using a hot solvent extraction method
that involved incubating the 5 gram soil samples in a 1:1
(volume) iso-octane and methanol mixture at 85°C for
24 hours. Previous studies performed at Georgia Tech
have shown that the hot solvent extraction procedure,
which is based on the methods of Sawhney et al. (1988)
and Huang and Pignatello (1990), is equivalent to
Soxhlet extraction for chlorinated benzenes (Prytula,
1998).
3.2.3 Experimental Procedures
3.2.3.1 Experimental Series 1
Nine (9) experiments were completed with one quartz-
tube apparatus assembly for tube oven temperatures of
24, 40, 60, 120, 180, 240, 300, 360, and 420°C at 1 atm
of carrier gas pressure during the first experimental
series. These experiments were conducted to determine
if TCE could be degraded within the quartz tube, and
if the quartz tube could withstand operating in the
temperature range from ambient to 600°C. The first
experimental series involved passing dry air (Airgas-
South, Inc., Marietta, GA) through a gas-washing bottle
(250 mL Pyrex) filled with TCE-NAPL at 22°C. The
TCE saturated air flowed through the quartz tube that
contained 100 grams of acid washed 20-30 mesh Ottawa
sand positioned on quartz-glass wool (Technical Glass
Products, Mentor, OH) and held at the midpoint of
the quartz tube by a slotted quartz-glass shelf (quartz
tube #1). The TCE-saturated air was passed through
the reactor system for a period of approximately
16 minutes, which represented 3 pore volumes (320 mL
total quartz tube pore volume assuming sand porosity
of 0.3), and resulted in approximately 450 mg of TCE
being transferred into the quartz tube. The mass of TCE
delivered to the tube was determined gravimetrically by
measuring the weight (PG503-S, Mettler-Toledo, Inc.,
Columbus, OH) of the TCE-filled gas washing bottle
before and after each 16 minute ran period. The airflow
rate was approximately 60 mL/min, which resulted in
a residence time of approximately 5 minutes at 25°C.
The effluent from the heated quartz tube passed through
a vial containing dichloromethane (DCM) to trap all
condensable products; no gas samples were collected.
At the end of 16 minutes, the airflow was stopped and
the gas pressure within the quartz tube apparatus was
monitored for a period of 5 minutes to test for gas
leaks. No additional volume of air was passed through
the quartz tube after each 16 minute TCE introduction
period meaning that at least 1 pore volume of TCE
saturated air remained in the tube at the end of each
16 minute trial. The temperature of the quartz tube
was increased, allowed to stabilize for 30 minutes, and
the 16 minute TCE introduction period was repeated.
The apparatus was not disassembled between each
isothermal ran.
3.2.3.2 Experimental Series 2
The second experimental series involved passing dry
air that was saturated with TCE through the quartz
tube used in the first experimental series (quartz tube
#1), but without sand present (empty). The empty
reactor experiments were performed to replicate
results reported by Yasuhara and Morita (1990), who
passed TCE-saturated air through an empty quartz tube
over a temperature range of 300 to 800°C. Nine (9)
experiments were completed (including 2 replicates at
120 and 240°C) at temperatures of 24, 120, 240, 300,
360, 420, and 480°C at 1 atm of carrier gas pressure.
The quartz tube apparatus was disassembled and
decontaminated between each isothermal trial completed
above 300°C due to the presence of degradation
products observed during the first experimental series.
The airflow rate was approximately 60 mL/min, which
resulted in a residence time of approximately 5 minutes
at 25°C. The TCE-saturated air flowed through the
empty tube reactor for a period of 20 minutes, which
represented 3 pore volumes (400 mL total quartz tube
pore volume) and resulted in approximately 700 mg of
TCE being transferred into the quartz tube. Although
these experiments were intended to include temperatures
up to 600°C, experiments were only completed up to
480°C because the quartz tube shattered into many
small pieces while heating to 540°C. The destruction
of the quartz tube at 540°C was unexpected as these
tubes were rated to 1,200°C (Technical Glass Products,
Mentor, OH). Repeated attempts to operate the reactor at
temperatures above 500°C resulted in destruction of the
quartz tubes.
3.2.3.3 Experimental Series 3
The third experimental series involved passing dry
air saturated with TCE through a quartz tube that was
partially or completely filled with Ottawa sand. The
purpose of this experiment was to determine if filling
the empty volume of the tube with sand had any effect
-------
on the amount of TCE degraded and the degradation
products formed. These experiments were completed
at a single temperature of 420°C, since operation of the
reactor at this temperature in the second experimental
series was found to degrade a significant amount of TCE
and produce detectable amounts of degradation products.
The partially sand-filled quartz tube experiment was
completed with 100 g of acid treated 20-30 mesh Ottawa
sand located on a quartz shelf (quartz tube #1), and was
operated under the same flow conditions (16 minute TCE
introduction period) as in the first experimental series.
The completely sand-filled experiment was conducted
with 700 grams of acid treated 20-30 mesh Ottawa sand
located in the quartz without the glass shelf (quartz
tube #2). TCE-saturated air was passed through the
completely filled quartz tube operated at a temperature
of 420°C and 1 atm carrier gas pressure for a period
of approximately 16 minutes. This time represents
5.4 tube pore volumes (177 mL pore volume assuming
a porosity of 0.3) and resulted in 551 mgof TCE
being introduced into the tube. The airflow rate was
approximately 60 mL/min, which resulted in a residence
time of approximately 3 minutes at 25°C. At the end
of each 16 minute run, airflow was stopped and the
system pressure was monitored for a period of 3 minutes
to test for gas leaks. Hence, at least 1 pore volume of
TCE saturated air remained in the apparatus. A second
trapping sequence was then completed by passing dry
air without TCE through the apparatus for a period of
20 minutes to flush any residual gas-phase TCE from
the tube. After effluent trapping was completed, the tube
was capped and allowed to cool to room temperature
overnight. The apparatus was disassembled the following
day and 5 gram sand samples were collected from near
the entrance, at the mid-point, and exit of the sand-filled
quartz tube. A sample of the glass wool located at the
exit of the quartz tube was also collected. The sand and
glass wool samples were processed using a hot solvent
extraction method described above (Section 3.2.2).
3.2.3.4 Experimental Series 4
The fourth experimental series involved passing
humidified air and gas-phase TCE through an empty
quartz tube operated at 420°C. The experiments were
completed using three different carrier gas humidity
levels; 0, 25, and 100% relative humidity (RH). The 25%
RH experiment used a 1:3 ratio of air that had passed
through a water-filled gas washing bottle at 22°C and air
saturated with TCE at 22°C. The 100% RH experiment
involved passing air through a gas washing bottle that
contained an approximate 1:1 (volume) mixture of
TCE-NAPL and water at 22°C. A 1.6 L Tedlar® bag
was used to capture all of the gas leaving the DCM trap.
The gas within the Tedlar® bag was analyzed for CO2
content using a gas chromatograph (GC) equipped with a
thermal conductivity detector (TCD).
3.2.3.5 Experimental Series 5
The fifth experimental series was designed based
on the results of the initial four experimental series
with the goal of accounting for all TCE degradation
products in an effort to close the mass balance. The
experiments were planned so that the amount of each
TCE degradation product could be determined as a
function of three experimental variables; 1) quartz tube
temperature, 2) oxygen content, and 3) water vapor
content (Table 3.2). The quartz tube temperatures were
fixed at 120, 240, and 420°C to reduce the number of
individual experiments in the series.
For the fifth experimental series, TCE was introduced
into the pre-mix chamber as neat liquid TCE at a fixed
rate of 0.68 mL/hr using a syringe pump (Model 11,
Harvard Apparatus, Holliston, MA). This allowed the
rate of TCE introduction to be fixed while adjusting the
amount of water entering the quartz tube to vary the
chlorine to hydrogen ratio inside the apparatus. Ultra
zero grade air (Airgas-South, Inc., Marietta, GA) or
nitrogen (Airgas-South, Inc., Marietta, GA) was used as
the carrier gases. Ultra zero grade air (UZA) was used
as received, while the nitrogen was passed through an
oxygen trap (Alltech Associates, Inc., Deerfield, IL)
before entering the quartz tube.
The residence time through the quartz tube was fixed at
approximately 4.3 minutes for all isothermal trials during
the fifth experimental series. This represented a gas
flow rate of approximately 85 mL/min (at 22°C) with
the empty quartz tube at 120°C, which was the upper
measurement limit of the mass flow meter and thus
Table 3.2 Fifth Experimental Series Matrix
Tube (°C)
120, 240, 420
120, 240, 420
Inlet (°C)
20, 80, 100
20, 80, 100
Reactor
Contents
Empty
Empty
Carrier Gas
N2
Zero air
Runs
3x3x1x1 = 9
3x3x1x1 = 9
Sum 9x2 = 18
Variable
baseline
oxygen
-------
fixed the residence time for all subsequent experiments
completed at temperatures greater than 120°C. The
gas flow rates to achieve a 4.3 minute residence time
were calculated using the ideal gas law to correct for
the gas expansion within the quartz tube at elevated
temperatures. The gas flow rate used with the empty
quartz tube at 240°C was approximately 65 mL/min (at
22°C), and approximately 48 mL/min (at 22°C) with the
quartz tube at 420°C.
3.2.4 Analytical Methods
The concentration of TCE in the DCM trap fluid was
determined by collecting three, 2 mL DCM samples,
which were placed in separate autosampler vials. An
internal standard, 1,1,1-trichloroethane, was added to
each vial, which was then capped with Teflon®-lined
septa affixed by a crimp seal. The analysis of DCM
fluids consisted of using an automatic liquid sampler
(HP6890) to inject 1 uL of DCM into a GC (Hewlett
Packard 6890) equipped with a 30 m by 0.32 mm OD
DB-5 column (Agilent Technologies, Palo Alto, CA)
that was connected to a Flame lonization Detector
(FID). The GC inlet was operated at 9.45 psi in the split
mode (10:1) at 200°C with helium as the carrier gas
and a constant column flow rate of 2 mL/min. The GC
oven temperature was isothermal at 50°C for 8 minutes
followed by a 20°C/min ramp to 150°C. The FID was
operated at 300°C with 400 mL/min of air, 30 mL/min
hydrogen, and 40 mL/min of nitrogen as the makeup
gas.
TCE calibration standards in the concentration range
from 8,000 to 20,000 mg/L were analyzed by GC/FID
to determine the amount of TCE in the DCM trap
fluid. The calibration standards were prepared by first
adding approximately 30 mL of DCM to 50 mL glass
volumetric flasks (50+0.05mL at 20°C), which were
sealed with ground-glass stoppers. The initial weight of
the flasks and DCM was determined using an analytical
balance (Model# AG245, Mettler-Toledo, Columbus,
OH) after allowing the flasks to stand for a period of
30 minutes. Neat TCE was then introduced into each
flask using a gas tight syringe, the stopper inserted into
each flask, and then the weight of each 50 mL flask with
TCE was recorded. Each flask was filled to the indicator
mark with DCM, sealed and inverted several times to
mix the solution. The concentration of each calibration
solution was calculated using the weight of TCE added
and the volume of DCM. GC/FID analysis of an EPA
8240B/8260 A Matrix Spike Mix (Sigma-Aldrich
#47412) spiked into DCM was performed to verify TCE
retention time and concentration.
The identity of compounds associated with unidentified
chromatographic peaks from the GC/FID analysis of
DCM trap fluids was determined using a GC (Varian
Star 3600CX) equipped with a 30 m by 0.25 mm OD
CP-Sil 8 CB Low Bleed/MS capillary column (Varian)
connected to a Varian Saturn 2000 Ion Trap Mass
Spectrometer (MS). Compounds were identified using
software (SaturnView ver. 5.41, Varian, Inc., Palo Alto,
CA) that matched their mass spectra with reference mass
spectra in the NIST/EPA/NIH Mass Spectral Library
(NIST98). Compounds were identified when their mass
spectrum fit with a matching NIST98 library spectrum
with purity of greater than 700. The mass spectrometer
was tuned to optimize the detector voltage (EM-Voltage)
and mass axis calibrated using perflurorotributylamine
(FC-43) prior to each use.
3.3 Experimental Results
3.3.1 Results of Experimental Series 1-4
The amount of TCE recovered in the DCM trap fluid
of the first experimental series is shown in Figure 3.2
as function of temperature. At temperatures less than
240°C, the amount of TCE recovered in the DCM trap
fluid was less than the amount delivered to the quartz
tube. The missing mass of TCE was attributed, in part,
to the residual gas-phase TCE that remained within the
quartz tube as no attempt was made to flush TCE from
the apparatus after the 16 minute introduction period.
At temperatures of 300°C and above, PCE and carbon
tetrachloride (CC14) were detected in the DCM trap fluid,
and the amounts of PCE and CC14 detected continued to
increase as the tube temperature was raised to 420°C.
The experimental series was terminated prior to reaching
600°C due to repeated failure of the quartz tube at
temperatures of 480°C and above. These data indicate
that TCE underwent thermally induced degradation
when flowing through a quartz tube containing 100 g
of Ottawa sand and heated to temperatures greater than
300°C. PCE and CC14 were the degradation products
detected in the DCM trap and represent chlorinated
oxidation products that were expected during TCE
degradation because the C1:H ratio was equal to 3 and
TCE was the only source of chlorine and hydrogen
(Mulholland et al., 1992, see also Section 2.3.6)
Results of the second experimental series, in which
gas-phase TCE was passed through an empty quartz
tube reactor at temperatures up to 480°C, are shown in
Figure 3.3. The lower mass recoveries at 22 and 120°C
occurred prior to implementation of flushing step to
capture residual gas phase mass in the reactor. Dry air
without TCE was passed through the quartz tube for
15 minutes following the 20 minute TCE introduction
period in all experiments completed above 240°C.
Consequently, the amount of TCE collected in the DCM
trap for the 240°C yielded a mass recovery of-98%,
indicating that TCE was not degraded at temperatures
less than 240°C for these experimental conditions.
-------
0 100 200 300 400 500
Reactor Temperature (°C)
Figure 3.2 Amounts of TCE, PCE and carbon
tetrachloride (CC14) recovered during the
first experimental series.
I
s
•6
Reactor Temperature (°C)
Figure 3.3 Amounts of TCE, tetrachloroethylene
(PCE) and carbon tetrachloride (CC14)
recovered during the second experimental
senes.
Starting at temperatures of 300 to 360°C, TCE, PCE and
carbon tetrachloride (CC14) were detected in the effluent
DCM trap. The extent of TCE degradation increased as
the temperature was raised to 480°C, with no measurable
amounts of TCE observed at 480°C. These findings are
consistent with data presented by Yasuhara and Morita
(1990), which followed similar trends; a sharp decline
in TCE coincident with the appearance of PCE and CC14
between 300 and 500°C, and no detectable amounts of
TCE at temperatures of 500°C and above.
Results of the second experimental series indicate
that TCE degradation products in addition to PCE
and CC14 are likely to have formed at 420°C as the
amount of TCE, PCE, and CC14 detected in the DCM
trap accounted for less than 35% of the amount of TCE
introduced into the quartz tube reactor. The missing
degradation products are hypothesized to be CO and
CO2 based on the experimental work described in
Section 2.3.5 (Chang and Senkan, 1989). Experimental
procedures were modified in experimental series five to
collect and measure CO and CO2.
The results of the third experimental series, in which
TCE was introduced into quartz tube reactors that were
either partially- or completely-filled with Ottawa sand
at 420°C, are summarized in Table 3.3. The completely
sand-filled tube produced more PCE and had lower
TCE recovery than the tube containing 100 grams of
sand (Table 3.3). No CC14 was detected in the DCM
trap during the third experimental series as compared
to the significant amount detected during the initial
experimental series. However, the first and third
experimental series results are not directly comparable
since a decontamination step was not completed between
each isothermal experiment during the first experimental
series. PCE was the only compound detected in the iso-
octane extracts and only from the sand sample collected
at the tube exit. An initial sand-filled experiment was
completed for tube temperatures of 120 and 240°C,
however, the sand-filled quartz tube shattered at 400°C
while heating to 420°C. Based on this experience and
the quartz tube failures during preliminary experiments,
it was concluded that sand-filled quartz tubes are not
capable of consistently withstanding temperatures
greater than 400°C.
In the fourth experimental series, gas containing both
water and TCE passed through empty quartz tube
reactors at 420°C. Results from these experiments
indicate that increasing the quartz tube water-vapor
content led to an increase in the amount of CO2 and
CC14 detected (Table 3.4). However, the amount of
CO2 represented less than 5% of the total amount of
carbon introduced into the quartz tube as TCE. Thus,
we hypothesize that additional degradation products
formed during the fourth experimental series. This
observation led to the development of a method to detect
carbon monoxide (CO) and CO2 along with the use of an
additional liquid filled trap to determine the amount of
phosgene (COC12) formed.
The ratio of chlorine to hydrogen in the fourth
experimental series was greater than one. Based on the
prior experimental results described in Section 2.3.6
(Mulholland et al., 1992; Werner and Cool, 2000), a
C1:H of less than 1.0 provides an insufficient amount
of hydrogen to reduce the amount of chlorinated
degradation products. This line of reasoning provides
justification to conduct experiments with chlorine to
hydrogen ratios of less than one (i.e., more water vapor)
to determine if TCE could be degraded without forming
chlorinated degradation products.
-------
Table 3.3 Amounts of TCE, PCE, and CC14 from the Sand Filled Quartz Tube at 420°C
Amount of Sand
100 grams
partially filled
(Exp. Series 1)
100 grams
partially filled
(Exp. Series 3)
700 grams completely
filled
(Exp. Series 3)
Amount TCE
Introduced (mg)
456
503
551
Amount TCE
Recovered (mg)
273
226
172
TCE
Recovered (%)
60
45
31
Amount PCE
(mg)
63
156
250
Amount
CC14 (mg)
28.5
None*
Nonef
* Other degredation products detected included hexachloroethane, penta- and hexachloro-propene, and penta- and hexachloro-
butadiene.
f Other degredation products detected included penta- and hexachloroethane.
Table 3.4 Amount of TCE, PCE, CC14, and CO2 from an Empty Quartz Tube at 420°C
%RH
0
25
100
TCE Introduced
(mmol)
9.51
5.96
8.69
TCE Recovered
(mmol)
4.23
2.64
4.69
PCE (mmol)
0.29
0.19
0.34
CC14 (mmol)
0.18
0.38
0.39
CO2 (mmol)
0.08
0.11
0.38
3.3.2 Results of Experimental Series 5
Empty quartz-tube experiments were completed for
tube temperatures of 120, 240, and 420°C at carrier
gas pressures of 1 atm. Separate experiments were
completed with the inlet at 22, 60, and 100°C for each
tube temperature to evaluate the effect of increasing the
quartz-tube water content on TCE degradation and the
degradation products formed. Separate experiments were
completed with nitrogen and air (UZA) as the carrier
gas to evaluate the effect of oxygen on TCE degradation
(Table 3.5). The following sections describe the recovery
of TCE after being introduced into the empty heated
quartz-tube, along with the identity and quantity of TCE
degradation products detected in the DCM trap, water
rinse, iso-octane rinse, and Tedlar® bag. The latter
sample devices were installed in an attempt to collect
and identify a wider range of potential degradation
products and to improve mass balance closure.
Table 3.5 Experiments Completed for the Fifth Experimental Series
Tube (°C)
120, 240, 420
120, 240, 420
Inlet (°C)
20, 80, 100
20, 80, 100
Tube Contents
Empty
Empty
Carrier Gas
N2
Zero air
Variable
baseline
oxygen
-------
3.3.2.7 TCE Recovery
The amount of TCE detected in the DCM trap with
respect to the amount introduced into the quartz tube as
a function of quartz tube temperature and carrier gas is
shown in Figure 3.4. The average recovery of TCE with
nitrogen as the carrier gas was greater than 94% at all the
reactor temperatures and inlet stream relative humidities.
With air as the carrier gas, the average recovery of TCE
was greater than 94% for tube temperatures of 120
and 240°C but dropped to approximately 53% for the
tube at 420°C. The amount of TCE recovered at each
tube temperature shown in Figure 3.4 represents the
average for the three experiments completed at different
quartz-tube water contents (inlet temperatures). Table
3.6 contains the average amount of TCE recovered as
a function of quartz-tube water content with nitrogen
as the carrier gas. The low TCE recovery observed for
the 60°C inlet temperature and 120°C tube temperature
was due to a leak in the experimental apparatus, and this
value was not used to calculate the average recovery
at 120°C shown in Figure 3.4. The amount of TCE
recovered as a function of quartz-tube water content
with air as the carrier gas is shown in Table 3.7. The
average values reported in Tables 3.6 and 3.7 are shown
graphically in Figure 3.4.
I I Nitrogen
100% Recovery
Quartz Tube Temperature (°C)
Figure 3.4 Recovery of TCE with nitrogen or air as
the carrier gas, averaged over three relative
humidities (inlet temperatures).
Tables 3.6 and 3.7 also contain the Pearson correlation
coefficient (R2) that describes the variability in TCE
recovery as a function of quartz tube water content along
with the P-Value, which indicates the significance of
the correlation between the quartz-tube water content
and TCE recovery. The statistical calculations were
performed using MINITAB software (Release 14,
Minitab Inc., State College, PA). A linear relationship
between the quartz tube water content and the amount
of TCE recovered was not obtained (R2<0.3 and
P-Value>0.5) for the 240°C experiment with nitrogen
as the carrier gas (Table 3.6). In contrast, the increase
in TCE recovery was linearly related (R2=0.994 and
P-Value=0.05) to the increase in water content for the
420°C experiment. With air as the carrier gas, there was
no linear relationship (R2<0.2 and P-Value>0.5) between
the quartz tube water content and the amount of TCE
recovered for the 120°C and 420°C experiments. The
decrease in TCE recovery was linearly related (R2=0.998
and P-Value=0.03) to the increase in water content
for the 240°C experiment with air as the carrier gas
(Table 3.7).
Table 3.6 TCE Recovery with Nitrogen as the
Carrier Gas
(%Recovery = TCE in DCM Trap H- TCE injected x 100)
Inlet
Temperature
(°C)
(Relative
Humidity, %)
22(2)
60 (20)
100 (95)
Average ±
Standard
Deviation
Correlation
Coefficient
(R2)
Quartz Tube Temperature (°C)
120
91.2+6.0
*68.0+4.6
96.9+4.5
94.0+7.6
NA
240
98.7+1.3
101.7+2.9
96.2+1.0
98.8+3.3
0.227
(P-Value
=0.68)
420
93.1+1.9
95.5+1.5
99.1+1.9
95.9+3.1
0.994
(P-Value
=0.05)
* Leak in experimental system, average based on 22 and
100°C inlet temperatures.
Table 3.7 TCE Recovery with Air as the Carrier Gas
(%Recovery = TCE in DCM Trap H- TCE injected x 100)
Inlet
Temperature
(°C)
(Relative
Humidity, %)
22(2)
60 (20)
100 (95)
Average ±
Standard
Deviation
Correlation
Coefficient
(R2)
Quartz Tube Temperature (°C)
120
95.3+1.2
97.7+2.1
96.3+1.7
96.4+3.0
0.144
(P-Value
=0.75)
240
96.2+1.2
94.1+0.6
92.0+1.5
94.1+2.0
0.998
(P-Value
=0.03)
420
52.4+1.2
54.6+0.2
51.7+0.9
52.9+1.5
0.052
(P-Value
=0.85)
-------
TCE recovery of less than 100% is one indication that
TCE degradation had occurred within the experimental
apparatus. However, accounting for the amount of
missing TCE with the amount of degradation products
detected (i.e., mass balance) provides a greater level
of confidence to conclude that TCE was degraded as
opposed to experimental loss which could result from a
gas leak. The amount of each TCE degradation product
detected is presented in Sections 3.3.2.2 through 3.3.2.5,
and the mass balance between the missing amount of
TCE from each isothermal experiment and the amount
of degradation products detected are provided in
Section 3.3.2.6.
3.3.2.2 Compounds in theDCM Trap
The effluent carrier gas leaving the quartz tube reactor
passed through an ice cooled, 40 mL vial filled
with DCM to trap condensable TCE degradation
products that were soluble in DCM. These anticipated
TCE degradation products included PCE, CC14,
hexachlorobutadiene, and hexachlorobenzene based
on the results of the first four experimental series and
work by Yasuhara and Morita (1990) presented in
Section 2.3.5. Samples from the DCM trap for each iso-
thermal trial were initially analyzed by GC/FID. TCE
was the only compound detected in the DCM trap for the
120 and 240°C experiments regardless of carrier gas or
quartz tube water content. Thus, at temperatures below
240°C no TCE degradation products were detected
in the DCM trap, consistent with results obtained in
experimental series 2.
Chromatograms obtained from the analysis of DCM
trap samples from each 420°C experiment with nitrogen
as the carrier gas contained up to four unidentified
peaks in addition to the TCE peak. Subsequent GC/MS
analysis of the DCM samples showed that the four
other compounds were titanium tetrachloride (TiCl4),
pentachlorobutadiene (C4HC15), hexachlorobutadiene
(C4C16), and pentachlorobenzene (C6HC15). These
compounds are thought to represent the primary TCE
degradation products since no other organic compounds
were detected in the DCM rinse of the quartz tube
apparatus. Titanium tetrachloride was detected in all
three experiments performed at 420°C without oxygen
present. Titanium tetrachloride was thought to form due
to a reaction between the gas phase chlorine from the
degradation of TCE and titanium in the 316-SS Cajon
connectors. The amount of titanium tetrachloride was
estimated using the response factor that was determined
for CC14.
The gas phase concentrations (ppmv or uL/L gas)
of each detected compound are reported in
Table 3.8. The amounts of pentachlorobutadiene and
pentachlorobenzene shown in Table 3.8 were estimated
using a GC/FID response factor of 3.0 (concentration
of compound/chromatogram area). This response factor
was based on the average response factor (3.01+0.41)
determined from calibration solutions of TCE, PCE,
hexachloroethane, and hexachlorobutadiene. The
actual concentration values were calculated using the
ideal gas law to convert the moles of each compound
detected into a gas phase volume at 25°C and 1 atm.
The calculated volume of gas for each compound
was normalized to the duration of TCE injection
(30 minutes) and the gas residence time (4.3 minutes)
in the quartz tube (500 mL). These data are intended
to represent the concentration of each compound that
would be present in a gas sample collected from the
quartz tube during TCE injection. The purpose of
Table 3.8 Concentration (ppmv) of Compounds Detected in the DCM Trap for the 420°C Experiments with
Nitrogen as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity %)
TCE*
titanium tetrachloride (TiCl4)
pentachlorobutadiene (C4HC15)
hexachlorobutadiene (C4C16) *
pentachlorobenzene (C6HC15)
Total
22
(2)
24,010+491
864+37
395+7.0
14+0.0
22+0.4
24,571
60
(20)
24,517+397
524+30
305+5.0
5+0.3
1<
24,905
100
(95)
25,992+501
186+26
175+3.0
1<
1<
26,192
Values reported as ppmv in the quartz tube (mL gas phase compoundn-30 minx4.3 minn-500 mL).
* Amount determined using calibration solutions. Amount of other compounds was estimated.
-------
expressing these results as gas phase concentrations
is to gain insight into conditions during thermal
remediation where these compounds are anticipated to
be present in the gas phase. Thus, TCE would be the
dominant compound present (>97% by volume) with
relatively minor amounts (<1% by volume) of penta-
and hexachlorobutadiene, and pentachlorobenzene in a
representative volume containing TCE that was heated
to 420°C and was absent of oxygen.
The relative humidity of the injected gas stream was
varying the inlet temperature from 22 to 100°C to
evaluate the effects of water on TCE degradation
product formation. It was hypothesized that a reduction
in the C1:H ratio would result in the formation of less
chlorinated TCE degradation products. Increasing the
quartz tube water content led to a decrease in the number
and amount of degradation products detected (Table 3.8),
similar to the trend noted with the increase in TCE
recovery (Table 3.6, R2=0.994 and P-Value=0.05).
Thus, increasing the amount of water in the apparatus
at 420°C with nitrogen as the carrier gas did not shift
the TCE degradation products toward less chlorinated
compounds, but had the effect of decreasing the amount
of TCE degraded based on the DCM trap results.
With air as the carrier gas, up to 14 peaks, in addition
to TCE, were observed in the GC/FID chromatograms
of DCM trap samples for the reactor experiments
conducted at 420°C. The identities of the compounds
associated with the unknown GC/FID peaks were
determined by GC/MS analysis (Table 3.9), with
the exception of the peak with retention time of
2.05 minutes, which could not be identified due to
interference from DCM co-elution. The compound
with retention time of 2.05 minutes may have been
dichloroethane (C2H4C12) based on the elution order
for a test mix of chlorinated solvents available in the
Table 3.9 Concentration (ppmv) of Compounds Detected in the DCM Trap for the 420°C Experiments with Air
as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity %)
*TCE
Unknown (reported as C2H4C12)
*chloroform (CHC13)
*carbon tetrachloride (CC14)
*tetrachloroethylene (C2C14)
trichloroacetic acid methyl ester
(C13C2O2CH3)
tetrachloropropene (C3H2C14)
pentachlorocyclopropane (C3HC15)
perchlorocyclobutenone (C4C14O)
*hexachloroethane (C2C16)
tetrachlorobutadiene (C4H2C14)
titanium tetrachloride (TiCl4)
pentachlorobutadiene (C4HC15)
hexachloropropene (C3C16)
*hexachlorobutadiene (C4C16)
hexachlorobutene (C4H2C16)
pentachlorobenzene (C6HC15)
Total
221
(2)
15,339+1,395
675+128
276+12
1,639+52
822+26
34+2
19+1
93+5
30+2
226+8
92+5
1<
38+2
30+0
65+4
7+0
1<
19,386
22 II
(2)
13,709+1,722
854+54
257+3
1,145+54
829+35
37+1
18+1
109+3
39+5
288+5
97+3
1<
33+1
28+1
69+3
11+0
1<
17,523
60
(20)
14,474+1,450
683+45
209+9
984+41
863+17
1<
19+0
110+0
52+2
261+1
102+1
1<
35+0
24+1
57+0
12+0
1<
17,883
100
(95)
13,568+1,442
95+25
205+5
281+34
365+18
1<
21+1
27+1
24+1
81+1
44+1
1<
42+2
6+0
18+1
1<
1<
14,777
Values reported as ppmv in the quartz tube (mL gas phase compoundn-30 minx4.3 mirH-500 mL)
* Amount determined using calibration solutions. Amount of other compounds was estimated.
-------
chromatogram library. The concentrations of chloroform,
CC14, PCE, hexachloroethane, hexachlorobutadiene, and
hexachlorobenzene were determined using calibration
solutions prepared from ACS grade, high purity
reagents (Sigma-Aldrich, Inc., Milwaukee, WI). The
concentrations of other compounds were estimated using
a response factor of 3 as discussed above.
Two experiments were completed with the inlet at 22°C
(221 and 2211) (Relative Humidity of 0.02%) with air as
the carrier gas and the quartz tube operated at 420°C.
The 22 II experiment was completed with the addition
of an aniline trap located in-line after the effluent DCM
trap to determine the amount of phosgene exiting the
quartz tube. Although the two experiments (221 and
2211) could be considered replicates, the addition of
the aniline trap resulted in a pressure increase within
the quartz tube from 1.058+0.001 to 1.072+0.013 bar.
Based on the results shown in Table 3.9, a gas sample
collected from the quartz tube at 420°C with air as the
carrier gas would primarily contain TCE (>80% by
volume). The gas sample would also contain significant
amounts of CC14 (2 to 8% by volume), PCE (2.5 to 5%
by volume), hexachloroethane (0.5 to 2% by volume),
and chloroform (-1.5% by volume).
Increasing the water content (Relative Humidity from
0.02 to 0.95) in the quartz tube reactors operated at
420°C with air as the carrier gas did not affect the
amount of TCE degraded (Table 3.9, R2<0.5 and
P-Value>0.5), in contrast to the results obtained with
nitrogen as the carrier gas (Table 3.8). However,
there was a decrease in some of the chlorinated TCE
degradation products with increasing water content
(Table 3.9), most notably CC14 and hexachloroethane.
Non-chlorinated TCE degradation products such as
ethane (C2H6) were not detected in the DCM trap,
which were anticipated due to the decrease in C1:H ratio
with the addition of water. Thus, increasing the water
content of the quartz reactor did not result in a shift from
chlorinated to non-chlorinated TCE degradation products
as originally hypothesized. Water did, however, affect
the amounts of other TCE degradation products formed
as discussed in the following sections.
3.3.2.3 Compounds Detected in Tedlar®Bags
The entire volume of carrier gas that passed through the
experimental apparatus during each isothermal run was
collected in Tedlar® bags to determine TCE degradation
products that were not retained within the DCM trap.
The anticipated degradation products included CO, CO2,
and phosgene (COC12) based on prior experimental
results as described in Section 2.3.5. The amount of CO
and CO2 formed was determined by GC/TCD analysis
of a 60 mL gas sample from each 1.6 L Tedlar® bag.
Carbon monoxide (CO) and CO2 were detected only
when passing TCE through the quartz tube heated to
420°C with air as the carrier gas (Table 3.10).
A 250 uL gas sample from the Teldar® bag was collected
during the 420°C experiment with inlet temperature
of 22°C (i.e., 22 I) and was analyzed by GC/MS. The
presence of phosgene (COC12) was identified by mass
spectrum match with the NIST98 library. An aniline
trap was added to determine the amount of phosgene
formed as per EPA method TO-6 (U.S. EPA, 1999) and
the experiment was repeated (i.e., 22 II). The amount of
phosgene formed was determined gravimetrically and
by determining the concentration of carbanilide formed.
The concentration of phosgene reported in Table 3.10
was calculated using the ideal gas law from the moles of
phosgene detected in the 1.6 L volume Tedlar® bag at
25°C and 1 atm.
The aniline trap was used for experiments completed at
420°C with air as the carrier gas to evaluate the effect
of increasing water content on the amount of phosgene
produced. Phosgene concentrations were found to
decrease with increasing quartz-tube water content
Table 3.10 Concentration of Compounds Detected in the Tedlar® Bag and Phosgene Trap for the 420°C
Experiments with Air as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
22 I (2)
22 II (2)
60 (20)
100 (95)
Correlation Coefficient
R2
CO (uL/L)
(%ofTCE-C)
8640 (15.4)
8712(15.3)
9410 (16.3)
8846 (15.5)
0.16
(P-Value=0.61)
CO2(uL/L)
(%ofTCE-C)
2120 (3.8)
1755(3.1)
2285 (4.0)
4395 (7.7)
0.85
(P-Value=0.08)
Phosgene
(%ofT
gravimetric
NA
7964 (7.0)
929 (0.8)
15 (0.3)
0.84
(P-Value=0.26)
(uL/L)
:E-O
UV254
NA
1067 (0.9)
836 (0.7)
345 (0.3)
0.96
(P-Value=0.13)
-------
(Table 3.10). However, the compound formed after
passing the quartz tube effluent through the aniline trap
may not have been due to phosgene alone. For example,
O'Mara et al. (1971) found that gas phase HC1 formed
during the combustion of vinyl chloride caused aniline
to polymerize in a liquid trap and form a compound
that had a UV absorbance of 254 nm which interfered
with the detection of phosgene. While the concentration
of HC1 in the quartz tube effluent was not determined,
the amount of chloride found in the water rinse (see
Section 3.3.2.4 and Figure 3.5) suggests that gas phase
HC1 was present in the quartz tube effluent. Thus, the
decrease in phosgene concentration with increase in
quartz tube water content shown in Table 3.10 may have
been due to phosgene hydrolysis alone, or may represent
a reduction in effluent HC1 concentration along with
phosgene hydrolysis. Hydrolysis of phosgene can occur
in the gas- and aqueous phases, and is reported to yield
CO according to (Ryan et al., 1996):
COCL + HO -> CO. + 2HC1
(3.2)
An increase in the amount of CO2 produced with
increasing in quartz tube water content was apparent.
In contrast, the concentration of CO in the effluent
(8,902+349 ppmv) and TCE recovery remained
consistent (Table 3.7), implying a shift in degradation
product distribution with phosgene being converted to
CO2 as expected based on Equation 3.2. The amounts
(moles) of CO2 and phosgene formed along with the
difference between the amount of CO2 found with the
inlet at 22°C (22 II), at 60°C, and at 100°C are shown in
Table 3.11. The increase in CO2 production with increase
in quartz tube water content (i.e., CO2 Gain, Table 3.11)
was approximately 33% of the amount of phosgene
lost between the inlet temperatures of 22°C and 100°C
based on the phosgene gravimetric analysis, but was
6 times greater than the amount of phosgene lost based
on the UV 254 analysis. The gravimetric results suggest
that the increase in CO2 was primarily due to phosgene
hydrolysis. However, the UV 254 analysis results
suggest that not all the solids formed in the aniline trap
represented carbanilide.
3.3.2.4 Compounds Detected in the Water Rinse
After each isothermal run, approximately 30 mL of
freshly dispensed DI water were used to rinse the quartz
tubes once they had cooled to room temperature (22°C).
The water rinse was performed to determine the water-
soluble TCE degradation products formed after passing
TCE through the heated quartz tube. The anticipated
degradation products included chloride, due to the loss
of chlorine atoms from TCE (i.e., dechlorination), and
haloacetic acids such as dichlororacetate, based on the
past experimental work described in Section 2.3.2 (e.g.,
McKinney et al., 1995).
The amounts of chloride in the water rinse as a function
of quartz tube temperature, carrier gas, and inlet
temperature (water content) is shown in Figure 3.5.
Chloride was detected in the water rinse from each
isothermal experiment regardless of carrier gas used.
This result suggests that TCE was degraded, to some
extent, in all of the quartz tube experiments performed
in the fifth experimental series. The amount of chloride
in the 120°C experiment with the inlet temperature at
22°C and nitrogen as the carrier gas was 0.06 umol,
which was below the method detection limit (MDL) of
0.07 umol, and the concentration of chloride with air as
the carrier gas was 0.10 mg/L with a MDL of 0.05 mg/L;
all other chloride concentrations were at least an order-
of-magnitude above their MDL. The amount of chloride
increased with increasing quartz tube water content (i.e.,
inlet temperature), even for experiments completed at
the lowest quartz tube temperature of 120°C where TCE
recovery was greater than 94% (Figure 3.4), and no TCE
degradation products were detected in the DCM trap.
The amount of chloride measured in the 240 and 420°C
experiments with air as the carrier gas was greater than
the amount of chloride detected in experiments with
nitrogen as the carrier gas, which was consistent with
the lower TCE recovery noted in Figure 3.4 with air as
the carrier gas. The amount of chloride was greatest in
the water rinse after the 420°C experiments where TCE
degradation products were detected in the DCM trap.
Table 3.11 Change in the Amount of CO2 and Phosgene Detected with Increase in Water Content for the 420°C
Experiments with Air as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
22 I (2)
22 II (2)
60 (20)
100 (95)
CO2(mmol)
0.28
0.23
0.30
0.58
CO2 Gain
CO2 - CO2 (2211)
-0.05
0.00
0.07
0.35
Phosgene
gravimetrically
NA
0.52
0.06
0.00
; (mmol)
UV254
NA
0.07
0.05
0.02
NA - not analyzed
-------
1000
100
E"x7l 100°C Inlet
F^73 60°C Inlet
^^ 22°C Inlet
0.1 -
Carrier Gas:
Quartz Tube:
Figure 3.5
Nitrogen Air
120°C
Nitrogen Air
240°C
Nitrogen Air
420°C
Amount of chloride detected in the post
experiment water rinse.
The amount of dichloroacetate (C12HC2OO") as a
function of quartz tube temperature, carrier gas, and inlet
temperature (water content) is shown in Figure 3.6. As
mentioned in Section 2.3.2, DCAA has been classified
as a probable human carcinogen (US EPA, 1998). No
haloacetic acids (e.g., dichloroacetate) were detected
in the water used to rinse the quartz tube operated at
120°C with nitrogen as the carrier gas and the minimum
concentration of DCAA detected with air as the carrier
gas was 0.006 mg/L which was near the MDL of
0.005 mg/L. The amount of dichloroacetate (DCAA)
detected in the water rinse with air as the carrier gas
was greater than the amount of DCAA detected in
experiments completed with nitrogen as the carrier gas.
100°C Inlet
60°C Inlet
I 20°C Inlet
Carrier Gas:
Quartz Tube:
Nitrogen Air
120°C
Nitrogen Air
240°C
Nitrogen Air
420°C
Figure 3.6 Amount of dichloroacetate detected in the
post experiment water rinse.
Trichloroacetate (C13C2OO") was also detected in the
water rinse from the quartz tube operated at 240 and
420°C with air as the carrier gas (Table 3.12), whereas
no trichloroacetate (TCAA) was detected in experiments
with nitrogen as the carrier gas. TCAA was identified
by mass spectrum match with the NIST02 library after
GC/MSD analysis and the concentration of TCAA
was estimated based on the ratio of chromatogram
peak areas between DCAA and TCAA along with the
concentration of DCAA that was determined using
calibration solutions. The amounts of TCAA and DCAA
were similar for the 240°C experiments, whereas the
amount of TCAA exceeded that of DCAA for the 420°C
experiments with inlet temperature of 22 and 100°C. The
water rinse from each 420°C experiment with air as the
carrier gas had a pale yellow color and a strong solvent
odor, whereas the water rinse with nitrogen as the carrier
gas was clear.
After processing the water rinse solutions for haloacetic
acid analysis, the MTBE extract was analyzed by
GC/MSD, which revealed the presence of additional
chlorinated compounds (Table 3.13). Each compound
was identified by mass spectrum match with the NIST02
library and the mass of each compound was estimated
based on the ratio of chromatogram peak area to the
peak area for DCAA. Based on the results shown in
Figure 3.5 and Table 3.12, a water sample collected
from the quartz tube experiment operated at 240°C and
420°C with nitrogen as the carrier gas would contain
chloride and dichloroacetric acid (DCAA). With air as
the carrier gas, a water sample collected from the quartz
tube operated at 240°C would contain DCAA and TCAA
(Table 3.12) and a water sample from the quartz tube
at 420°C would contain the chlorinated hydrocarbons
3,4-dichloro-3 -butene-2-one, pentachlorobutadiene,
hexachlorobutene, and pentachlorobenzene (Table 3.13),
in addition to DCAA and TCAA.
3.3.2.5 Compounds Detected in the Iso-Octane Rinse
The DI water rinse was immediately followed by a
30 mL iso-octane rinse for period of 5 minutes. The
iso-octane rinse was performed to determine organic
TCE degradation products that had condensed within the
experimental apparatus while passing TCE through the
heated quartz tube. The anticipated degradation products
included hexachlorobutadiene and hexachlorobenzene
based on the past experimental work described in
Section 2.3.5 (e.g., Froese and Hutziner, 1994). The iso-
octane rinse samples were initially analyzed by GC/MS
to identify the TCE degradation products present, while
the mass of each product in each iso-octane rinse was
determined by GC/FID analysis. No TCE degradation
products were detected in the iso-octane rinse for the
120 and 240°C experiments regardless of carrier gas or
water content.
Up to three products were detected in the 420°C
experiment with nitrogen as the carrier gas (Table 3.14).
Increasing the quartz tube water content (i.e., inlet
temperature) led to a decrease in the number and amount
of degradation products detected with nitrogen as the
carrier gas, similar to the increase in TCE recovery noted
-------
Table 3. 12 Amount of Haloacetic Acids in the Water Rinse from 240°C and 420°C Experiments with Air as the
Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
20(2)
60 (20)
100 (95)
Quartz Tube Temperature (°C) - Air as Carrier Gas
240
DCAA (umol)
0.53±0.00
1.48±0.01
0.78±0.01
TCAA (umol)*
0.30
0.66
0.30
420
DCAA (umol)
0.53+0.00
8.12+0.02
1.08+0.04
TCAA (umol)*
1.92
6.02
3.42
*TCAA concentration estimated based on ratio of TCAA to DCAA chromatogram peak area.
Table 3.13 Estimated Amounts (umol) of Compounds in Water Rinse from 420°C Experiment with Air as the
Carrier Gas
Compound
tetachloroethylene (C2C14)
1,1,2,2-tetrachloroethane (C2C14H2)
hexachloroethane (C2C16)
3 ,4-dichloro-3 -butene-2-one (C4C12OH4)
pentachlorobutadiene (C4HC15)
hexachlorobutadiene (C4C16)
pentachloro-1-propene (C3C15)
hexachlorobutene (C4H2C16)
pentachlorobenzene (C6HC15)
hexachlorobenzene (C6C16)
tetrachloro- 1 , 3 -cy clopentadiene-5 -
dichloromethylene (C6C16)
Inlet Temperature (°C)
(Relative Humidity, %)
22 I (2)
nd
nd
nd
0.36
0.09
nd
0.46
nd
0.08
nd
nd
22 II (2)
nd
nd
nd
0.11
0.11
nd
nd
0.54
0.11
0.02
0.16
60 (20)
0.25
nd
0.07
1.40
0.08
0.03
0.14
0.76
0.23
0.08
0.04
100 (95)
0.49
0.09
nd
11.63
0.08
nd
nd
0.10
0.14
0.01
nd
nd - not detected in the chromatogram
Table 3.14 Amounts (umol) of Compounds in the Iso-Octane Rinse from the 420°C
Experiments with Nitrogen as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
hexachlorobutadiene (C4C16)
hexachlorobutene (C4H2C16)
pentachlorobenzene (C6HC15)
heptachlorocyclohexane (C6H5C17)
hexachlorobenzene (C6C16)
22
(2)
O.OK
O.OK
0.75+0.02
0.55+0.01
5.03+0.02
60
(20)
O.OK
O.OK
O.OK
O.OK
1.58+0.02
100
(95)
O.OK
O.OK
O.OK
O.OK
O.OK
-------
in Table 3.6 (R2=0.994). There were no degradation
products detected by GC/FID analysis in the iso-octane
rinse of the 420°C experiment with the inlet temperature
at 100°C which represented the maximum water content
(relative humidity of 95%) for the fifth experimental
series.
There were up to five compounds detected in the 420°C
experiments with air as the carrier gas (Table 3.15).
Hexachlorobutadiene and hexachlorobutene were
detected in addition to penta- and hexachlorobenzne in
the iso-octane rinse for the inlet temperature of 22°C.
Penta- and hexachlorobenzene were detected in the iso-
octane rinse with the inlet at 100°C. These compounds
were also detected in the MTBE extract of the water
rinse (Table 3.13) that was completed prior to the iso-
octane rinse for the 420°C experiment with air as the
carrier gas. Based on the results presented in Tables
3.14 and 3.15, hexachlorobenzene is likely to be the
most prominent degradation product formed under these
experimental conditions, and increasing the amount of
water vapor entering the reactor would decrease the
amount of TCE degradation products formed.
Table 3.15 Amounts (umol) of Compounds in
the Iso-Octane Rinse from the 420°C
Experiments with Air as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
hexachlorobutadiene
hexachlorobutene
(C4H2C16)
pentachlorobenzene
(C6HC15)
heptachlorocyclohexane
(C6H5C17)
hexachlorobenzene
221
(2)
0.66
1.55
0.95
O.OK
4.29
22 II
(2)
0.47
1.26
0.75
O.OK
4.23
60
(20)
NA
NA
NA
NA
NA
100
(95)
O.OK
O.OK
0.65
O.OK
1.56
NA - not analyzed, sample broken during storage
3.3.2.6 Mass Balance
The amount of TCE as moles of carbon (moles carbon =
2 x moles TCE) detected in the DCM trap with respect
to the moles of TCE as carbon that were injected into the
experimental apparatus operated at 420°C with nitrogen
as the carrier gas is provided in Table 3.16 as "%Carbon
in Feed." Similarly, the amount of TCE recovered as
moles of chlorine (moles chlorine = 3 x moles TCE)
with respect to the moles of TCE as chlorine that were
injected into the apparatus is also provided in Table
3.16 as "%Chlorine in Feed." These measures of TCE
recovery are analogous to those presented for TCE in
Section 3.3.2.1, and inFigure 3.4 and Table 3.6.
The amount of TCE degradation products detected in
the DCM trap and quartz tube rinses are also reported in
Table 3.16 as moles of carbon and chlorine with respect
to the amount of carbon and chlorine delivered to the
quartz tube apparatus as TCE. The purpose of reporting
the amount of carbon or chlorine detected is to show
the distribution of each TCE degradation product in the
apparatus and to determine if all the carbon and chlorine
atoms were accounted for (i.e., mass balance). For
example, while 93.1% of the carbon delivered during
the experiment with the inlet at 22°C was detected
in the DCM trap as TCE (Table 3.16), 3.4% of the
carbon delivered was detected in the DCM trap as TCE
degradation products that were presented in Table 3.8,
and 0.5% of the carbon delivered was detected in the
quartz tube rinses as TCE degradation products as listed
in Table 3.14. Thus, the net recovery of the TCE injected
with the inlet at 22°C increased from 93.1% when TCE
recovery alone was considered to 97.1% on a carbon
basis, and increased to 101.2% on a chlorine basis when
the TCE degradation products were included.
TCE was the predominant (>93%) compound detected
in the DCM trap, and there were more chlorinated TCE
degradation products condensed within the DCM trap
than found in the water or iso-octane rinses of the 420°C
experiments with nitrogen as the carrier gas. Increasing
the quartz-tube water content resulted in a decrease in
the amount of TCE degradation products in the DCM
trap while the amount of chloride found in the water
rinse increased. Overall, very good mass recovery of
TCE on a carbon (>97%) and chlorine (>100%) basis
was obtained in the 420°C experiment with nitrogen as
the carrier gas.
The distribution of carbon and chlorine in the
experimental apparatus operated at 420°C with air as
the carrier gas is provided in Table 3.17. Approximately
18-23% of the carbon introduced as TCE was converted
to CO and CO2, while the other half consisted of
chlorinated hydrocarbons detected in the DCM trap and
as phosgene for the experiment with inlet temperature of
22°C. Increasing the quartz tube water content resulted
in a decrease in the amount of chlorinated carbon
compounds in the DCM trap and in the water and iso-
octane rinses of the quartz tube, along with an increase
in the amount of chloride found in the water rinse.
The amount of carbon and chlorine recovered in the
DCM trap, water and iso-octane rinses, and in the
Tedlar® bag decreased with increasing quartz tube
water content for the 420°C experiments with air as the
carrier gas (Table 3.17). The amounts of missing carbon
and chlorine were nearly equal, suggesting that the
missing compounds could have consisted of chlorinated
-------
Table 3.16 Distribution of Carbon and Chlorine for the 420°C Experiments with Nitrogen as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
Phase
Compounds in
DCM Trap
Compounds in
Gas Phase
Condensed in
Quartz Tube
Compound
TCE
All Other
C0/C02
Water + Iso-
Octane Rinses
Net Recovery
22
(2)
60
(20)
100
(95)
%Carbon in Feed
93.1+1.9
3.4+0.1
<1.4
0.5+0.0
97.1+1.9
95.5+1.5
2.4+0.0
<1.4
0.1+0.0
98.1+1.5
99.1+1.9
1.3+0.0
<1.4
0.0+0.0
100.4+1.9
22
(2)
60
(20)
100
(95)
%Chlorine in Feed
93.1+1.9
7.3+0.2
<1.4
0.8+0.1
101.2+1.9
95.5+1.5
4.7+0.2
<1.4
0.9+0.1
101.1+1.6
99.1+1.9
1.1+0.0
<1.4
2.2+0.3
102.4+1.9
Table 3.17 Distribution of Carbon and Chlorine for the 420°C Experiments with Air as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
Phase
Compounds in
DCM Trap
Compounds in
Gas Phase
Condensed in
Quartz Tube
Compound
TCE
All Other
CO/CO2
Phosgene
Water + Iso-
Octane Rinses
Net Recovery
22 II
(2)
60
(20)
100
(95)
% Carbon in Feed
52.4+1.2
13.2+0.3
18.4
7.0
0.6+0.0
91.5+1.2
54.6+0.2
11.9+0.2
20.3
0.8
0.6+0.0
88.1+0.3
51.7+0.9
3.2+0.1
23.2
0.0
1.0+0.0
79.1+0.9
22 II
(2)
60
(20)
100
(95)
% Chlorine in Feed
52.4+1.2
20.3+0.4
0.0
14.0
2.0+0.0
88.8+1.3
54.6+0.2
18.0+0.3
0.0
1.1
6.7+0.4
80.5+0.5
51.7+0.9
4.9+0.1
0.0
0.0
17.8+1.8
74.6+2.0
hydrocarbons. It is possible that the unaccounted for
chlorinated hydrocarbons were present in the Tedlar®
bag and went undetected during the GC/TCD analysis
for CO/CO2 content due to adsorption on the Carboxen
1010 capillary column.
3.4 Discussion
The goal of the quartz tube experiments was to
determine the identity and amount of TCE degradation
products formed after exposing gas phase TCE to
temperatures from 60 to 480°C. The following sections
present potential TCE degradation mechanisms based on
the quartz tube experimental results.
3.4.1 Nitrogen as the Carrier Gas at 420°C
The 420°C experiment with nitrogen as the carrier gas
involved passing TCE and water vapor through the
quartz tube heated to 420°C. Nitrogen and water are inert
relative to oxygen and are not thought to have caused
TCE degradation under the experimental conditions.
Therefore, the degradation of TCE, as indicated by the
detection of carbon compounds and chloride ions in the
DCM trap and quartz tube rinses is hypothesized to have
occurred via unimolecular dissociation of TCE rather
than a bimolecular reaction with nitrogen or water.
The compounds formed after passing gas-phase
TCE at 320°C through a laser beam include HC1,
dichlorovinylidene (:C2C12), dichloroacetylene (C2C12),
vinyl radicals (HC1C2C1'), and chlorine atoms (Cl)
according to (Yokoyama et al., 1995):
-------
Cl Cl
\ _ / T>420°C
/ \ Gas Phase
H Cl
H—Cl
H—Cl
Cl
dichlorovinylidene
- Cl—C=C—Cl
dichloroacetylene
H
+ /c=(v
cr ci
dichlorovinyl radical
(a)
(b)
(c)
(3.3)
While Yokoyama et al. (1995) found spectroscopic
evidence to suggest that dichlorovinyl radicals (3.3c)
had formed after passing TCE through the laser beam,
the branching ratio for the Cl (3.3c) to HC1 elimination
(3.3a and 3.3b) reactions was 0.17, which indicates that
the HC1 elimination pathway produced approximated
5 times more TCE degradation products than the Cl
elimination reaction pathway under these experimental
conditions. The predominance of the HC1 elimination
pathway (3. 3 a and 3.3b) is supported by the results
of Reiser et al. (1979) which indicate that HC1 and
dichloroacetylene were the primary products from the
photolysis of TCE at 25°C.
The elimination of HC1 from TCE (Equations 3. 3 a and
3.3b) was proposed to occur at elevated temperatures
after passing TCE through a flame (Chang and Senkan,
1989) and after passing gas-phase TCE through a heated
quartz tube (Wu and Lin, 2004) based on the detection
of dichloroacetylene. Passing TCE through the quartz
tube apparatus used herein at 420°C with nitrogen as the
carrier gas is proposed to have caused the unimolecular
dissociation of TCE described by Equation 3.3. The
products in Equation 3.3 are reactive compounds
thought to have rapidly transformed into the chlorinated,
4 and 6 carbon compounds that were detected in
experiments completed herein using the empty quartz
tube with nitrogen as the carrier gas (Tables 3.8 and
3.14). Goodall and Hewlett (1954) also found HC1 and
hexachlorobenzene as the primary TCE degradation
products after passing gas phase TCE through a Pyrex
tube heated between 385 and 445°C with nitrogen as the
carrier gas.
Increasing the water content of the quartz tube at 420°C
with nitrogen as the carrier gas resulted in an increase
in TCE recovery, an increase in the amount of chloride
detected in the water rinse, and a decrease in the amount
of degradation products detected in the DCM trap and in
the iso-octane rinse (Table 3.18). These results indicated
that increasing the amount of water in the quartz tube
at 420°C resulted in a decrease in the amount of TCE
degraded.
The role that water played in reducing TCE degradation
is not known. Water may have reacted with the
unimolecular degradation products shown in Equation
3.3 and prevented them from reacting with TCE, thereby
decreasing the amount of TCE degraded. For example,
chlorine initiated TCE degradation has been shown to
occur at 25°C and is thought to proceed by forming
tetrachloroethyl radicals (CL,HC2C12') according to
(Catoire et al., 1997):
Cl Cl
/C=C
H Cl
cr
T=25°C
1
Gas Phase
C'\ fl
CH—C'
C/ }»
(3.4)
Increasing the quartz-tube water content in the
experiments completed at 420°C with nitrogen as the
carrier gas could have prevented the chlorine radicals
produced by the unimolecular dissociation of TCE
(Equation 3.3c) from reacting with TCE. However,
chlorine radicals are not thought to react with water
based on the equilibrium coefficient for the aqueous
phase reaction as shown in Equation 3.5 which is
1.4xlO-7 at 24°C (Yu et al., 2004).
l- + HO -> C1HO- + H+
(3.5)
Table 3.18 Carbon in the DCM Trap and Iso-Octane Rinse, and Chloride in the Water Rinse for the 420°C
Experiments with Nitrogen as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
TCE Recovery (%)
Carbon in DCM trap other than TCE (umol)
Carbon in iso-octane rinse (umol)
Carbon in water rinse (umol)
Chloride in water rinse (umol)
22
(2)
93.1±1.9
251.6
38.0
0.01
52.0
60
(20)
95.5±1.5
176.7
9.4
0.01
83.9
100
(95)
99.1±1.9
99.7
0.0
0.84
242.7
Note: No CO or CO2 was detected in the experiments with nitrogen as the carrier gas.
-------
While chlorine radicals are not expected to react with
water, they are known to react with chloride-ions
dissolved in water to form dichlorine radical anions
(CL;-) (Yu and Barker, 2003).
cr + ci- -> ci -
(3.6)
Since the amount of chloride increased with quartz tube
water content, water may have reduced the amount
of gas-phase chlorine radicals and thus reduced the
amount of TCE degraded. The source of the chloride
ions is thought to be from the HC1 produced during
the unimolecular dissociation of TCE according to
Equations 3.3a and 3.3b since HC1 readily ionizes in
water. For example, the presence of water molecules in
a gas stream with HC1 molecules at 25°C was shown
to cause an increase in the hydrogen to chlorine bond
length at the water to HC1 molar ratio of 1:1 and the
complete ionization of HC1 with the water to HC1 molar
ratio at 5:1 (Farnik et al., 2003).
HC1 has been proposed as a source of chlorine radicals
in the post-flame zone of combustion chambers
(Procaccini et al., 2003). However, the experimental
work completed to date suggests that HC1 is stable at
420°C, with an estimated dissociation half-life of 3 x 109
years (Baulch et al., 1981). Thus, HC1 is not expected
to yield chlorine radicals at 420°C with nitrogen as
the carrier gas. Even though HC1 is not expected to
dissociate to yield chlorine radicals at 420°C there is
experimental evidence to suggest that HC1 reacts with
organic compounds at elevated temperatures. The
presence of HC1 caused the chlorination of gas phase
hexachlorodibenzo-p-dioxin (HCDD) at 248°C with the
formation of hepta- and octachlorodibenzo-p-dioxins
whereas less chlorinated dibenzo-p-dioxins were found
in experiments completed without HC1 (Addink et
al., 1996). Procaccini et al. (2003) found that adding
gas-phase HC1 and benzene to a post ethene (C2H4)
combustion zone at 640°C resulted in the formation of
chlorobenzene and chlorophenols, demonstrating that
HC1 could react with benzene. The reaction between
HC1 and benzene was proposed to involve chlorine
radicals based on the similarity in product distribution
after using C12 gas and HC1 in combination with benzene
(Procaccini et al., 2003). Thus, HC1 may be reacting
with TCE in the experiments completed herein at 420°C
with nitrogen as the carrier gas. Increasing the water
content of the 420°C experiment with nitrogen as the
carrier gas is proposed to have decreased the amount of
HC1 and chlorine radicals available to react with TCE.
An alternative explanation is that the increase in water
content could cause an increase in the HC1 elimination
pathways (Equations 3.3a and 3.3b) and a decrease in the
chlorine radical pathway (Equation 3.3c).
3.4.2 Air as the Carrier Gas at 420°C
The 420°C experiment conducted with air (21% oxygen)
as the carrier gas involved passing TCE and water vapor
through the quartz tube heated to 420°C. The presence
of molecular oxygen in the 420°C experiments resulted
in a decrease in TCE recovery and an increase in the
number and amount of TCE degradation products
detected as compared to the experiments completed
with nitrogen as the carrier gas. Thus, the presence of
oxygen in the 420°C experiments resulted in an increase
in the amount of TCE degraded in excess of the amount
of TCE degraded by unimolecular dissociation alone
(Equation 3.3). The increase in TCE degradation with
oxygen present is thought to involve the formation of
peroxyl radical species. Molecular oxygen is suspected
to have reacted with tetrachloroethyl radicals produced
from the reaction between chlorine and TCE as shown in
Equation 3.4 to form peroxyl radicals (CL,HC2C12OO )
according to (Catoire et al., 1997; Nimlos et al., 1993):
c, a
H—C— C—O-
I I
CI CI
(3.7)
•o-
The peroxyl radicals are suspected to react with TCE in
a radical chain mechanism resulting in an increase in the
amount of TCE degraded at 420°C as compared to the
amount degraded with nitrogen as the carrier gas.
The TCE degradation products detected in the DCM trap
(Table 3.9), water rinse (Table 3.13), and iso-octane rinse
(Table 3.15) are thought to have formed by reactions
involving radicals such as the tetrachloroethyl radical,
by reactions involving non-radical compounds such
as dichlorocarbene, or a combination of radical and
non-radical interactions. For example, trichloroacetate
(TCAA) was detected in the DCM trap and in the water
rinse (Table 3.12) of the quartz tube after the 420°C
experiments with air as the carrier gas. The formation
of 1,1,1,2-tetrachloroethyl radicals (Cl3C2HCf) from
the chlorine initiated TCE degradation (Equation 3.8)
is suspected as the key intermediate that reacted with
molecular oxygen to yield TCAA.
/c
(3.8)
The formation of tetrachloroethyl radical isomers
may have been due to the additional chlorine radicals
produced during the peroxyl induced degradation of
TCE. TCAA is known to yield dichlorocarbene (:CC12)
and CO2 upon heating (Kaberdin and Potkin, 1994;
p. 250, Smith and March, 2001). Dichlorocarbene is a
reactive compound that is known to combine with TCE
to yield pentachlorocyclopropane (Sepiol and Soulen,
-------
1975), which was detected in the DCM trap (Table 3.9).
Dichlorocarbene has been suggested to dimerize to form
PCE, to react with chlorine to form CC14, and with HC1
to form chloroform (CHC13) (Zhu and Bozzelli, 2003),
all compounds that were detected in the DCM trap
(Table 3.9).
Dichloroacetylene (DCA) is another intermediate
compound that may have contributed to the formation
of the TCE degradation products observed in the DCM
trap and rinses. Reichert et al. (1980) synthesized DCA
(C2C12) from TCE and then exposed DCA to air at room
temperature (22° C). DCA decomposed on contact with
air and formed phosgene, PCE, hexachlorobutadiene,
trichloroacryloyl chloride, trichloroacetyl chloride, CC14,
and chloroform. Several of these products including
PCE, hexachlorobutadiene, CC14, and chloroform were
also detected in the 420°C experiment with air as the
carrier gas, suggesting that DCA may have been present.
Increasing the amount of water entering the quartz tube
by increasing the inlet temperature from 22°C to 60°C
resulted in a decrease in TCE recovery for the 221 to
60 experiments, whereas there was an increase in TCE
recovery for the 2211 to 60 experiments (Table 3.19).
The lowest TCE recovery occurred when the inlet was
operated at 100°C (relative humidity of 95%). While
there was no clear trend in TCE recovery with increasing
water content (R2=0.052, see Table 3.7), increasing the
quartz tube water content led to a decrease in the amount
of degradation products in the DCM trap and iso-octane
rinse, with an increase in the products detected in the
gas phase and an increase in the amount of chloride
in the water rinse (Table 3.19). The increase in CO2
coupled with the decrease in phosgene may have been
related to the increase in quartz tube water content as
described in Section 3.3.2.4. However, the amount of
chloride detected in the water rinse for the 100°C inlet
experiment (1954 umol) was approximately twice the
amount expected if all the phosgene had reacted with
water (2x521=1042 umol). The amount of chloride
detected above the amount expected from phosgene
hydrolysis (1954 - 1042 = 912 umol) might represent
chlorine that was prevented from reacting with TCE due
to the presence of water in the quartz tube. Thus, water
may have hydrolyzed phosgene and served to remove
reactive chlorine radicals and HC1 from the gas phase
while not impacting the amount of TCE degraded.
3.4.3 Experiments Conducted at 120 and 240°C
The small amount of chloride (<0.02% of TCE in
the feed) detected in the water rinse from the 120°C
experiments completed with nitrogen and air as the
carrier gas (Table 3.20) was initially thought to represent
background chloride from the laboratory air. However,
the detection of DCAA in water rinses from the 120°C
experiment with air as the carrier gas indicated that
TCE degradation was occurring. Table 3.20 contains
the amount of chloride (umol) and DCAA detected
(nmol) in the water rinse from the 120°C experiments.
The detection limit for chloride was 0.07 umol, and
was determined using the standard deviation of 12
measurements of a 2 uM calibration standard collected
over a one month period and the student's t value of
2.718 (n=ll, alpha=0.01). The detection limit for the
DCAA was determined using the standard deviation
of 12 measurements of a 12 ug/L calibration standard
analyzed over a 15 day period.
Table 3.19 Carbon in the DCM Trap and Iso-Octane Rinse, and Chloride in the Water Rinse for the 420°C
Experiments with Air as the Carrier Gas
Inlet Temperature (°C)
(Relative Humidity, %)
TCE Recovery (%)
Carbon in DCM trap other than TCE (umol)
Carbon in iso-octane rinse (umol)
Carbon as CO (umol)
Carbon as CO2 (umol)
Carbon as Phosgene (umol) gravimetrically
Carbon as DCAA (umol)
Carbon as TCAA (umol)
Total Carbon other than TCE (umol)
Chloride in water rinse (umol)
221
(2)
59.5+0.9
967.4
40.3
1130.7
277.5
NA
0.6
0.7
2417.2
104.7
22 II
(2)
52.4+1.2
981.0
36.8
1140.2
228.4
521.1
0.5
1.9
2909.9
105.1
60
(20)
54.6+0.2
897.4
NA
1231.6
299.1
60.8
8.1
6.0
2503.0
725.1
100
(95)
51.7+0.9
240.8
13.3
1157.8
575.1
1.0
1.1
3.4
1992.5
1953.9
-------
Table 3.20 Amount of Chloride and DCAA in the
Water Rinse from the 120°C Experiments
Inlet
Temperature
(°C)
(Relative
Humidity, %)
22(2)
60 (20)
100 (95)
Nitrogen as
Carrier Gas
ci-
(umol)
0.07
0.74
0.94
DCAA
(nmol)
0.5
O.5
0.8
Air as Carrier Gas
ci-
(umol)
0.11
0.45
0.48
DCAA
(nmol)
2.70+0.03
1.89+0.06
2.08+0.05
No estimate of uncertainty, only one Q- measurement
performed.
If the chloride detected in the 120°C experiments
represents the degradation of TCE, then carbon
degradation products in addition to DCAA should
have been detected. Dichloroacetylene (C2C12) is the
expected product after elimination of HC1 from TCE
(Equations 3.3a and 3.3b). With nitrogen as the carrier
gas, dichloroacetylene should have been collected in
the Tedlar® bag; however, the bag was only analyzed
for CO and CO2 content and none was detected. With
air as the carrier gas, dichloroacetylene was expected
to react with oxygen to form CO and CO2 and, based
on the amount of chloride detected, there should have
been from 2 to 30 ppmv of CO2 in the Tedlar® bag. The
CO2 content of the Tedlar® bag for the experiment at
120°C with air as the carrier gas was determined using
a GC/TCD method; however, the detection limit for this
method was 500 ppmv, and no CO or CO2 was detected.
Increasing the tube temperature to 240°C did not
significantly increase the amount of chloride detected
with nitrogen as the carrier gas (Table 3.21) as compared
to the amount detected in the 120°C experiments
(Table 3.20). The presence of oxygen in the 240°C
experiments resulted in an increase in the amount
of DCAA, TCAA, and chloride detected relative
to the experiments completed with nitrogen as the
carrier gas (Figures 3.7 and 3.8). This suggests that
chlorine initiated TCE degradation occurred to yield
the 1,1,1,2-tetrachloroethyl radical, which reacted
with oxygen to yield TCAA. No other chlorinated
hydrocarbon compounds were detected indicating that
there was insufficient thermal energy for the radical
chain reaction to propagate.
Increasing the water content of the quartz tube in the
240°C experiment with air as the carrier gas resulted in a
decrease in TCE recovery (R2=0.998 and P-Value=0.03,
see Table 3.7), while no trend in TCE recovery was
apparent for the 240°C experiment completed with
nitrogen as the carrier gas (R2=0.227 and P-Value=0.68,
see Table 3.6). Thus, water did have an effect on the
recovery of TCE when combined with oxygen, which
may have been to induce TCE degradation. However, the
amount of chloride, DCAA, and TCAA detected in the
water rinse did not significantly increase with quartz tube
water content. Additional carbon degradation products
should have been detected to confirm that water was
causing TCE degradation since the amount of DCAA
and TCAA detected in the 240°C experiment with air as
the carrier gas represented less than 1% of the missing
TCE as carbon. The range of CO2 concentrations,
assuming the missing TCE was completely converted
to CO2, would have been from 2170 to 4596 ppmv, well
above the GC/TCD method detection limit; however,
no CO or CO2 was detected. The two explanations for
the decrease in TCE recovery with increasing water
content for the 240°C experiments with air as the carrier
gas are: 1) there were additional chlorinated carbon
reaction products, such as dichloroacetylene, that went
undetected or 2) more TCE partitioned into the water
in the quartz tube with air as the carrier gas than with
nitrogen as the carrier gas.
Table 3.21 Amount of Chloride, DCAA, and TCAA in the Water Rinse from the 240°C Experiments
Inlet Temperature (°C)
22
60
100
Nitrogen as Carrier Gas
Cl- (umol)
0.27
0.36
0.34
DCAA (nmol)
2.9+0.1
2.5+0.1
2.9+0.1
Air as Carrier Gas
Cl- (umol)
11.2+0.9
15.5+2.2
14.9+0.2
DCAA (nmol)
528+4
1481+8
711+13
TCAA (nmol)
299
662
305
-------
3.5 Summary
The average recovery of TCE with nitrogen as the
carrier gas was greater than 94% at all the experimental
temperatures. Carbon-based TCE degradation products
were only detected in the experiments completed at
420°C with nitrogen as the carrier gas. Up to four
degradation products were identified in the DCM trap,
two in the water rinse, three in the iso-octane rinse, and
no CO or CO2 was detected for experiments completed
at 420°C with nitrogen as the carrier gas. The amount
of TCE recovered as carbon for the 420°C experiments
with nitrogen as the carrier gas was greater than 97%,
with 93% as TCE. The amount recovered as chlorine
was greater than 100% with up to 7% as chlorinated
degradation products. The degradation products detected
contained 4 and 6 carbon atoms, with greater than
5 chlorine atoms per molecule. TCE degradation was
proposed to be initiated by thermal induced unimolecular
dissociation but was also influenced by chlorine induced
degradation. Increasing the quartz tube water content
resulted in an increase in TCE recovery which was
suggested to indicate a decrease in chlorine induced TCE
degradation.
The average recovery of TCE was greater than 94%
with air as the carrier gas for the 120 and 240°C
experiments, but dropped to approximately 53% in
the 420°C experiments. The small amount (<0.02%
TCE) of chloride detected in experiments completed
at 120°C and at 240°C was thought to represent TCE
degradation. However, no carbon degradation products
were identified that could account for the missing TCE.
Carbon based TCE degradation products were detected
in the 240 and 420°C experiments with air as the carrier
gas. There were three degradation products identified in
the water rinse from the 240°C experiments. Up to 13
degradation products were detected in the DCM trap,
13 in the water rinse, and five in the iso-octane rinse
in the 420°C experiments. The degradation products
ranged from single carbon compounds with 3 chlorine
atoms (i.e., chloroform) to compounds with 6 carbons
and 6 chlorine atoms (i.e., hexachlorobenzene). Carbon
monoxide (CO), CO2, and phosgene were detected in
the gas phase of the 420°C experiments. The amount
of carbon recovered for the 420°C experiments with
air as the carrier gas was 91.5% for the 22°C inlet
experiment but decreased to 79.1% for the experiment
completed with the inlet at 100°C. The amount of
chlorine recovered followed a similar trend with 88.8%
recovered for the 22°C inlet experiment and just 74.6%
recovered with the inlet at 100°C. TCE degradation
was proposed to be initiated by thermal induced
unimolecular dissociation but was also influenced by
the formation of peroxyl radicals due to the presence of
oxygen. Increasing the quartz tube water content in the
420°C experiments with air as the carrier gas may have
hydrolyzed phosgene and served to remove reactive
chlorine radicals and HC1 from the gas phase while not
impacting the amount of TCE degraded.
The experiments reported herein represent one
of only a few efforts to quantify gas-phase TCE
degradation and reaction product formation under
well-controlled thermal treatment conditions. Several
of the important results of this work are: (1) little, if
any, TCE degradation occurred at temperatures below
240°C; (2) at 420°C, up to 34 degradation products
were detected in the effluent solvent trap (e.g., CC14,
PCE, hexachloroethane), solvent and water rinses of
the reactor (DCAA, TCAA), and effluent gas (e.g., CO,
CO2, and phosgene); (3) at 420°C, with nitrogen as the
carrier gas degradation products accounted for 1 to 4%
of the TCE-carbon feed, and with air (22% oxygen)
as the carrier gas degradation products accounted for
28 to 38% of the TCE-carbon feed, with 18 to 23% of
the carbon attributed to the formation of CO and CO2.
Here, it is important to recognize that these results are
specific to the experimental conditions employed in
these laboratory studies, and do not precisely replicate
the field conditions. For example, the following
differences between the reported laboratory studies and
thermal treatment conditions in the field could alter TCE
degradation and product formation: (1) natural minerals
and organic matter present in subsurface soils could
either facilitate or quench specific reaction pathways,
(2) temperatures near thermal wells can be much higher
(e.g., 600-800°C) than those studied here (22 to 420°C),
which could lead to complete oxidation of reaction
products, (3) oxygen may be depleted in the thermal
treatment swept zone which could alter reaction product
distributions. Nevertheless, the experimental results
reported here provide quantitative measurements of gas-
phase TCE degradation and reaction product formation
in heated, flow-through reactors, and provide important
insight into the reaction products and pathways that
could potentially occur during thermal treatment of
TCE-contaminated soil.
3.6 Quality Assurance Summary for the
Flow-Through Experiments
These experiments involved passing a carrier gas that
contained TCE through a quartz tube heated to between
25 and 480°C. The quality assurance efforts for these
experiments focused on:
1. Assessing system cleanliness prior to each
experiment (pre-rinse/pre-trap)
2. Estimating the variability in sample collection and
analyses (replicates)
3. Demonstrating analysis method performance
relative to methylene chloride (matrix spike)
-------
4. Determining if contaminants were introduced during
sample storage (storage blanks)
System Cleanliness. The apparatus was assembled and
rinsed with freshly dispensed dichloromethane (DCM)
prior to each experimental run. Then carrier gas was
passed through the apparatus followed by a DCM
filled vial prior to TCE introduction. Samples from the
DCM rinse and trap were then analyzed to determine
if the decontamination methods were adequate. During
preliminary experiments it was discovered that rinsing
the apparatus with iso-octane was inadequate for
removing all the products formed after passing TCE
through the apparatus operated at 300°C; the results
of this experimental trial were discarded and not used.
Consequently, the decontamination method was modified
so that the apparatus was disassembled, washed in 45°C
soapy water, rinsed with deionized water, and heated in a
drying oven at 240°C for at least an hour. No compounds
were detected in samples from either the DCM rinse
or carrier gas trap collected prior to each subsequent
experiment.
Sample Collection and Analyses Variability. At least
two samples were collected from the DCM trap, iso-
octane rinse, and methyl-fer/-butyl ether (MTBE) used to
extract compounds from a water rinse of the apparatus.
These replicates samples were used to assess the range of
compounds in the samples and any variability introduced
during sample collection and analysis. Variability was
low, less than 10% relative standard deviation (RSD)
for all replicates, and less than 5% RSD for select
experiments. Given that analytical grade solvents were
used in these experiments, the low variability was
expected.
Method Performance. Assessing method performance
involved adding 1 mL of EPA 8240B/8260 A Matrix
Spike Mix (Sigma-Aldrich #47412) to the solvent
being analyzed (i.e., DCM, iso-octane, or MTBE) and
analyzing the matrix spike. The resulting analyses were
within 10% of the expected concentrations. This was
not unexpected as the matrix used for these experiments
were analytical grade solvents.
Storage Blanks. Vials filled with freshly dispensed
solvent were stored with each batch. No compounds
were detected in any of the storage blanks.
-------
4.0
TCE Degradation in Heated Ampules
4.1 Introduction
Four experimental series were performed to determine
the rate of TCE degradation and degradation products
formed after heating dissolved phase TCE to 120°C
over periods of up to 40 days. The experiments were
completed using glass ampules filled with TCE
contaminated water and sealed by melting the ampule
neck with a propane-oxygen torch (flame sealed).
Approximately three-quarters of the ampule volume was
filled with TCE contaminated water with the remaining
one-quarter volume contained gas, thus TCE was present
in both the dissolved- and gas-phase within the ampules
during the experiments. The first ampule experiment
was performed to demonstrate that dissolved oxygen
levels could be maintained in flame-sealed ampules by
measuring the dissolved oxygen concentration before
and after heating water-filled ampules to 120°C over
a period of 6 days (Table 4.1). The second ampule
experiment involved demonstrating analytical methods
to determine the aqueous phase concentrations of TCE
and dichloroacetic acid (DCAA), one of the anticipated
TCE degradation products. The third ampule experiment
introduced room temperature control ampules and
solids into the ampules along with an evaluation of the
method used to introduce TCE into the ampules. The
fourth ampule experiment was designed to determine the
rate of TCE disappearance along with the identity and
amount of each TCE degradation product as a function
of 1) dissolved oxygen concentration, 2) hydroxide ion
concentration, and 3) ampule solids content.
The following section (Section 4.2) describes the ampule
experimental system along with the methods used to
prepare the ampules. Experimental methods and results
specific to each of the experimental series are given in
Section 4.3. The final two sections provide a discussion
of the ampule results in terms of potential TCE
degradation mechanisms (Section 4.4) and a summary of
the ampule experiment results (Section 4.5).
4.2 Experimental Materials and
Methods
4.2.1 Preparation of Solids
Two solids compositions were used in the fourth ampule
experiment including 20-30 mesh Ottawa sand and a
mixture of 20-30 mesh Ottawa sand and 1% goethite.
The solids were prepared by soaking approximately
2,000 g of sand from Ottawa, IL (ASTM 20-30 Sand,
U.S. Silica Co., Berkeley Springs, WV) in 1 N nitric
acid solution as described in Section 3.1.3. However,
the DI rinse method employed was improved to remove
residual nitric acid from the sand. The DI rinse consisted
of placing small volumes of sand into the top of a
20-30-100 mesh ASTM sieve stack and running DI
water over the sand. The sand was then placed back into
a drying tray and DI-Nanopure water was added to cover
the sand. The pH of the DI-Nanopure water covering the
sand was measured with a pH probe (Accumet Model
50, Fisher Scientific, Fair Lawn, NJ) and the water rinse
was repeated until the pH of the standing DI-Nanopure
water was 7. The sand was then placed into a drying
Table 4.1 Ampule Experimental Matrix
Experimental
Series
1
2
3
4
Variables
Dissolved Oxygen, Hydroxide Ions
Dissolved Oxygen
Dissolved Oxygen
Solids
Dissolved Oxygen, Hydroxide Ions
Solids
Purpose
Demonstrate that ampules can retain oxygen
Identify reaction products
Demonstrate sample analysis techniques
Room temperature controls, solids, evaluation of TCE
introduction method
Determine rate of TCE disappearance and degradation products
as function of oxygen, hydroxide ion, and solids
-------
oven and heated to 130°C for 3 hours to remove excess
moisture and then baked at 200°C for 2 hours. The oven
temperature was lowered to 100°C and the sand was
allowed to cool for 3 hours.
Approximately 1,000 g of the acid-washed sand was
placed into a second glass drying tray to which 10 grams
of goethite powder was added to create a uniform
1% (wt) mixture. Research grade goethite chips,
approximately 1 gram each, were obtained from Ward's
Natural Science (Rochester, NY), and were reported to
have been collected from Grants County, New Mexico.
The goethite chips were ground into a fine powder (silt
to clay size particles) using a mortar and pestle prior
to mixing with the sand. The drying trays were then
autoclaved with steam at 17 psi (121°C) for 25 minutes
and the water from the autoclave process was allowed
to vent from the trays for a period of approximately
30 minutes. Ampules that had been autoclaved and
cooled in a desiccator according to procedures outlined
in Section 4.1.1 were loaded with approximately
20 grams of either Ottawa sand or Ottawa sand+1%
goethite and then sealed with aluminum foil.
4.2.2 Preparation of Aqueous Solutions
All aqueous solutions were prepared with deionized
(DI) water that was freshly dispensed from a Nanopure®
analytical deionization system (model D4741, Barnstead
International, Dubuque, IA). The Nanopure® system has
four inline purification cartridges that produce organic
free, Type I reagent grade water in accordance with
the specifications provided in the ASTM D1193-99el,
"Standard Specification for Reagent Water." The DI
water was dispensed only after the electrical resistance
of the water was greater than 18 M£l-cm at room
temperature (22°C) and through a 0.2 um pore size filter.
Aqueous solutions with low dissolved oxygen content
(<0.3 mg/L), referred to as anoxic water, were prepared
prior to each experiment by sparging freshly dispensed
DI-Nanopure water with argon gas (Airgas-South, Inc.,
Marietta, GA) after passing through an oxygen trap
(part* 4001, Alltech Associates, Inc., Deerfield, IL). The
anoxic water was sparged with argon for at least 1 hour
and had a dissolved oxygen concentration between 0.2
and 0.3 mg/L as indicated by the Rhodazine D method
(part* K7501, CHEMetrics, Inc., Calverton, VA).
Oxygen-saturated water, referred to as oxic water, was
prepared by sparging DI-Nanopure water with ultra zero
grade air (UZA) (Airgas-South, Inc., Marietta, GA). The
oxic water was sparged with UZA for at least 1 hour
and had a dissolved oxygen concentration between 8
and 10 mg/L as indicated by the Indigo Carmine method
(part* K7512, CHEMetrics, Inc., Calverton, VA). Gas
sparging was accomplished by passing the carrier gas
through a glass tube fitted with a fritted glass disk
that generated small bubbles to enhance gas transfer.
Sparging was completed within a 4 L aspiration carboy
that had been autoclaved with 17 psi of steam (121°C)
for 25 minutes prior to each use.
Stock solutions of TCE were prepared by transferring
either argon- or UZA-sparged DI-Nanopure water from
a 4 L carboy (Section 4.1.2) via gravity drainage into
2 L volumetric flasks. The 2 L flasks were prepared
prior to use by autoclaving with steam at 17 psi
(121°C) for 25 minutes, rinsing with DI-Nanopure
water (>18 MQ/cm), and drying at 200°C for 2 hours.
The flasks were allowed to cool to room temperature
within the drying oven and were then flushed with argon
gas prior to filling with sparged DI-Nanopure water.
Argon was used instead of nitrogen due to the greater
density of argon, 0.98 g/mL for argon vs. 0.68 g/mL for
nitrogen at 25°C (calculated using the ideal gas law),
which was thought to minimize the introduction of
atmospheric oxygen during anoxic ampule preparation.
A Teflon-coated stir bar was placed into each water-filled
flask, to which neat TCE was added using a gas-tight
syringe. Approximately 1.37 mL of TCE was added
to create 1,000 mg/L solutions (1.37 mLxl.46 g/mL^
2 L = 1000 mg/L), and 0.14 mL of TCE was added to
create 100 mg/L solutions (0.14 mix 1.46 g/mL^2 L =
102 mg/L). The 2 L flasks were then sealed with glass
stoppers and then the stoppers were wrapped with
parafilm. The 2 L flasks were wrapped in aluminum foil
to minimize exposure to light, and placed on a magnetic
stir plate where the contents of each flask were mixed
at room temperature for at least 24 hours. The pH
10 stock solution was prepared by adding 10 mL of a
NaOH solution (901.4 mg/L) to the 2 L flask containing
100 mg/L of TCE just prior to ampule loading. The
resulting NaOH concentration was 0.26 mM in the 2 L
volume with 100 mg/L of TCE. American Chemical
Society (ACS) certified NaOH obtained from Fisher
Scientific (Fair Lawn, NJ) was used to prepare the stock
solution.
4.2.3 Preparation of Ampules
The ampule experiments were conducted in clear, 25 mL
(Kimble-Kontes, Vmeland, NJ) or 50 mL (Wheaton
Science Products, Millville, NJ) borosilicate glass
ampules (Figure 4.1). The 25 mL ampules were used
for the initial three experimental series and the 50 mL
funnel-top ampules were used for the fourth experiment
to minimize the amount of carbon monoxide and
carbon dioxide (CO/CO2) introduced during the flame
sealing process. The ampules were autoclaved with
17 psi of steam (121°C) for 25 minutes, then rinsed
with deionized (DI) water (>18 MQ/cm), and dried in
an oven at 200°C for 2 hours. The ampules were then
removed from the oven and placed in a glass desiccator
that contained approximately 100 grams of indicator
drierite (97 % CaSO4 and 3% CoCl2) to maintain water
-------
free conditions. For the anoxic and oxic experiments,
the desiccator was evacuated to 750 mm Hg of vacuum
and then backfilled with either argon gas or ultra
zero grade air (UZA), respectively. Each ampule was
then flushed with argon or UZA just prior to filling
with aqueous solution. Each water-filled ampule was
temporarily sealed with aluminum foil until a complete
batch of ampules was prepared (~10 minutes). Ampules
containing solids were prepared in an identical fashion,
but filled with either sand or sand+1% goethite as they
were removed from the desiccator.
The ampules were flame sealed using a propane-oxygen
torch (BernzOMatic, Medina, NY) that has a maximum
flame temperature of approximately 2,500°C. The flame
sealing process consisted of heating the ampule neck
using the outer portion of the torch flame to vaporize
any water droplets present within the neck followed by
melting the glass using the inner portion of the torch
flame. The flame seal location was approximately 3 cm
above the gold band (Figure 4.1) in accordance with
Wheaton Science instructions. The vaporization of water
required approximately 10 seconds, while melting the
glass to form the seal required less than 5 seconds. The
sealed ampule was then placed in a rack and allowed to
cool to room temperature.
Flame Seal
Gold Band'
Pre-ScoredNeck
10cm
Figure 4.1 Picture of the 50 mL funnel-top ampule
before and after sealing.
Each ampule was labeled with a permanent marker to
indicate the sequential ampule number, preparation
date, and ampule contents. The ampules were then
weighed using an analytical balance (Model# AG245,
Mettler-Toledo, Columbus, OH) after checking the
balance accuracy with an ASTM E617 class 2 certified
traceable 20±0.0001 gram weight (Cat. # 820000.2,
Denver Instruments, Denver, CO).
4.2.4 Description of Ampule Experiments
4.2.4.1 Ampule Experiment 1
The first ampule experiment consisted of filling each
of 4, 25 mL ampules with 20 mL of DI-Nanopure
water (>18 MQ-cm), leaving approximately 5 mL of
gas headspace in each ampule. Two (2) ampules were
filled with nitrogen sparged water with initial dissolved
oxygen (DO) concentration of less than 0.5 mg/L
(anoxic water) and 2 ampules were filled with ultra
zero grade air (UZA) sparged water with initial DO
concentration of 8.17 mg/L (oxic water). Approximately
1.4 uL of neat TCE was then injected into each of the
4 ampules through a temporary aluminum film seal
to create aqueous solutions containing approximately
100 mg/L of TCE. The ampule with anoxic water and
TCE represented the control since Knauss et al. (1999)
had found that dissolved oxygen affected the rate of
TCE disappearance. Two (2) of the 4 ampules were
amended with solid sodium hydroxide (NaOH) chips to
adjust the solution pH to approximately 11. No duplicate
ampules were prepared in this first ampule experiment.
All 4 ampules were flame-sealed and placed in an oven
maintained at 120°C for a period of 6 days.
At the end of 6 days the oven was turned off and allowed
to cool for 12 hours to room temperature (22° C). The
ampules were destructively sampled and the gas-phase
CO2 concentrations was determined using a GC (HP
6890) equipped with a gas sampling valve, a Supel-Q
PLOT capillary column (Supelco, Bellefonte, PA), and
a thermal conductivity detector (TCD). The Supel-Q
PLOT column was only capable of separating CO2
from the ampule gas and using this column resulted
in a detection limit of 200 ppmv for CO2. The DO
concentration of each sample was determined using
a membrane-covered voltammetric sensor (YSI 5010
BOD Probe, YSI, Inc. Yellow Springs, OH). Aqueous
samples were collected from each ampule and injected
into a GC (Varian 3600CX) equipped with a Varian
Saturn 2000 mass spectrometer (MS) to analyze for TCE
degradation products.
4.2.4.2 Ampule Experiment 2
The second ampule experiment consisted of filling
each of 8, 25 mL ampules with 20 mL of DI-Nanopure
water, leaving approximately 5 mL of gas headspace
in each ampule. Four (4) ampules were filled with
nitrogen sparged water with initial DO concentration
of 0.68 mg/L (anoxic water) as measured using the YSI
voltammetric sensor, and 4 ampules were filled with
UZA sparged water with initial DO of 9.8 mg/L (oxic
water). Approximately 1.4 uL of neat TCE was added to
-------
7 of the 8 ampules to achieve an initial aqueous phase
TCE concentration of approximately 100 mg/L, while
one ampule with oxic water was TCE-free to serve as
a control. All of the ampules were flamed sealed and
placed in an oven at 120°C for a period of 10 days.
One ampule with anoxic water and TCE was broken
during the flame sealing process, leaving 7 ampules for
incubation.
4.2.4.3 Ampule Experiment 3
The third ampule experiment was designed to
incorporate a set of room temperature controls, introduce
solids (20-30 mesh Ottawa sand) into the ampules,
and evaluate the method of introducing TCE into the
ampules. The experiment involved 25 mL ampules
that contained 20 mL of DI-Nanopure water and
approximately 5 mL of gas headspace. Twelve (12)
ampules contained anoxic water and nitrogen gas, and
a matching set of 12 ampules contained oxic water and
UZA gas for a total of 24 ampules in all. Approximately
10 grams of 20-30 mesh Ottawa sand were added to
8 ampules, 4 with anoxic water and 4 with oxic water.
Approximately 1.4 uL of neat TCE was introduced
into 10 of the 12 ampules with anoxic water, while
the remaining 2 ampules were TCE free. TCE was
introduced into 10 of the 12 oxic ampules as a 100 mg/L
aqueous solution that was prepared by adding neat TCE
to a 250 mL volumetric flask filled with UZA-sparged
DI-Nanopure water and a Teflon-coated stir bar. The
250 mL flask was sealed, placed on a magnetic stir plate,
and the contents mixed for 12 hours at room temperature
(22°C) before filling each ampule with 20 mL of the
solution via glass pipette.
After filling the ampules with aqueous solution, a
100 uL aqueous sample was collected from each ampule
in an effort to determine the initial concentration
of TCE. The samples were analyzed using a mass
spectrometer (ITS40, Thermo-Finnegan, Waltham,
MA) equipped with a vial sparge module that allowed
TCE to be purged from water in a 40 mL vial directly
into the mass spectrometer (MS) per EPA method
8265 (U.S. EPA, 2002). The vial sparge method was
being evaluated as an alternative to the direct GC
injection method for determining aqueous-phase TCE
concentration. After collecting a sample to determine
the initial TCE concentration, each ampule was flame
sealed using a propane torch and 19 of the ampules
were placed in an oven maintained at 120°C. The
remaining 5 ampules were wrapped with aluminum
foil and stored in a vented hood at room temperature
(22°C). The 19 ampules were removed from the oven
after 10 days and allowed to cool to room temperature
and then all 24 ampules were destructively sampled on
the same day in numerical order. The gas phase from
each ampule was initially analyzed for CO2 content
followed by the analysis of aqueous phase samples to
determine the concentration of TCE, DO, DCAA, and
chloride ions. The concentration of TCE in each ampule
after the 10 day period was determined using the vial
sparge method and a headspace method that involved
collecting a 1 mL aqueous sample from each ampule and
injecting the sample into a previously sealed headspace
vial for analysis by a GC equipped with a headspace
autosampler.
4.2.3.4 Ampule Experiment 4
The fourth ampule experiment was undertaken to
determine the rate of TCE degradation and the TCE
degradation products formed after incubating dissolved-
phase TCE at 120°C for up to 40 days. The experimental
matrix was expanded compared to the initial three
experiments and included the following experimental
variables: 1) dissolved oxygen concentration,
2) hydroxide ion concentration, and 3) solids content
(Table 4.2).
Table 4.2 Experimental Matrix Used for the Fourth Ampule Experiment
Liquid Content
(-50 mL)
Anoxic water, pH 7
Oxic water, pH 7
Anoxic water, pH 10
Anoxic water, pH 7
Anoxic water, pH 7
Solids Content
(20 g total solids)
None
None
None
ASTM 20-30 Sand
20-30 Sand +
1% Goethite
Headspace Gas
(-20 mL)
Argon
UZA
Argon
Argon
Argon
Experimental
Variable
Control
Oxygen
Hydroxide Ion
Solids
Goethite
-------
Of particular note, the solids were expanded to include
goethite, a common Fe-containing soil mineral which
is known to be chemically reactive. The experimental
methods were refined or modified based on the results
of the initial three ampule experiments. Several batches
of ampules were prepared over a period of two weeks to
evaluate the experimental variables listed in Table 4.2.
The ampules prepared with TCE were designed to
evaluate the rate of TCE degradation and the degradation
products formed. Ampules prepared without TCE or
solids were intended to determine if CO and CO2 were
being introduced during the flame sealing process.
Ampules without TCE but with solids were intended to
determine the amount of CO2 that could be attributed to
the presence of solids. The number of ampules prepared
for each experimental batch are listed in Table 4.3.
An initial TCE concentration of 100 mg/L was intended
for all experiments, however, a miscalculation led to the
preparation of a 1,000 mg/L TCE solution used for the
initial 2 batches of ampules (Ampules 1-61), which were
added to the experimental matrix. Three (3) additional
ampules (Nos. 3a, 6a, and lOa) were prepared in the first
batch to replace ampules that had to be re-sealed. These
3 re-sealed ampules were retained to evaluate the effect
of exposing the ampule contents to the propane-oxygen
flame. Only 6 TCE-free ampules were prepared in the
second batch instead of the 18 that were planned since
the 1,000 mg/L TCE concentration was in addition to
the planned 100 mg/L concentration. In Batch 6 and 7,
one of the TCE-free sand and one of goethite-containing
ampules cracked open during preparation and were
not replaced. In all, 242 ampules were prepared for the
fourth ampule experiment.
The initial concentration of TCE in the ampules was
determined by analyzing aqueous samples from 3 of
the ampules in each batch just prior to flame sealing
(Section 4.5.2). The average initial TCE concentration
for Batch 1 was 893±21 mg/L, 878±44 mg/L for Batch
2, 95.2±3.5 mg/L for Batch 3, 85.3±2.5 mg/L for Batch
5, and 77.0±6.4 mg/L for Batch 4 based on the direct
GC analysis method as the headspace GC equipment
was unavailable. The solids-containing ampules had an
average initial TCE concentration of 87.5±3.6 mg/L,
as both the Ottawa sand and Ottawa sand+l%goethite
ampules (Batches 6 and 7) were prepared using the
same anoxic TCE solution. The average initial TCE
concentration for the 1,000 mg/L solutions prepared for
Batch 1 and 2 was less than 1,000 mg/L, as was the TCE
concentration in the 100 mg/L solutions prepared for
Batches 3 through 7 (Table 4.4). Thus the 1,000 mg/L
and 100 mg/L ampules are in name only (i.e., nominal),
with the actual initial concentrations provided in
Table 4.4.
Table 4.3 Summary of Ampules Prepared for the Fourth Ampule Experiment
Ampule
Batch No.
1
2
3
4
5
6
7
Number of Ampules
Prepared
with
TCE
21
18
18
18
18
18
18
without
TCE
18
6
18
18
18
17
18
Nominal
Initial TCE
Concentration
(mg/L)
1,000
1,000
100
100
100
100
100
Oxygen
Content
Anoxic
Oxic
Oxic
Anoxic
Anoxic
Anoxic
Anoxic
Solids Content
None
None
None
None
None (pH 10)
Sand+
1% Goethite
Sand
Ampule
Number
1-36
37-61
62-97
98-133
134-170
171-188
209 - 225
189 - 207
226 - 243
-------
Table 4.4 Schedule for Convection Oven Ampule Incubation
Initial TCE
Concentration
(mg/L)
893±21
878±44
95.2±3.5
77.0±6.4*
85.3±2.5
87.5±3.6
87.5±3.6
Oxygen
Content
Anoxic
Oxic
Oxic
Anoxic
Anoxic
Anoxic
Anoxic
Batch No.
No. of Ampules
1
18 [9]
2
12 [9]
o
3
18 [9]
4
18 [9]
5
18 [9]
6
18 [9]
7
18 [9]
10 day
No. of Ampules
6 [3]
4 [3]
6 [3]
6 [3]
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
20 day
No. of Ampules
6 [3]
4 [3]
6 [3]
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
30 day
No. of Ampules
oven exploded
6 [3] Lost
oven exploded
4 [3] Lost
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
oven exploded
6 [3] Lost
*Analyzed using the direct GC injection method, all other ampules by headspace analysis.
(#) - Batch number corresponds to Table 4.10.
[#] - Number of ampules that contained TCE.
After flame-sealing the ampules in each batch, the
ampules were divided in half so that 18 ampules (9
with TCE and 9 without TCE) were incubated at
120°C in a convection oven (VWR Model 1320, VWR
International, West Chester, PA), while the remaining
18 ampules (9 with TCE and 9 without TCE) were stored
at room temperature (22°C) in the dark. The ampules
for oven incubation were contained in a 3 L Pyrex glass
drying tray that was placed in the convection oven at
120°C where the temperature was measured using a
certified traceable thermometer (part* 15-060-223,
Fisher Scientific, Fair Lawn, NJ) located within the
oven enclosure. The ampules stored at room temperature
were placed in a storage rack located inside an insulated
container (ice chest) to minimize exposure to light.
Ampules with TCE and without TCE were removed
from the oven and room temperature storage at 10 and
20 day intervals for destructive sampling according to
the schedule listed in Table 4.4 and 4.5. Four days after
placing the solids filled ampules in the convection oven,
an explosion occurred that destroyed the 82 ampules
being incubated within the oven. The 82 ampules that
were stored at room temperature (Table 4.5) were
moved to a 4°C chamber to minimize the formation of
potentially explosive TCE degradation products.
Following the explosion, the remaining 1,000 mg/L
ampules that had been stored at room temperature
(Table 4.5) along with the 20-day, 100 mg/L oxic
ampules that had been removed from the oven just
prior to the explosion were destructively sampled to
determine if dichloroacetylene (C2C12) had formed.
Dichloroacetylene (DCA) is reportedly a spontaneously
explosive compound (Urben et al., 1999) and was
suspected to have caused the oven explosion. DCA was
one of the TCE degradation products detected during
the high temperature degradation of TCE in quartz tubes
(Wu and Lin, 2004; Kim and Choo, 1983) but was not
anticipated to form in the ampules at 120°C with water
present. DCA was detected in the gas phase from the
100 mg/L oxic ampules that had been incubated at 120°C
in the convection oven and thus the ampule experiment
was considered to be an explosion hazard. Butler and
Hayes (2001) warn of the possibility of ampules popping
or exploding as a result of the buildup of gases such as
hydrogen when metals are present in the ampules. As a
consequence, an explosion-resistant incubation apparatus
was constructed to incubate the 72, 100 mg/L ampules
that were being stored in the 4°C chamber.
The explosion-resistant apparatus consisted of a block
of aluminum into which 2 inch diameter holes were
drilled to accommodate the ampules. The aluminum
-------
Table 4.5 Schedule for Room Temperature Ampule Storage
Initial TCE
Concentration
(mg/L)
893±21
878±44
95.2±3.5
77.0±6.4
85.3±2.5
87.5±3.6
87.5±3.6
Oxygen
Content
Anoxic
Oxic
Oxic
Anoxic
Anoxic
Anoxic
Anoxic
Batch No.
No. of Ampules
1
21 [12]
2
12 [9]
3
18 [9]
4
18 [9]
5
18 [9]
6
17 [9]
7
18 [9]
10 day
No. of Ampules
8 [5]
4 [3]
6 [3]
6 [3]
6 [3]
6 [3]
6 [3]
20 day
No. of Ampules
7 [4]
4 [3]
6 [3]
6 [3]
6 [3]
6 [3]
6 [3]
30 day
No. of Ampules
6 [3]
4 [3]
6 [3]
6 [3]
6 [3]
5 [3]
6 [3]
(#) - Batch number corresponds to Table 4.3.
[#] - Number of ampules that contained TCE.
block was heated using a standard bench top hot plate
(Fisher Scientific, Fair Lawn, NJ) and the temperature of
the block was determined using a K-type thermocouple
connected to a data logger (Model# CR23X, Campbell
Scientific, Inc., Logan, UT), which automatically
recorded the temperature at 15-minute intervals. As
shown in Figure 4.2, the hot plate and aluminum block
were located behind a 2 by 2 foot, 1/2 inch thick sheet
of polycarbonate (Part* 8574K57, McMaster-Carr,
Atlanta, GA) that served to shield laboratory personnel
from exploding ampules. The ampules were manipulated
behind the polycarbonate sheet using an 18-inch long
pair of metal tongs. In addition, laboratory personnel
wore a polycarbonate face shield (Fisher Scientific, Fair
Lawn, NJ) when manipulating ampules.
Polycarbonate
Shield
Aluminum Block
Tongs
Figure 4.2 Explosion-resistant ampule incubation apparatus.
-------
Incubation of the ampules that were stored in the 4°C
chamber was done according to the schedule given
in Table 4.6. The ampules had been stored in the 4°C
chamber for approximately 5 months prior to incubating
in the explosion-resistant apparatus. The ampules that
contained Ottawa sand+1% goethite (Batch 6) were
incubated first with the duration of incubation at 120°C
limited to 4 days since the oven explosion occurred
4 days after placing the solids-filled ampules in the oven.
The ampules containing pH 10 water (Batch 5) were
then incubated over a 4-day period. The ampules with
sand (Batch 7) were incubated for 40 days at 120°C
after significant levels of DCA had been detected in the
pH 10 ampules. The remaining oxic ampules (Batch 3)
were then incubated at 120°C for 30 days and the
anoxic ampules (Batch 4) for 30 and 41 days. A limited
number of ampules from each batch were stored at room
temperature including: 4 of the sand+goethite, 4 of the
pH 10, 4 of the sand, 2 of the oxic, and 4 of the anoxic
ampules.
4.2.4 Ampule Sampling Methods
The ampule sampling process was initiated by removing
the ampules from the oven and placing them in a
darkened vent hood to allow the hot ampules to cool
to room temperature. The ampules were then weighed
using the same analytical balance used to determine
the initial ampule weight. The ampule opening method
consisted of inverting the ampule and measuring the
distance of the gas-filled space to estimate the volume
of gas in each ampule (Step 1, Figure 4.3). The ampule
neck was then broken by hand along the pre-scored line
(Step 2, Figure 4.3). The water within the ampule body
did not drain out since gas could not flow past the water
that blocked the ampule opening which was smaller in
diameter than the ampule body. The opened ampule was
then placed into a custom made sampling collection
apparatus that was filled with a stream of argon gas
flowing at 100 mL/min to minimize the introduction of
oxygen and carbon dioxide during sample collection
(Step 4, Figure 4.3). The sample collection apparatus
consisted of a 60 mL plastic syringe body that was cut in
half with an 18-gauge, 30 cm long stainless steel needle
affixed to the syringe body. A 10 mL syringe (Becton
Dickinson and Co., Franklin Lakes, NJ) was attached to
the 30 cm long needle via a Luer lock connection in an
effort to collect a gas sample from the inverted ampule.
Table 4.6 Schedule for Explosion-Resistant Apparatus Ampule Incubation
Ampule Content
Sand+Goethite
(6)
pHIO
(5)
Sand
(7)
Oxic
(3)
Anoxic
(4)
No. of Ampules
12 [6]
14 [7]
14 [7]
3 [2]
8 [4]
Time (days)
No. Ampules
1
4 [2]
1
4 [2]
10
4 [2]
30
3 [2]
30
4 [2]
Time (days)
No. Ampules
2
4 [2]
2
4 [2]
30
4 [2]
~
41
4 [2]
Time (days)
No. Ampules
3
4 [2]
3
4 [2]
40
6 [3]
~
~
Batch 4
No. Ampules
~
4
2[1]
~
~
~
(#) - Batch number corresponds to Table 4.10.
[#] - Number of ampules that contained TCE.
-------
V
i
^
CHEMets®
1
60 mL plastic
syringe body
filled with argon
Ampule
/ argon — »• — 1=
j
1
1
1
si
\ 1
D
D
0
1
1
1
P
1
1
argon bubbles
^
^ 18-gauge needle
10-mL syringe
/
S
1. Invert ampule and 2. Crack open ampule 3. Measure dissolved 4. Collect 10 mL gas
measure length of gas along pre- scored neck oxygen from ampule sample
filled space neck
Figure 4.3 Illustration of the ampule gas sample collection method.
4.2.5 Analytical Methods
The dissolved oxygen (DO) concentration was
determined using a membrane-covered voltammetric
sensor (YSI 5010 BOD Probe, YSI, Inc. Yellow Springs,
OH) during the initial three experiments. This method
required transferring the aqueous phase from each
ampule into a second vial that could accommodate the
YSI probe body. The dissolved oxygen concentration of
ampules in the fourth experiment was determined using
the Rhodazine D method (part* K7501, CHEMetrics,
Inc., Calverton, VA) for DO between 0 and 1 mg/L or
the Indigo Carmine method (part# K7512, CHEMetrics,
Inc., Calverton, VA) for DO between 1 and 10 mg/L.
The CHEMets method was contained within a self-
filling ampule that was inserted into the ampule neck
(Step 3, Figure 4.3) to determine the DO concentration
while waiting for argon to flush ambient air from the
60 mL syringe body. A 10 mL sample of the gas within
the ampule was then collected by slowly retracting the
syringe plunger (10 mL in 30 seconds) and allowing
the argon gas to bubble through the liquid-filled portion
of the ampule and backfill the gas-filled space as the
gas sample was being removed. The 10 mL gas sample
was analyzed within 60 seconds of collection using a
gas chromatograph (HP 6890) equipped with a thermal
conductivity detector (TCD). The ampule was then
removed from the 60 mL syringe body and placed
upright to allow for collection of liquid samples. The
water within the ampule was then transferred into a
pre-washed 50 mL borosilicate glass vial and sealed
with a Teflon® lined septum affixed with an open-top
screw cap. The pH of the ampule water was measured
in the 50 mL glass vial by placing a DI water rinsed pH
probe (Fisher Scientific, Fair Lawn, NJ) into the vial and
waiting approximately 5 minutes before recording the
pH value. The vial was then labeled using a permanent
marker and stored in a 4°C chamber.
Aqueous phase concentrations of TCE were determined
by gas chromatography (GC) using a headspace method
and a direct GC injection method. The headspace
method consisted of transferring a 1 mL aqueous sample
into a 20 mL vial that was sealed with a Teflon-lined,
butyl rubber stopper (West Pharmaceutical Services,
Inc., Lionville, PA) affixed with an aluminum crimp
cap. The headspace vials were placed in an autosampler
(HP 7694) that was programmed to heat each sample to
70°C for a period of 15 minutes prior to transferring the
headspace gas into an HP 6890 GC for analysis. The GC
was equipped with a30mx0.32 mm DB-624 column
(Agilent Technologies, Palo Alto, CA) connected to a
flame ionization detector (FID). Calibration standards
containing 60, 80, and 100 mg/L TCE were analyzed
with each experimental sample batch. The calibration
standards were prepared by injecting small volumes of
10,000 mg/L TCE methanol stock solution into 100 mL
flasks that contained DI-Nanopure water cooled to 4°C.
The direct GC injection method involved introducing
1 uL of aqueous solution into a HP 6890 GC equipped
with a 990 uL inlet liner, a 30 m x 0.32 mm OD DB-5
column (Agilent Technologies, Palo Alto, CA), and a
FID. The inlet was operated at 200°C with a constant
helium pressure of 20 psi and a 2:1 split ratio: these inlet
parameters served to minimize water vapor back-flash
after injecting the 1 uL aqueous sample.
Immediately prior to flame sealing, the initial aqueous-
phase concentration of TCE in the ampules was
determined by collecting a 1 mL water sample from
-------
three randomly-selected ampules in each experimental
batch of 18 ampules. The TCE concentration measured
immediately prior to sealing the ampules was
considered to be representative of the initial aqueous
phase TCE concentration. Following incubation, the
aqueous phase TCE concentrations were determined
by collecting 3, 1 mL water samples from each ampule
immediately after collecting a 10 mL gas sample. Two
(2) of the 3, 1 mL water samples, which contained
1,1,1-trichloroethane as an internal spike, were analyzed
by the direct GC injection method and the third 1 mL
water sample was analyzed by the headspace method.
The concentrations of formate (CHOO"), glycolate
(HOH2C2OO-), sulfate (SO/-), and chloride (CT)
in aqueous phase samples from the ampules were
determined using a Dionex DX-100 Ion Chromotograph
(1C) equipped with an AS 14 A lonPac column with 8 mM
Na2CO3/l mM NaHCO3 eluent concentrate flowing
at 1 mL/min. Organic acid calibration standards were
prepared from 1 M stock solutions in the concentration
range from 0.02 to 0.50 mM. Formate and glycolate
solutions were prepared from 99% grade solids (glycolic
acid and sodium formate, ACROS Organics, Morris
Plains, NJ). One limitation of 1C analysis is the relatively
high detection limit: the average method detection
limits for glycolate and formate were 0.86 mg/L and
0.31 mg/L, respectively. The chloride ion content of
aqueous samples was determined by 1C analysis to allow
for comparisons with chloride concentrations measured
using the titration method of Bergmann and Sanik
(1957). The DO content of water held within the ampule
neck was measured immediately after opening using the
CHEMets ampules (Step 3, Figure 4.2). The aqueous
phase concentration of DCAA was determined using the
modified EPA method 552.2.
4.3 Experimental Results
4.3.1 Results of Ampule Experiment 1
Carbon dioxide (CO2) was detected in the 2 ampules
without NaOH, regardless of the initial dissolved oxygen
content (Table 4.7). No TCE was detected in water
samples collected from the NaOH amended ampules
while TCE was present in the ampules without NaOH.
The concentration of TCE was not determined. The
absence of CO2 and TCE in the NaOH amended ampules
indicated that other TCE degradation products had
formed, which was suspected to be dichloroacetic acid
(DCAA) based on the past experimental work presented
in Section 2.3.2. Additional aqueous phase samples were
collected from the NaOH amended ampules and the pH
of the samples were adjusted to less than 1 by adding
concentrated sulfuric acid. The pH adjustment was
performed to convert any organic ions that may have
been present from the anionic to the acid form because
organic acids can be detected using an ultraviolet (UV)
light spectrophotometer. After adjusting the pH of each
sample to less than 1, the water samples from the NaOH
amended ampules absorbed light at 270 nm in a Varian
UV-visible spectrophotometer, which, based on the
similarity to work by Mertens and von Sonntag (1994),
suggested that DCAA was present.
The first ampule experiment demonstrated that DO
levels in the ampules could be maintained with greater
than 95% of the initial amount of the DO detected
after incubating the sealed ampules for 6 days at
120°C. In addition, the importance of NaOH on the
rate of TCE disappearance was demonstrated in that
no TCE was detected in NaOH amended ampules
after 6 days at 120°C, while TCE was detected in the
ampules without NaOH. It was evident that a method to
determine the DCAA content of water solutions needed
to be developed based on the observation of UV light
absorbance in water samples from the NaOH amended
ampules. A method to determine the concentration of
TCE in water was also found to be necessary in effort to
determine if small changes in TCE content were equal to
the amount of CO2 detected.
4.3.2 Results of Ampule Experiment 2
After cooling to room temperature, the ampules were
destructively sampled to determine the amount of CO2
in the gas phase and the concentration of DO, TCE,
and DCAA in the aqueous phase. Gas samples from
Table 4.7 Results of the First Ampule Experiment After 6 Days at 120°C
Ampule Contents
Water and TCE
Water and TCE
Water, TCE, and NaOH
Water, TCE, and NaOH
Initial DO (mg/L)
0.5
8.17
<0.5
8.17
Final DO (mg/L)
0.5
7.8
O.5
7.9
Final pH
6.74
6.60
11.24
10.75
CO2 in Gas Phase
(uL/L)
1032
1000
nd
nd
nd - below detection limit (-200 uL/L)
-------
3 of the 7 ampules were not analyzed because these
ampules were damaged during the opening process
which resulted in the ampule contents being exposed to
ambient air (-500 uL/L CO2) and potentially biasing the
amount of CO2 in those samples. The TCE content of
the ampule aqueous phase was determined by injecting a
1 uL water sample from each ampule directly into a gas
chromatograph (GC) equipped with a flame ionization
detector (FID). The DCAA concentration of the
ampule water was determined using the modified EPA
method 552.2.
The ampules with oxic water had DO at concentrations
ranging from approximately 7.5 to 8.6 mg/L after
10 days at 120°C (Table 4.8) as determined using the
YSI voltammetric sensor. The similarity between initial
and final DO concentrations of the control ampule
(9.08 vs. 8.52 mg/L), and minimal changes in ampule
weights (< 0.0006%) were taken to indicate that the
flame-sealed ampules provided a gas-tight environment
over the 10-day, 120°C incubation period. For ampules
that contained TCE, the pH decreased from 7 to
approximately 6.2, and CO2 was detected in the ampules
with both anoxic and oxic water. As anticipated, CO2
was not detected in the gas headspace of the TCE-free
control ampule. Dichloroacetic acid (DCAA) was
detected at concentrations near the method detection
limit (~5 ug/L) in 3 of the 7 ampules.
The pH of the TCE-free control ampule increased
from 7.15 to 8.19, which was attributed to the
thermal enhanced dissolution of silica (SiO2) from the
borosilicate glass ampule walls. As SiO2 dissolves into
water it forms silicic acid (H4SiO4), which is a weak
acid with an initial dissociation constant of pK =9.5
(H4SiO4^H3SiO; + H+) (Stumm and Morgan, 1996).
Thus the dissociation of silicic acid would cause the pH
of the solution to increase as SiO2 was dissolved from
the ampule walls.
The initial concentration of TCE was not determined
using the direct GC injection technique but was in
the 95 to 115 mg/L range as estimated based on the
mass of TCE added to each ampule and the volume of
water in each ampule (Table 4.9). The concentration
after incubating the ampules at 120°C for 10 days was
determined using the direct GC injection technique and
was less than the initial, calculated TCE concentration.
The relative standard deviation (%RSD= standard
deviation •*• average x 100) is one measure of the
precision associated with an analysis method. A %RSD
of less than 15% is considered adequate for determining
the concentration of TCE in aqueous samples (U.S.
EPA, 1996). The %RSD values shown in Table 4.4 were
determined by analyzing three separate water samples
collected from each ampule and shows that the direct GC
injection method was capable of determining aqueous
phase TCE concentration with adequate precision.
There appeared to be a significant reduction in the
amount of TCE after 10 days at 120°C based on the
results shown in Table 4.9; however, the small amount
of CO2 detected and DCAA indicated that the apparent
reduction in TCE was due to an overestimation in the
initial concentration of TCE. This led to the conclusion
that the initial concentration of TCE should be measured
rather than estimated and an additional set of ampules
should be maintained at room temperature to help in
quantifying any temperature induced changes in TCE
content.
Table 4.8 Results of Second Ampule Experiment After 10 Days at 120°C
Ampule Contents
Water and TCE*
Water and TCE*
Water and TCE
Water and TCE
Water and TCE*
Water and TCE
Water
Initial DO
(mg/L)
0.68
0.68
0.68
9.08
9.08
9.08
9.08
Final DO
(mg/L)
na
na
na
7.5
8.55
8.43
8.52
Initial
pH
7.34
7.34
7.34
7.15
7.15
7.15
7.15
Final pH
na
6.1
na
6.3
6.47
6.16
8.19
Weight Change
(%)
0.0006
0.0006
0.0006
0.0003
0.0006
0.0000
0.0006
CO2 (uL/L)
1007
na
na
na
879
739
nd
DO - dissolved oxygen
* DCAA detected in water samples near detection limit of 5 ug/L.
na - not analyzed, contents lost after breaking ampule neck
nd - not detected, below detection limit of -200 uL/L
-------
Table 4.9 Initial and Final TCE Concentrations in the Second Ampule Experiment
Ampule Contents
Water and TCE*
Water and TCE*
Water and TCE
Water and TCE
Water and TCE*
Water and TCE
Water
Initial TCE (mg/L)*
110.0
95.0
105.0
100.0
100.0
115.0
not added
Final TCE (mg/L)f
57.9
66.5
42.9
49.5
60.2
65.3
nd
Relative Standard Deviation
(%RSD)
3.4
0.5
5.2
0.6
3.2
4.7
not determined
* Calculated based on the mass of TCE added to each ampule and volume of water in each ampule.
f Determined using direct GC injection technique.
nd - not detected
4.3.3 Results of Ampule Experiment 3
The initial concentration estimated by the mass of TCE
added was greater than the concentration determined
by the vial sparge method (Table 4.10). The difference
between the estimated and measured TCE concentration
was thought to be due to non-equilibrium between
the neat TCE droplet that was added to each ampule
and the ampule water because the sample used to
determine the initial TCE concentration was collected
within 30 minutes of injecting the neat TCE. The non-
equilibrium condition between the neat TCE droplet and
ampule water was indicated by the significant difference
(P-value=0.01, alpha=0.05) between the initial TCE
concentration determined for Ampule 1 (46.9+1.0 mg/L)
and for the other identically prepared ampules
(69.4+3.8 mg/L for Ampule 2 and 69.4+3.8 mg/L for
Ampule 3) even though a similar amount of neat TCE
was added to each ampule (2.1, 2.3, and 2.2 mg TCE,
respectively). The difference in the initial concentration
between ampules indicated that the neat TCE in Ampule
1 was not completely dissolved into the ampule water
at the time when the water sample was collected in an
effort to estimate the initial TCE concentration. The
concentration of TCE in Ampule 1 measured after
10 days at 120°C increased from 46.9 mg/L to 66.7 mg/L
which supports the conclusion that a portion of the
neat TCE was not dissolved when the sample used to
establish the initial TCE concentration was collected.
In addition to the non-equilibrium condition, there must
have been a loss of TCE mass after injecting the neat
TCE droplet into each ampule since the concentration,
as determined by the vial sparge method, was less than
the concentration estimated from the mass of TCE added
to each ampule. Thus it was concluded that injecting
neat TCE to establish an initial mass of TCE was not
desirable since estimating the initial concentration of
TCE was complicated by the non-equilibrium condition
and loss of TCE mass during the loading process.
Table 4.10 Initial and Final TCE Concentration in Anoxic Ampules from the Third Ampule Experiment (1.4 uL of
TCE added to each ampule, Initial DO = 0.79 mg/L)
Contents
TCE, water
TCE, water, sand
Temperature
(°C)
120
22
120
22
Initial TCE (mg/L)
Estimated
110+7.1
110
107+7.6
110
Vial Sparge
58.2+15.9
77.8
87.9+12.2
84.8
Final TCE (mg/L)
Vial Sparge
62.4+6.2
80.6
62.5+5.1
80.6
Headspace
61.5+3.7
69.9
63.2+5.7
81.2
Error represents one standard deviation calculated from duplicate ampule results.
-------
Table 4.11 contains the initial and final concentration
of TCE in the ampules filled with oxic water. Here,
instead of amending each ampule with 1.4 uL of neat
TCE, as done with the ampules filled with anoxic water
(Table 4.10), the ampules with oxic water were filled
with an aqueous mixture that had been prepared by
mixing neat TCE with water at room temperature for a
period of 12 hours. The average concentration of TCE
in the mixture was 71. 7±4.1 with a %RSD of 5.7. This
average was calculated based on the analysis of a 100 uL
sample collected from each of the 10 ampules after
filling with the aqueous TCE mixture, which shows the
adequate precision of the vial sparge analysis method
and degree of TCE homogeneity in the mixture. The vial
sparge method was not used in subsequent experiments
since there was no substantial improvement in analysis
precision over the direct GC injection or headspace
methods and because the vial sparge method was time-
consuming due to the manual sample-injection step.
TCE was introduced into each ampule in the subsequent
experiment as an aqueous solution that was prepared
by mixing neat TCE with a 2 L volume of water for
24 hours at room temperature.
The concentration of TCE increased in the ampules
with anoxic water from 58.2 to 62.4 mg/L after 10 days
at 120°C while it decreased from 87.9 to 62.5 mg/L in
ampules that contained Ottawa sand and anoxic water
(Table 4.10); however, these changes are in comparison
to the initial TCE concentration which, as described
above, did not represent the true initial condition due
to non-equilibrium between neat TCE and water. In
ampules with oxic water (Table 4.11), there was a
significant decrease in TCE concentration (P-value=
0.01, alpha=0.05) from 72.5 to 57.2 mg/L after 10 days
at 120°C and from 69.9 to 58.3 mg/L in ampules with
sand and oxic water, which may have indicated that
TCE was being degraded. However, there was also a
decrease in TCE concentration in the matching ampules
stored at 22°C where TCE was not expected to degrade
over 10 days. No statistical comparison between the
22°C and 120°C results was possible as only one ampule
was stored at 22°C for each experimental variable.
Thus the variability (the basis for statistical tests) of the
TCE concentration in the ampules stored at 22°C was
unknown. A matching number of 22°C control ampules
were prepared in the subsequent experiment to allow for
statistical comparisons between 22°C and 120°C results.
If TCE was being degraded in the ampules incubated at
120°C, the rate of disappearance, based on the results
shown in Table 4.11, was approximately 1.5 mg/L-day,
which meant that 24 days would have been required to
degrade 50% of the initial amount of TCE, assuming a
zero-order reaction model. The subsequent experiment
was extended up to 40 days in effort to observe a greater
change in TCE concentration potentially caused by the
degradation of TCE.
Table 4.11 Initial and Final TCE Concentration in Oxic Ampules from the Third Ampule Experiment (TCE
mixed into 250 mL of water prior to addition to each ampule, Initial DO = 8.22 mg/L)
Contents
TCE, water
TCE, water,
sand
Temperature
(°C)
120
22
120
22
Initial TCE (mg/L)
Vial Sparge
72.5±6.0
73.3
69.9±3.5
71.4
Final TCE (mg/L)
Vial Sparge
57.2±3.6
69.1
58.3±3.2
58.0
Headspace
59.9±1.2
65.7
64.7±1.3
66.8
Error represents one standard deviation calculated from duplicate ampule results.
-------
The final CO2 content of the gas phase, and the aqueous
phase concentration of DO and DCAA along with the
final pH for the ampules with anoxic water are shown
in Table 4.12 and in Table 4.13 for the ampules with
oxic water. The pH of the aqueous solution decreased
from 7.73 to 6.46 in ampules with anoxic water that
contained TCE and where incubated at 120°C, while
the pH decrease was less than 0.3 pH units in the TCE-
free ampule with anoxic water (Table 4.12). A similar
decrease in the pH of ampules with oxic water was
observed with the pH decreasing from 7.80 to 6.55 for
ampules incubated at 120°C (Table 4.13). The pH also
decreased in ampules that contained both anoxic and
oxic water along with TCE and were stored at 22°C, the
pH decreased to values lower than expected based on the
results from ampules incubated at 120°C.
The decrease in pH was matched with the detection of
CO2 in the gas phase. This could either indicate that TCE
was being degraded or that CO2 was introduced during
the flame sealing process from the torch combustion
products; CO2 can dissolve in water to form carbonic
acid and thus decrease the pH. Since TCE was not
expected to degrade in the ampules stored at 22°C where
the pH decreased and CO2 was detected, it was suspected
that the ampule contents were being exposed to the torch
flame combustion-products during the flame sealing
process. To test this hypothesis, several empty 25 mL
ampules were flame sealed and CO was subsequently
Table 4. 12 Third Ampule Experiment Results for Anoxic Ampules ( 1 .4 uL of TCE added to each ampule, Initial
DO = 0.79 mg/L)
Contents
water
TCE, water
TCE, water
sand, water
TCE, water, sand
TCE, water, sand
Temperature
(°C)
120
120
22
120
120
22
Final pH
(Initial pH=7.73)
7.46
6.46±0.18
4.85
3.17
3.86
3.14±0.06
Final DO
(mg/L)
4.5
6.7±0.5
na
na
3 (n=l)
na
CO2
(uL/L)
na
838±104
2,320
2,783
3,493±144
646
DCAA (ug/L)
nd
0.7±0.6
5.4
nd
6.0±1.6
10.3
na - not analyzed
nd - below detection limit (-0.2 ug/L)
Table 4.13 Third Ampule Experiment Results for Oxic Ampules (TCE mixed into 250 mL of water prior to
addition to each ampule, Initial DO = 8.22 mg/L)
Contents
water
TCE, water
TCE, water
water, sand
TCE, water, sand
TCE, water, sand
Temperature
(°C)
120
120
22
120
120
22
Final pH
(Initial pH=7. 80)
7.63
6.55±0.07
6.45
3.09
3.25±0.03
3.94
Final DO
(mg/L)
7.7
6.7±0.6
6.9
na
2.8±0.2
6.0
C02
(uL/L)
nd
847±369
1,103
2,845
2,187±392
1,733
DCAA
(ug/L)
nd
nd
nd
nd
nd
nd
na - not analyzed
nd - below detection limit (-0.2 ug/L)
-------
detected in the ampules after destructive sampling.
Thus the CO2 detected in the ampules may have been
introduced during the flame sealing process and not from
the degradation of TCE. This finding prompted the use
of "funnel top" ampules in all subsequent experiments,
which can be flame sealed without exposing the ampule
contents directly to the torch flame.
The pH of the water in the ampules that contained
sand decreased from approximately 7.7 to less than 4.0
regardless of TCE content, initial DO concentration,
or incubation temperature. It was suspected that the
20-30 mesh Ottawa sand, which had been treated by
soaking in a 0.5 N nitric acid solution and then rinsed
with DI water, contained residual nitric acid due to an
inadequate DI rinse. A separate test was performed
by placing 100 grams of the acid treated Ottawa sand
in a beaker and adding enough freshly dispensed
DI-Nanopure water to cover the sand. The pH of the
DI water decreased from 7.7 to 4.2 indicating that
the DI rinse after the 0.5 N nitric acid treatment was
not sufficient to remove the residual nitric acid. This
result prompted additional rinsing procedures for the
subsequent experiment that included rinsing the sand
until the rinse water was at pH 7.
The chloride content of the water samples was
determined using an ion selective electrode (ISE).
However, the ISE failed to yield consistent values for the
1 mM chloride solution used to check probe response
during measurements. As a result, chloride data are
not reported here, and alternative analytical methods
that included a colorimetric technique by Bergmann
and Sanik (1957) and use of Ion Chromatography
as described in Section A. 5 were employed in the
subsequent experiment. The DO concentration in
ampules filled with anoxic water increased from 0.79 to
greater than 4.5 (Table 4.12). The DO concentration was
determined using the YSI voltammetric sensor that was
used in the first and second ampule experiments, which
required exposing the sample to air followed by vigorous
stirring thus potentially introducing oxygen into the
samples. To minimize exposure to oxygen during the
measurement process, a colorimetric method (Chemets)
for measuring dissolved oxygen was adopted for the
subsequent ampule experiment.
4.3.2 Results of Ampule Experiment 4
The following sections describe results obtained
from 38 ampules that were incubated at 120°C in the
convection oven prior to the oven explosion (Table 4.4),
51 ampules incubated at 120°C in the explosion resistant
apparatus (Table 4.6), and 67 ampules that were stored
at 22°C (Table 4.5). No results were obtained from the
convection oven incubation of the pH 10 or the solids-
filled ampules (Batch 5, 6 and 7) because these ampules
were destroyed in the oven explosion.
The dissolved oxygen data collected from these ampules
showed evidence of problems with the analysis.
The data could not be used to draw conclusions on
TCE reactions, but is presented in Appendix B for
completeness.
4.3.2.1 Change in TCE Content
The average amount of TCE in the 1,000 mg/L
(nominal) ampules with time is shown in Table 4.14.
The amount of TCE in each ampule was calculated
based on the aqueous phase TCE concentration and
the estimated gas phase TCE concentration. The gas
phase TCE concentration was estimated using Henry's
law (C = HCw), assuming equilibrium conditions,
with Cw equal to the TCE aqueous phase concentration
as determined by GC analysis, and the dimensionless
Henry's Law constant (H) equal to 0.318 at 25°C
(Staudinger and Roberts, 1996). The inclusion of TCE
in the gas phase accounts for differences in gas phase
volume between ampules, which typically represented
less than 15% of the total TCE content. These estimates
were performed to correct for the differences in the
volume of gas (headspace) between ampules that were
due to slight differences in the volume of water initially
added to each ampule. The amount of TCE (i.e., micro
moles or umol) in each ampule was then calculated
based on the concentration of TCE and the volume of
water and volume of gas in each ampule. The volume of
water in each ampule was determined by the difference
between the weight of the sealed ampule and the weight
of the empty ampule and neck after destructive sampling
along with a density of water equal to 0.997 g/mL. The
volume of gas in each ampule was estimated based on
the length of gas filled space and the ampule diameter
that was equal to 3 cm. The amount of TCE in the
ampules shown in Table 4.14 is the average along with
the standard deviation (i.e., average±S.D.) calculated
from the number (n) ampules at each time period.
There was more TCE in the 1,000 mg/L ampules
incubated at 120°C than stored at 22°C (Table 4.14),
with the exception of the oxic 20-day ampules. However,
a paired t-test performed on these data suggest that the
difference between the average amount of TCE in the
1,000 mg/L ampules incubated at 120°C and stored at
22°C was not significant (P-Values>0.22, alpha=0.05).
Thus there was no apparent change in the TCE content
for the 1,000 mg/L ampules. The one exception was
Ampule 58 from Batch 2, which had a 50.9% reduction
in the amount of TCE after 20 days at 120°C. The result
from Ampule 58 was not used in calculating the 20 day
average shown in Table 4.14.
The average amount of TCE in the 100 mg/L (nominal)
anoxic (Batch 4) and oxic (Batch 3) ampules, based
on the amount detected in the aqueous phase and that
estimated in the gas phase, is shown in Figure 4.4 as a
-------
Table 4.14 Amount (Aqueous and Gas Phases) of TCE in 1,000 mg/L (nominal) Anoxic (Batch 1) and Oxic
(Batch 2) Ampules
Content
Anoxic
Initial TCE (umol) =
368±8
Oxic
Initial TCE (umol) =
350±17
Day
10
20
30
10
20
30
Total TCE (umol)
120°C
202±49*
337±22
Lost
355±14
337±26
Lost
n
o
6
2f
3
2*
22°C
175±49*
276±55
317±12
338±2
343±22
342±11
n
3
3
3
3
3
o
6
P-Value
(120 vs. 22°C)
0.22
0.09
0.09
0.40
n - Number of ampules used to calculate average and standard deviation.
* TCE by direct GC injection.
Lost -Ampules destroyed in oven explosion.
f Ampule 7 contents lost during opening.
t Ampule 58 results excluded due to significant difference in TCE content.
function of time. There was little change in the amount
of TCE in the anoxic ampules (Batch 4) incubated at
120°C or stored at 22°C over the 41 day period. There
was no statistical difference (P-Values>0.05, Table 4.15)
between the amount of TCE in the anoxic ampules
incubated at 120°C and stored at 22°C over 41 days,
which supports the claim that there was little change in
the amount of TCE in the anoxic ampules.
o
2.
LU
O
—o- anoxic 22'C
—•- anoxic 120°C
—o— oxic22°C
-*— oxic120-C
20 30
Time (days)
Figure 4.4
Amount of TCE in the 100 mg/L anoxic
(Batch 4) and oxic (Batch 3) ampules
stored at 22°C and incubated at 120°C.
In contrast, there was a decrease in the amount of TCE in
the oxic ampules that contained 100 mg/L of TCE after
10, 20, and 30 days at 22°C and 120°C (P-Value<0.05)
as compared to the initial amount of TCE in Batch 3.
There was less TCE in the oxic ampules incubated at
120°C than was present in the ampules stored at room
temperature after 10 and 20 days (P-Values<0.05,
Table 4.15). These results suggest that the presence
of oxygen affected the amount of TCE found in the
ampules incubated at 120°C and stored at 22°C.
The average amount of TCE decreased in the anoxic
ampules that contained solids (Batches 6 and 7) as a
function of time as shown in Figure 4.5. The amount
of TCE detected in the ampules that contained either
Ottawa sand alone or Ottawa sand+1% goethite and were
incubated at 120°C was significantly less (P-Value>0.05)
than the amount in the corresponding ampules stored
at 22°C (Table 4.16). There was 22.8% less TCE in the
ampules with sand after 30 days at 120°C with 20.8%
less TCE in the ampules with sand+l%goethite after
4 days at 120°C as compared to the amount in the
ampules stored at 22°C. This result indicated that the
rate of TCE disappearance was greater in the ampules
that contained Ottawa sand+1% goethite as compared to
ampules with Ottawa sand alone.
-------
Table 4.15 Amount (Aqueous and Gas Phases) of TCE in 100 mg/L (nominal) Anoxic (Batch 4) and Oxic
(Batch 3) Ampules
Content
Anoxic
Initial TCE (umol) =
31.4±2.6*
Oxic
Initial TCE (umol) =
39.0±1.0
Day
10
30f
41f
10
20
30f
Total TCE (umol)
120°C
30.5±2.3
31.3±2.8
30.9±1.7
34.8±2.5
32.5±0.3
31.3±2.5
n
3
2
2
3
o
5
2
22°C
32.8±1.6
33.2
32.5
37.3±1.0
35.9±0.6
35.3
n
3
1
1
3
o
6
i
P-Value
(120 vs. 22°C)
0.12
0.34
0.29
0.06
0.01
0.21
n - Number of ampules used to calculate average and standard deviation.
* TCE by direct GC injection.
f Incubated using the explosion resistant apparatus, all other ampules incubated in convection oven.
Q.
ra
o
_=_
m
o
Sand 22'C
Sand120°C
Sand+Goethite 22'C
Sand+Goethite 120'C
mean TCE 22'C
0 10 20 30 40 50
Time (days)
Figure 4.5 Amount of TCE in anoxic 100 mg/L
ampules with Ottawa sand (Batch 6) and
Ottawa sand+1% goethite (Batch 7) stored
at 22°C and incubated at 120°C.
The average amount of TCE in the anoxic ampules that
contained water amended with NaOH to achieve a pH of
10 (Batch 5) decreased as a function of time as shown
in Figure 4.6. There was 10.8% less TCE in the pH 10
ampules after 4 days at 120°C compared to the amount
of TCE in the 22°C ampules (Table 4.17).
Table 4.16 Amount (Aqueous and Gas Phases) of TCE in 100 mg/L (nominal) Anoxic Ampules with Sand (Batch
6) and Sand+1% Goethite (Batch 7)
Content
Sand
Initial TCE (umol)
= 30.6±0.9
Sand + 1%
Goethite
Initial TCE (umol)
= 30.6±0.9
Day
10
20
30
1
2
o
J
4
Total TCE (umol)
120°C
22.5±0.4
17.3±0.5
19.6±2.2
25.4±1.7
20.6±1.9
19.8±2.3
n
2
2
3
2
2
2
22°C
25.4±1.0
25.0±0.5
n
2
2
P-Value
(120 vs. 22°C)
0.04
0.04
0.01
0.35
0.07
0.07
n - Number of ampules used to calculate average and standard deviation.
-------
Table 4.17 Amount (Aqueous and Gas Phases) of TCE in 100 mg/L (nominal) Anoxic Ampules Amended with
NaOH (0.26 mM) to pH 10 (Batch 5)
Content
NaOH
Initial TCE (umol) =
35.3±1.0
Day
1
2
3
4
Total TCE (umol)
120°C
28.5±0.7
26.8±1.7
24.4±4.1
28.0
n
2
2
2
1
22°C
31.4±1.9
n
2
P-Value
(120 vs. 22°C)
0.09
0.01
0.07
n - Number of ampules used to calculate average and standard deviation.
Time (days)
Figure 4.6 Amount of TCE in 100 mg/L anoxic
ampules amended with NaOH (0.26 mM)
to pH 10 (Batch 5) stored at 22°C and
incubated at 120°C.
4.3.2.2 Change in pH
The pH of ampules was expected to decrease with
the degradation of TCE. Assuming that all the TCE
initially present was transformed into CO2 and chloride,
then the pH in ampules would have decreased from
7 to approximately 2 based on the initial amount of
TCE present in the 1,000 mg/L ampules and from 7 to
3 in ampules with 100 mg/L. The aqueous phase pH
of anoxic (Batch 1) and oxic (Batch 2) ampules with
1,000 mg/L (nominal) of TCE are shown in Figure 4.7.
The pH decreased in both the anoxic and oxic ampules
from 7.0 to 4.34±0.08 (n=2) and 3.65±0.19 (n=2),
respectively, after 20 days at 120°C. The pH of the oxic
1,000 mg/L ampules stored at 22°C also decreased,
from 7.0 to 4.74±0.35 (n=3) after 20 days but did not
substantially decrease after an additional 10 day period.
In contrast, the pH of the anoxic 1,000 mg/L ampules
stored at 22°C increased from 7.0 to 7.43±0.82 (n=3)
after 20 days (Figure 4.7). The pH of the TCE-free
anoxic and oxic ampules (controls) increased from 7.0
to 8.45±0.16 (n=12), this average was calculated from
the pH of TCE-free ampules after 10, 20, and 30 days at
22°C and 120°C. The pH measurements for the TCE-free
controls were from 12 of the 24 ampules that were
initially prepared in Batch 1 and 2 (Table 4.4). The pH
increase in TCE-free ampules was thought to represent
the dissolution of SiO2 from the ampule glass walls as
discussed in Section 4.3, whereas the pH decrease in
TCE containing ampules was thought to represent the
release of hydrogen atoms from TCE molecules due to
the degradation of TCE.
-O- Anoxic 22'C
-•- Anoxic 120'C
Oxic 22'C
Oxic 120°C
Control {TCE Free, n=12)
10 20
Time (days)
Figure 4.7
The pH of anoxic (Batch 1) and oxic
(Batch 2) ampules with 1,000 mg/L of TCE
and TCE-free controls stored at 22°C and
incubated at 120°C.
The aqueous phase pH of anoxic (Batch 4) and oxic
(Batch 3) ampules with 100 mg/L (nominal) of TCE are
shown in Figure 4.8. The pH of the anoxic (Batch 4)
ampules incubated at 120°C decreased from 7.0 to
4.69±0.04 (n=3) after 10 days, to 6.60±0.04 (n=2)
after 30 days, and to 6.36±0.08 (n=2) after 40 days.
It should be noted that the 30 and 41 day results for
-------
the anoxic 100 mg/L ampules were obtained using the
explosion resistant incubation apparatus, while the 10
day results were obtained from the convection oven prior
to the explosion. There was a difference in the ampule
temperature profile between heating methods with the
ampule necks exposed to ambient temperature (22°C)
in the explosion resistant apparatus (see Figure 4.2)
while the entire ampule was exposed to 120°C in the
convection oven.
6 — Anoxic 22'C
• — Anoxic 120'C
Oxic22"C
Oxlc120'C
Control (TCE-Free, n=26)
20 30 40
Time (days)
Figure 4.8 The pH of anoxic (Batch 4) and oxic
(Batch 3) ampules with 100 mg/L of TCE
and TCE-free controls stored at 22°C and
incubated at 120°C.
In the oxic 100 mg/L ampules (Batch 3), the pH
decreased from 7.0 to 5.42±0.28 (n=3) after 10 days, to
4.95±0.04 (n=3) after 20 days, and to 6.66±0.40 (n=2)
after 30 days, with the 10 and 20 day results obtained
from the convection oven and the 30 day results
obtained from the explosion resistant apparatus. The
pH of the 100 mg/L anoxic and oxic ampules stored
at room temperature (22°C) increased to 8.0±0.32
(n=3) and 7.75±0.23 (n=3), respectively, after 10 days
but then decreased to 6.36 (n=l) and 7.34 (n=l),
respectively, after 30 days at 22°C. The average pH of
the anoxic and oxic TCE-free control ampules that were
stored at 22°C and incubated at 120°C was 8.00±0.70
(n=26) over 41 days which shows that the pH did not
substantially change in ampules without TCE. The pH
measurements for the TCE-free controls were from 26 of
the 36 ampules that were initially prepared in Batches 3
and 4 (Table 4.4).
The pH of anoxic ampules that contained acid washed,
20-30 mesh Ottawa sand (Batch 6) are shown in
Figure 4.9. After 40 days at 120°C, the pH decreased
in ampules with TCE and TCE-free ampules, from
an initial value of 7.00 down to 2.96±0.05 (n=3)
and 3.21±0.13 (n=3), respectively. The lower pH in
the ampules with TCE was significantly different
(P-Value=0.01, alpha=0.05) than the pH in the TCE-
free ampules indicating that TCE was being degraded.
However, the Ottawa sand also served as a source of
hydrogen atoms since the pH of the TCE-free control
ampules was less than 4 after 10, 20, and 40 days. The
pH decrease was not believed to be from the nitric acid
used during sand pre-treatment since the sand was rinsed
with DI-Nanopure water after the acid wash until the
rinse water in contact with sand yielded a pH of 7.
-O- Sand22°C
-•- Sand120"C
Control CTCE-Free, n=9)
Time (days)
Figure 4.9 The pH of anoxic ampules that contained
acid-washed Ottawa sand with 100 mg/L
of TCE (Batch 6) and TCE-free controls
stored at 22°C and incubated at 120°C.
One explanation for the pH decrease in the ampules with
sand is related to the detection of sulfate ions (SO42~) in
the ampule aqueous phase, which led to the hypothesis
that the Ottawa sand contained pyrite (FeS2). Pyrite
is known to dissolve in water and form sulfate and
increase water acidity with 16 H+ produced for every
molecule of pyrite (Stumm and Morgan, 1996, p. 691),
which could have caused the observed decrease in pH.
A sample of the acid treated Ottawa sand was sent to
the U.S. Silica Co. Corporate Laboratory in Berkeley
Springs, West Virginia, where the sand was analyzed by
X-ray diffraction analysis (XRD). The mineral phases
identified in the sand included pyrite, marcasite (FeS2
- polymorph of pyrite), and hematite (Fe2O3), however,
no estimate of the relative amount of the minerals (i.e.,
mg/kg) was provided. Ottawa sand was analyzed for
20 elements using ICP-MS (Thermo Jarrell-Ash, Enviro
36) after a double acid digest using a dilute HC1 and
H2SO4 solution by the Chemical Analysis Laboratory,
The University of Georgia, Athens, GA. Silica was the
predominant element in Ottawa sand and iron was the
major component in Ottawa sand amended with 1%
goethite(Table4.18).
-------
Table 4.18 Elements Present in Ottawa Sand and
Ottawa sand+1% Goethite
Element
Silica
Calcium
Iron
Zinc
Sodium
Aluminum
Nickel
Magnesium
Potassium
Ottawa
Sand
(mg/kg)
4.54
3.96
3.23
2.33
2.27
2.10
1.75
1.10
0.89
Ottawa
Sand+l%Goethite
(mg/kg)
11.31
32.25
119.4
4.51
4.64
3.61
0.14
2.40
2.46
Assuming all the iron in the Ottawa sand represented
FeS2 and that all pyrite present was oxidized to sulfate,
then the pH in the ampules would be expected to
decrease from 7 to 3.34 in ampules with Ottawa sand
based on the following calculations.
3.23 mg
Fe
kg
20 g
sand
ampule
16 H+
Fe
kg
1,000 g
mmol
Fe
55.84
mg
ampule
40mL
water
1,000
mL =0.029
T mMFe
0.029 mM
Fe n
u
.46 mM H+
ampule
pH = -log[0.46xlO-3 M + IxlO'7 M] = 3.34
Transformation of the sulfur in pyrite and marcasite
into sulfate requires a source of dissolved oxygen or
ferric iron (Fe3+) (Stumm and Morgan, 1996). While no
measurement for ferric iron was performed, the hematite
identified in the Ottawa sand may have served as a
source of ferric iron and thus accounts for the detection
of sulfate in the anoxic ampules.
Another source of acidity was the CO2 found after
incubating ampules amended with Ottawa sand
(Section 4.3.2.4). Assuming that gas phase CO2 was in
equilibrium with the ampule water at 25°C and using
a value of 0.0339 M/atm for the Henry's coefficient
between CO2(g) and dissolved phase CO2(aq), which
is assumed to transform into carbonic acid (H2CO3),
results in a decrease from pH 7 to approximately 4 in the
ampules with Ottawa sand. Thus there are at least two
potential sources of acidity in the Ottawa sand that could
account for the observed decrease in ampule pH.
The pH of anoxic ampules that contained acid-washed
Ottawa sand+1% goethite (Batch 7) is shown in
Figure 4.10. The pH decreased from 7.00 to 3.64±0.05
(n=2) in ampules with 100 mg/L of TCE after 4 days
at 120°C and to 4.44±0.18 after 3 days at 22°C. The
pH in TCE-free ampules stored at 22°C and incubated
at 120°C decreased to 4.79±0.35 (n=8). The pH of the
anoxic ampules that contained water amended with
NaOH to achieve an initial pH of 10 (Batch 5) is shown
in Figure 4.11. At 120°C, the pH of NaOH-amended
ampules that contained 100 mg/L of TCE decreased from
10.0 to 7.06 (n=l) after 4 days. In contrast, the pH of the
TCE-free control ampules remained at 9.03±0.11 (n=9)
and the ampules with 100 mg/L TCE were at pH of
9.37±0.04 (n=2) after 4 days at 22°C (Figure 4.11).
O Sand+Goethite 22°C
-•- Sand+Goethite 120'C
Control CTCE-Free, n=8)
Time (days)
Figure 4.10 The pH of anoxic ampules that contained
acid-washed Ottawa sand+l%goethite with
100 mg/L of TCE (Batch 7) and TCE-free
controls stored at 22°C and incubated at
120°C.
-------
Q.
- pH1022'C
- pH10 120-C
- Control (TCE-Free, n=8)
TCE-Free Control
Time (days)
Figure 4.11 The pH of anoxic ampules amended with
NaOH (0.26 mM) to pH 10 and 100 mg/L
of TCE (Batch 5) along with TCE-free
controls stored at 22°C and incubated at
120°C.
4.3.2.3 CO and CO2 in the Gas Phase
A 10 mL gas sample was collected from each ampule
using a gas-tight syringe and approximately 6 mL of
the gas sample was analyzed by a GC/TCD equipped
with a Carboxen 1010 capillary column capable of
separating CO and CO2 from the ampule gas samples.
Carbon monoxide (CO) and CO2 were anticipated TCE
degradation products based on the past experimental
work presented in Section 2.3.2. The amount (umol)
of CO and CO2 in each ampule was calculated based
on the gas-phase concentration determined for CO and
CO2, and the estimated volume of headspace gas in each
flame-sealed ampule. The amount of CO2 reported does
not include the amount of CO2 dissolved in the aqueous
phase, which was expected to be substantial given the
Henry's law coefficient of 0.83 M(aq)/M(g) for 25°C
(Stumm and Morgan, 1996, p. 214). Using the average
gas volume of 20 mL and liquid volume of 45 mL for
the ampules shows that the amount of dissolved CO2
is expected to be 1.7 times greater than the amount
in the gas phase. The additional amount of dissolved
CO2 was incorporated in calculating the rate of TCE
degradation based on the rate of carbon products formed
(Table 4.22).
The amounts of CO and CO2 in ampules that initially
contained 1,000 mg/L of TCE (Batch 1 and 2) and
were incubated at 120°C are shown in Figure 4.12.
The CO and CO2 content of the ampules increased
during the 120°C incubation period, with nearly
identical amounts of CO and CO0 detected in all of the
ampules after 10 days regardless of the initial oxygen
content (anoxic or oxic). After 20 days of incubation at
120°C, the amount of CO and CO2 detected in the oxic
ampules nearly doubled, while the amount detected
in the anoxic ampules remained similar to the 10-day
incubation values. The average amount of CO and CO2
shown in Figure 4.12 excludes the amount detected
in Ampules 9, 55, and 58, which were incubated at
120°C. In the case of Ampule 9, an anoxic ampule with
1,000 mg/L of TCE from Batch 1, there was a significant
difference (P-Value=0.02, alpha=0.05) between
the amount of CO and CO2 detected after 20 days
(CO+CO2 = 12,386 ppmv) compared to the amount
detected in the companion replicate ampules, Ampules
7 and 8 (2,057±387 ppmv). There was a significant
difference (P-Value=0.06, alpha=0.05) between the
amount of CO and CO2 detected after 10 days in Ampule
55 (4,250 ppmv) and in replicate Ampules 53 and 54
(1,983±419 ppmv), oxic ampules with 1,000 mg/L of
TCE from Batch 2. In this case, Ampule 55 had to be
flame sealed a second time during preparation thus
the ampule contents may have been exposed to the
propane-oxygen flame. The greatest difference between
replicate Batch 2 ampules was obtained after 20 days
for Ampule 58 (123,143 ppmv) and Ampules 56 and 57
(4,515±304 ppmv).
3 -
2
1 '
n -
-r
ffili
m
T
-r |--|
I
,„
\
P
6
6
^
1
I 1 CO Anoxic
E25S2 CO Oxic
Rgssasa COj Anoxic
K^^JI CO2 Oxic
I
\
I
1
!
Time (days)
Figure 4.12 Amounts of CO and CO2 in anoxic
(Batch 1) and oxic (Batch 2) ampules
with 1,000 mg/L of TCE and incubated at
120°C.
The amounts of CO and CO2 detected in anoxic
(Batch 4) and oxic (Batch 3) ampules that initially
contained 100 mg/L of TCE and were incubated at
120°C for up to 41 days are shown in Figure 4.13.
Carbon monoxide (CO) and CO2 were not detected
in ampules stored at 22°C that contained 100 mg/L
of TCE and were without TCE (TCE-free). Carbon
-------
monoxide (CO) was detected in ampules that contained
TCE and were incubated at 120°C using the convection
oven, whereas CO wasn't detected in ampules using
the explosion resistant apparatus until 41 days at
120°C. This result may indicate that the difference in
temperature profile between the two incubation systems
led to differences in TCE degradation rate.
CO 120"C
CO, 120'C
C02 without TCE 120'C
0.8 -
0.6 -
0.4 -
0.2 -
nn -
"
.
p
-r
i i
i EI
1 1
i ^
||
1
\ \ CO Anoxic
V/WA CO Oxic
W5h CO2 Anoxic
RWI CCj Oxic
|
|
i
§
1 ffi
i n
10 20 30 40
Time (days)
Figure 4.13 Amounts of CO and CO2 in anoxic
(Batch 4) and oxic (Batch 3) ampules with
100 mg/L of TCE and incubated at 120°C.
No CO or CO2 was detected in ampules
stored at 22°C.
The amounts of CO and CO2 detected in anoxic ampules
that contained 20-30 mesh Ottawa sand (Batch 6) are
shown in Figure 4.14. The average amount of CO
detected (0.97±0.08 umol) in ampules that contained
TCE and were incubated at 120°C over a 40-day period
was approximately twice the amount detected in the
ampules stored at 22°C for 40 days. Carbon monoxide
(CO) was not detected in ampules without TCE that were
stored at 22°C and incubated at 120°C. For the ampules
incubated at 120°C, the amount of CO2 increased with
incubation time, with amount of CO2 detected in ampules
with TCE in excess of the amount of CO2 detected in
ampules without TCE.
Time (days)
Figure 4.14 Amounts of CO and CO2 in anoxic ampules
with Ottawa sand and 100 mg/L of TCE
(Batch 6) stored at 22°C and incubated at
120°C.
The amounts of CO and CO2 detected in anoxic ampules
that contained 20-30 mesh Ottawa sand + 1% goethite
(Batch 7) are shown in Figure 4.15. The amount of
CO increased during incubation over 4 days at 120°C.
In contrast, no CO was detected in TCE-free ampules
that were stored at 22°C. At 120°C, the amount of CO2
increased with incubation time in excess of the amount
of CO2 detected in the TCE-free ampules. Relative the
data shown in Figure 4.14, the amount of CO2 generated
in the presence of 1% goethite was greater than
detected in ampules that contained Ottawa sand alone.
Figure 4.16 shows the amount of CO and CO2 detected
in the ampules that contained 0.26 mM of NaOH. The
amount of CO increased over the initial 2 days of the
incubation period at 120°C and then decreased over the
last 2 days. No CO was detected in the ampules without
TCE at 120°C and with TCE that were stored at 22°C.
The amount of CO2 increased with incubation time until
reaching a plateau of approximately 0.8 umol on day 3
and 4.
-------
4
3
2
1 -
I 1 CO 120'C
VZVA CO2 120'C
^m CO2 without TCE 120-C
jjj
\
n\
\
\
CO. without TCE 22QC £
C0222'C | I
\
^
T ^
\\
\i
\%
\ !
I i
ll I
I
i
i
i
i
=
=
=
=
=
» >
I si
I
=
E
i
t
;
;
t
I
:
^1
2 3
Time (days)
Figure 4.15 Amounts of CO and CO2 in anoxic
ampules with Ottawa sand+l%goethite and
100 mg/L of TCE (Batch 7) stored at 22°C
and incubated at 120°C.
0.8 -
0.6-
0.4-
0.2-
nn -
1 1 COwithTCE120°C
WJVJi CO2with TCE 120"C
Ka
1>
p
i
i
^\
-r-
%
Y
<&
/•
<6
1
|
i
i
^
|
^
g
1
i
>j
^
f
^
s
!
1
1234
Time (days)
Figure 4.16 Amounts of CO and CO2 in anoxic
ampules amended with NaOH (0.26 mM)
to pH 10 (Batch 5) and 100 mg/L of TCE
incubated at 120°C.
4.3.2.4 Other Gas Phase Compounds
After analyzing 6 mL of the 10 mL gas sample for
CO/CO2 content, the 10 mL syringe with 4 mL of gas
sample was removed from the GC/TCD injection loop
and the syringe needle tip was sealed with a rubber
septum. Approximately 1 mL of the gas sample was
then injected directly into the inlet of a Varian 3600CX
GC equipped with a 30 m long by 0.32 mm OD Varian
CP-Sil SMS column connected to a Varian Saturn 2000
mass spectrometer (MS). The purpose of injecting the
gas sample directly into the GC/MS was to screen for the
presence of compounds other than CO, CO2, and TCE.
This analysis step was added after the oven explosion
had occurred to determine if potentially explosive TCE
degradation products were present in the ampule gas
phase. The first gas samples analyzed by GC/MS were
from Ampules 83, 84, and 85 (Batch 3), oxic ampules
that contained 100 mg/L of TCE and had been incubated
at 120°C for 20 days in the convection oven. The gas
samples were found to contain dichloroacetylene (DCA),
CO2, and TCE. Dichloroacetylene was identified by mass
spectrum match with the NIST98 library (Figure 4.17).
Figure 4.17 Mass spectrum of the 2.05 min
chromatogram peak from the analysis of
1 mL of gas from Ampule 83.
The concentration of DCA was not determined,
as this compound is unstable and difficult to
prepare for calibration purposes. The abundance of
dichloroacetylene was less than 1% of the TCE present
(base peak). Dichloroacetylene (DCA) was not present
in the gas samples from Ampules 95, 96, and 97, the
matching 100 mg/L oxic ampules that were stored
at 22°C. Dichloroacetlylene (DCA) was consistently
identified in gas samples from ampules that were
incubated at 120°C for Batch 3 through 7, no gas
samples from Batch 1 or 2 ampules were analyzed by
GC/MS. The greatest amount of DCA was detected
in the gas phase of the pH 10 ampules (Batch 5) by
GC/MS analysis. An unidentified peak was present in the
headspace GC/FID chromatogram from the analysis of
1 mL water samples collected from the pH 10 ampules.
The identity of the GC/FID peak was assigned to DCA
based on the GC/MS results and the estimated dissolved
phase DCA concentration was 1.8±0.5 mg/L for the
pH 10 ampules incubated at 120°C, estimated using the
headspace response factor for TCE.
4.3.2.5 Aqueous Phase Compounds
The amount of chloride detected in both anoxic and
oxic ampules incubated at 120°C increased for all
-------
experimental conditions considered and in excess
of the amount measured in paired ampules stored at
22°C (Figures 4.18 through 4.22). Thus increasing
the temperature of ampules resulted in an increase in
the rate of TCE degradation since TCE was the only
source of chloride within the ampules. The presence
of oxygen appeared to have an affect on the amount of
TCE degraded in the 1,000 mg/L ampules (Batches 1
and 2) as there was a greater amount of chloride
(P-Value=0.001, alpha=0.05) in the 1,000 mg/L ampules
with oxygen after 10 and 20 days at 120°C (Figure 4.18).
However, there was no discernable difference
(P-Value=0.66, alpha=0.05) between the amount of
chloride in the anoxic (Batch 4) and oxic (Batch 3)
ampules with 100 mg/L of TCE after 10 and 30 days at
120°C (Figure 4.19). The amount of chloride detected
in the 100 mg/L oxic (Batch 3) ampules increased from
7.6±1.3 umol after 10 days to 9.9±0.6 umol after 20 days
at 120°C using the convection oven incubation but was
6.0±0.1 umol after 30 days using the explosion resistant
apparatus. There was a similar trend for chloride in the
anoxic 100 mg/L ampules with more chloride detected
after 10 days in the convection oven that after 30 and 41
days in the explosion resistant apparatus.
o 20
15
l=l Anoxic 120QC
V7Z7Z Oxic 120*C
3 Oxic 22°C
Control (TCE-fnee)
Time (days)
Figure 4.18 Amount of chloride in anoxic (Batch 1) and
oxic (Batch 2) ampules with 1,000 mg/L
of TCE stored at 22°C and incubated at
120°C.
o
12 -i
10
8
6
4
2
Convection Oven Incubation
j
u
F
1 1 Anoxic 120"C
mm oxici2o-c
Egsgss Anoxic 22aC
K^5^ Oxic 22°C
Control (TCE-free)
Explosion Resistant
Apparatus Incubation
1
L|
20 30
Time (days)
Figure 4.19 Amount of chloride in anoxic (Batch 4)
and oxic (Batch 3) ampules with 100 mg/L
of TCE stored at 22°C and incubated at
120°C.
The presence of 20-30 mesh Ottawa sand (Batch 6)
caused an increase in the amount of chloride
(Figure 4.20) in ampules incubated at 120°C compared
to the amount of chloride in the anoxic (Batch 4)
ampules that were without sand (Figure 4.19); both
batches were incubated in the explosion resistant
apparatus. There was 13.8±1.9 umol of chloride detected
in the anoxic ampules that contained 20-30 mesh Ottawa
sand after 30 days at 120°C as compared to 6.3±0.0 umol
of chloride in the ampules with anoxic water alone.
14 -
12 -
g 10 -
3
% 8-
z «
o
4 -
2
l=l Sand120°C
VJV/A Sand22°C
Control (TCE-free)
ft
fi
10 30
Time (days)
Figure 4.20 Amount of chloride in anoxic ampules
with Ottawa sand and 100 mg/L of TCE
(Batch 6) stored at 22°C and incubated at
120°C.
-------
The addition of goethite to the Ottawa sand resulted
in 6.8±1.7 umol of chloride (Figure 4.21) after 4 days
at 120°C while the amount of chloride in the pH 10
ampules was 13.8 umol (n=l) after 4 days at 120°C
(Figure 4.22).
Aqueous samples collected from the ampules were
analyzed for haloacetic acids content, including
chloroacetate (C1H2C2OO"), dichloroacetate
(CLHC2OO-), and trichloroacetate (C13C2OQ-). While
dichloroacetate (DCAA) was not detected (>2 ug/L)
in any of the ampules containing 100 mg/L of TCE
(Batches 3-7), DCAA was detected in several of the
ampules containing 1,000 mg/L of TCE (Batch 1 and 2).
In 3 of the anoxic ampules, DCAA was detected at the
following concentrations: 16 ug/L after 10 days at 22°C,
7 ug/L after 20 days at 120°C, and 6.1 ug/L after 30 days
at 22°C. The one exception was the 125 ug/L of DCAA
detected in Ampule 9, the 1,000 mg/L anoxic ampule
that also had more CO and CO2 than in its companion
replicate ampules. DCAA was detected in all of the oxic
ampules with 1,000 mg/L of TCE (Batch 2), with the
largest DCAA concentration detected in ampules that
were stored at 22°C (Figure 4.23). DCAA was detected
in the ampules that contained solids (Batch 6 and 7),
but at relatively low concentrations (i.e., < 5 ug/L).
Chloroacetate was detected in pH 10 ampules that
were incubated at 120°C (Figure 4.24), whereas the
chloroacetate concentration in the pH 10 ampules stored
at 22°C were less that 5 ug/L.
7 -
6-
"5 5_
^
^ 3-
O
2 -
1 -
0-
I I Sand+Goethte 120'C
vs/srA Sand+Goethite 22"C
Control (TCE-free)
I
IT
rh
1234
Time (days)
o
0
T3
O
18
14 •
12
10
8
6
4
5"
rh
I I pH10 120°C
^^ pH1022'C
Control (TCE-free)
lif
ii]
Time (days)
Figure 4.22 Amount of chloride in anoxic ampules
amended with NaOH (0.26 mM) to pH 10
with 100 mg/L of TCE (Batch 5) stored at
22°C and incubated at 120°C.
300 -
200 -
100 -
I 1 Anoxic 120°C
E?ZZI Anoxic 22-C
^m Oxic 120'C
RXita Oxic22°C
10
I I
20
Time (day
30
)
Figure 4.23 Concentration of dichloroacetate (DCAA)
in the anoxic (Batch 1) and oxic (Batch 2)
ampules with 1,000 mg/L of TCE stored at
22°C and incubated at 120°C.
Figure 4.21 Amount of chloride in anoxic ampules with £
Ottawa sand+l%goethite and 100 mg/L °
TCE (Batch 7) stored at 22°C and
incubated at 120°C.
ft
i
23
Time (days)
Figure 4.24 Concentration of chloroacetate in ampules
amended with NaOH (0.26 mM) to pH 10
and 100 mg/L TCE (Batch 5) incubated at
120°C.
-------
The aqueous solution that remained after the haloacetic
acid and chloride analysis (~5 mL) was analyzed by
Ion Chromatography (1C) to determine if other anionic
species were present. The 1C chromatogram contained
three elution peaks with retention times of 3.35, 3.73,
and 4.75 minutes. The 4.75 minute peak was attributed
to chloride based on comparison with the elution time
for a 1 mM chloride standard solution. The 3.35 and
3.73 minute peaks were attributed to formate (CHOO")
and to glycolate (HOH2C2OO"), respectively, based on
comparison with elution times for a solution containing
1 mM of glycolate, acetate, and formate. Glycolate and
formate were detected in all ampules that contained
1,000 mg/L of TCE (Batch 1 and 2) and were incubated
at 120°C (Figure 4.25). The average values shown in
Figure 4.25 do not include the results from Ampules
9, 55, and 58 as the results from these ampules were
significantly different from the other ampules as
discussed in Section 4.6.4. The amount of formate
increased with time at 120°C in the oxic and anoxic
ampules that contained 1,000 mg/L of TCE while the
amount of glycolate decreased with incubation time.
There was more glycolate than formate after 10 days
in all the 1,000 mg/L ampules, while there was more
formate than glycolate after 20 days. Glycolate and
formate were not detected in ampules stored at 22°C that
contained 1,000 mg/L of TCE or in the TCE-free control
ampules.
10 20
Time (days)
Figure 4.25 Amount of glycolate and formate in anoxic
(Batch 1) and oxic (Batch 2) ampules with
1,000 mg/L of TCE incubated at 120°C.
Glycolate and formate were detected in both the anoxic
(Batch 4) and oxic (Batch 3) ampules that contained
100 mg/L of TCE and were incubated at 120°C
(Figure 4.26). Formate was not detected in either the
anoxic or oxic ampules with 100 mg/L of TCE after
30 days at 120°C. However, it should be noted that
the 30 and 40 day results for the anoxic ampules were
obtained using the explosion resistant incubation
apparatus, whereas the 10 and 20 day results for the
oxic ampules were obtained using the convection oven.
Glycolate and formate were not detected in the anoxic
and oxic ampules that contained 100 mg/L of TCE and
were stored at 22°C or in the TCE-free control ampules.
The amount of formate detected in the anoxic ampules
that contained 20-30 mesh Ottawa sand and 100 mg/L
of TCE (Batch 6) decreased with time at 120°C, while
the amount of glycolate increased over the 40-day
incubation period (Figure 4.27). Formate was the only
organic acid detected in the ampules that contained
Ottawa sand + 1% goethite (Batch 7) and the amount of
formate increased over the 4-day incubation period at
120°C (Figure 4.28).
o
Convection Oven Incubation
Explosion Resistant
Incubation Apparatus
Time (days)
Figure 4.26 Amount of glycolate and formate in anoxic
(Batch 4) and oxic (Batch 3) ampules with
100 mg/L of TCE incubated at 120°C.
3 2
'',
';
''-. H
1 1 Glycolate Sand 120'C
^^ Formate Sand 120'C
m Formate Sand without TCE 120'C
ft
I I
Formate Sar
I I
\ n
1 n
^ n
[
I
ll
IE
d22'C
Time (days)
Figure 4.27 Amount of glycolate and formate in anoxic
ampules with Ottawa sand and 100 mg/L
of TCE (Batch 6) stored at 22°C and
incubated at 120°C.
-------
o
1 1 Formate Sand+Goethite 120°C
^ Formate Sand+Goethite without TCE 120°C
r
f
Formate Sand+Goethite 22°C
l
m
I
J
i e
1
] Glycolate pH10120"C
Time (days)
Figure 4.28 Amount of formate in anoxic ampules with
Ottawa sand+l%goethite and 100 mg/L
of TCE (Batch 7) stored at 22°C and
incubated at 120°C.
2 3
Times (days)
Figure 4.30 Amount of glycolate in anoxic ampules
amended with NaOH (0.26 mM) to pH 10
and with 100 mg/L of TCE (Batch 5)
incubated at 120°C.
Sulfate (SO42") was an additional anion detected in
ampules that contained solids. The amount of sulfate
detected in anoxic ampules with 20-30 mesh Ottawa
sand that were stored at 22°C and incubated at 120°C
is shown in Figure 4.29. Sulfate was present at aqueous
phase concentrations of greater than 1 mM and is
hypothesized to have been formed from the thermal
decomposition of pyrite and marcasite (FeS2), which
were found to be present in the 20-30 mesh Ottawa sand
(Dr. Matt Paige, facsimile communication, 30 August
2004) as discussed in Section 4.6.2. Similar amounts of
sulfate were detected in ampules that contained goethite
in addition to 20-30 mesh Ottawa sand (data not shown).
Glycolate was the only organic acid detected in the
100 mg/L ampules with NaOH (Batch 5) that were
incubated at 120°C as shown in Figure 4.30.
o
40 -
30 -
20
10 -
I I Sulfate Sand 120°C
V?m Sulfate Sand without TCE 120°C
I
I
}
\
\
*%
I
\
'fy
\
f
F
\
\
2
Sulfate
at22°C
/
I
\ /
10 30 40
Time (days)
Figure 4.29 Amount of sulfate in anoxic ampules
with Ottawa sand and 100 mg/L of TCE
(Batch 6) stored at 22°C and incubated at
120°C.
-------
4.3.2.6 Mass Balance
A balance between the loss of TCE and the moles of
carbon and chloride detected in ampules incubated at
120°C is provided in Table 4.19. The moles of TCE
lost were determined by subtracting the moles of TCE
detected after each 120°C incubation period from the
initial moles of TCE as determined immediately prior to
flame sealing the ampules. The moles of carbon lost were
calculated as twice the moles of TCE lost (2xTCE lost)
and the moles of chlorine lost as three times the moles
of TCE lost (3xTCE lost). The mass balance is a unitless
ratio of the moles of TCE lost divided by the moles
of degradation products detected with a value of 1
indicating an ideal balance between what was lost and
what was found. A value of less than zero corresponds to
a gain in the moles of TCE present in the ampule.
The results shown in Table 4.19 indicate that more TCE
was lost than could be accounted for by the moles of
carbon detected in the gas phase as CO and CO2, and in
the aqueous phase as formate, glycolate, and haloacetic
acids (i.e., mass balance > 1). There was also more
TCE lost than could be accounted for by the moles of
chlorine detected in the aqueous phase as chloride. The
failure to achieve mass balance could be attributed to
the following explanations: 1) more organic compounds
were present than were detected, 2) experimental error
occurred during the analysis of TCE, and/or degradation
products, or 3) losses occurred during the sample
collection process, in particular volatile degradation
products.
Table 4.19 Mass Balance Between Carbon and Chloride Lost as TCE and Detected as Degradation Products in
Ampules Incubated at 120°C
Ampule Content
(Batch No.)
Anoxic
1000 mg/L TCE
(1)
Oxic
1000 mg/L TCE
(2)
Oxic
100 mg/L TCE
(3)
Anoxic
100 mg/L TCE
(4)
Ottawa Sand
100 mg/L TCE
(6)
Ottawa Sand +
1% Goethite
100 mg/L TCE
(7)
pHIO
100 mg/L TCE
(5)
Incubation
Time (days)
10
20
10
20
10
20
30
10
30
41
10
30
40
1
2
4
1
2
3
4
C lost/C detected
6.4±13.0
3.7±2.7
-2.5±5.2
2.7±4.4
1.9±1.1
1.8±0.0
2.5±0.7
0.4±1.0
0.1±0.9
0.2±0.5
1.2±1.6
2.3±1.4
1.6±2.5
-0.2±1.8
2.0±2.2
1.8±2.5
2.5±0.2
1.7±0.2
2.1±1.2
1.4 (n=l)
Cl lost/Cl detected
6.9±14.4
2.9±0.6
-1.0±4.1
2.4±3.1
1.8±1.2
2.0±0.2
3.9±1.2
0.3±0.8
0.1±1.3
0.2±0.8
1.0±1.2
1.9±2.2
1.4±2.4
-0.4±2.0
3.2±2.0
2.7±2.4
3.5±0.3
1.9±0.3
2.6±1.8
1.4 (n=l)
-------
The ratio of the TCE degradation products—moles of
chloride divided by the moles of carbon as CO, CO2,
formate and glycolate—provides another measure to
determine if all degradation products were detected
and has the advantage of being independent of the
TCE analysis. The ideal ratio of chlorine to carbon
(C1:C), assuming TCE is the only parent compound, is
1.5 (3C1:2C). The values for C1:C shown in Table 4.20
are all close to 1.5, suggesting that the compounds
detected represent the majority of the TCE degradation
products. Therefore, the poor overall mass balance
results shown in Table 4.19 most likely reflect losses in
TCE during sample collection and analysis.
Table 4.20 Amount of Carbon and Chloride Detected as Degradation Products and the C1:C Ratio for Ampules
Incubated at 120°C
Ampule Content
(Batch No.)
Anoxic
1000 mg/L TCE
(1)
Oxic
1000 mg/L TCE
(2)
Oxic
100 mg/L TCE
(3)
Anoxic
100 mg/L TCE
(4)
Ottawa Sand
100 mg/L TCE
(6)
Ottawa Sand +
1% Goethite
100 mg/L TCE
(7)
pHIO
100 mg/L TCE
(5)
Incubation Time
(days)
10
20
10
20
10
20
30
10
30
41
10
30
40
1
2
4
1
2
3
4
C detected (umol)
6.5±0.3
6.0±0.7
5.3±1.7
9.0±2.0
4.5±0.5
7.3±1.0
6.1±0.2
4.7±0.6
6.6±1.0
7.1±0.1
5.2±1.2
7.1±0.8
7.3±0.8
3.3±0.5
4.5±1.1
5.9±0.8
5.5±0.1
9.8±0.7
11.1±2.0
10.1 (n=l)
Cl detected (umol)
5.6±0.3
13.0±1.5
9.5±0.3
21.1±6.3
7.7±1.3
9.9±0.6
6.0±0.1
8.3±1.1
6.3±0.0
6.6±0.4
8.8±0.4
13.1±1.9
12.7±0.6
2.7±0.9
4.3±0.6
6.0±0.3
5.8±0.1
13.6±0.7
14.2±4.8
16.2 (n=l)
C1:C
0.9±0.1
2.2±0.3
1.8±0.6
2.3±0.9
1.7±0.2
1.4±0.1
l.OiO.O
1.8±0.2
l.OiO.l
0.9±0.1
1.7±0.4
1.8±0.3
1.7±0.2
0.8±0.3
1.0±0.3
l.OiO.l
l.liO.O
1.4±0.1
1.3±0.5
1.6
* Corrected for the moles of carbon found in paired ampules without TCE
-------
4.4 Discussion
The primary objectives of the ampule experiments
were to determine the rate of TCE degradation and
the TCE degradation products formed after exposing
dissolved-phase TCE to temperatures of 22 and 120°C
for periods of up to 40 days. The primary gas phase
degradation products observed in these experiments
were CO and CO2 (>99%), while the primary aqueous
phase products included chloride, hydronium ions,
glycolate, and formate (>99%). The following sections
provide a discussion of possible TCE degradation
initiation mechanisms, including oxygen addition to
TCE and hydrogen elimination from TCE. A reaction
sequence is presented to explain the oven explosion, and
a separate section provides comparisons between the
ampule experimental results obtained herein and those of
Knauss etal. (1999).
4.4. 1 Oxygen Initiated TCE Degradation
Based on the review of thermal reaction experiments
presented in Sections 2.3. 1, passing pure oxygen through
liquid TCE-NAPL heated to 70°C is known to yield
hydrochloric acid (HC1), carbon monoxide (CO), and
phosgene (COC12) as gas phase products, and TCE
epoxide (C12COCHC1) and dichloroacetyl chloride
(C12HC2OC1) as NAPL products. This reaction can be
stated as (McKinney et al., 1955):
COCL(g)
H2O,25 C
CO0 + 2HC1
(4.2)
C12C=CHC1 + O2(g)
HCl(g) + CO(g)
COCl2(g) + C12COCHC1 + C12HC2OC1 (4. 1)
McKinney et al. (1955) obtained these results after
adding benzoyl peroxide or partially -oxidized TCE to
liquid TCE-NAPL and Kucher et al. (1990) observed
the same reaction products after adding azo-bis-
isobutyronitrile (AIBN) to liquid TCE-NAPL. Benzoyl
peroxide and AIBN are known to produce radical
initiator compounds upon heating to temperatures
greater than 70°C (Fossey et al., 1995). The reaction
between TCE and oxygen is thought to involve a radical
chain reaction mechanism (Kaberdin and Potkin,
1994). No radical initiator compounds were added to
the ampule experiments. However, it is possible that
partially-oxidized TCE was generated during the flame
sealing process. The radical initiation mechanism for
the aqueous oxidation of TCE is unknown and was not
determined in the experiments herein.
If the reaction products including HC1, phosgene,
dichloroacetyl chloride, and TCE epoxide shown in
Equation 4. 1 were to form in the presence of an aqueous
phase, they would likely undergo subsequent degradation
reactions as discussed in Section 2.3.2. Hydrochloric
acid gas would generate hydronium and chloride
ions in the aqueous phase. Phosgene would undergo
hydrolysis to form CO2 according to Mertens and von
Sonntag (1994):
&=6(s-1)at25°C
Dichloroacetyl chloride would be expected to undergo
hydrolysis to form dichloroacetate (C12HC2OO")
(Kivinen, 1972). TCE epoxide has been synthesized
and the hydrolysis products included CO in the gas
phase, and formate, dichloroacetate, and oxoacetate
(HOC2OO") in the aqueous phase (Cai and Guengerich,
1999). Dichloroacetate (DCAA) is a stable end-product
at room temperature, as indicated by its detection in the
oxic ampules that contained 1,000 mg/L of TCE and
were stored at 22°C (see Figure 4.24). As discussed in
Section 2.3.2, Prager et al. (2001) demonstrated that
DCAA is rapidly (t1/2 = 0.14 day or 3.4 hrs) hydrolyzed
to oxoacetate (HOC OO") at 120°C according to:
CLHCOO-
H,O,120°C
HOCOO + 2HC1
(4.3)
Oxoacetate was not detected in any of the ampules
incubated at 120°C. Prager et al. (2001) suggest that
glycolate (HOH2C2OQ-) and oxalate (HOOC2OQ-)
were likely products from the aqueous phase thermal
degradation of oxoactetate. While oxalate was not
detected, glycolate and formate (CHOO"), two closely
related organic acids, were found in the ampule aqueous
phase and are thought to be the products of DCAA
degradation in the ampule experiments reported herein.
Decarboxylation of glycolate (HOH2C2OO") is thought to
form CO2 and formaldehyde (CH2O) according to Belsky
etal. (1999):
HOH2C2OO-
CHO + CO+H, (4.4)
Formaldehyde would then be hydrolyzed to formate
(CHOO") and hydroxide ions (OH") according to:
CH O + HO -> CHOO- + OH- + 2H+
(4.5)
Formate was suggested as a stable end-product below
temperatures of 150°C by Shende and Mahajani
(1997). However, any one of these organic acids could
be present during the thermal remediation process as
they have been cited as the end products after thermal
treatment of organic compounds by wet air oxidation
(Mishra et al., 1995) and supercritical water oxidation
(Buhleretal. 2002).
Dichloroacetyl chloride, TCE epoxide, and phosgene
were not detected in the aqueous or gas phase samples
collected from the ampules incubated at 120°C in
the experiments reported herein. While the exact
mechanism of oxygen initiated TCE degradation is not
known (Kaberdin and Potkin, 1994), phosgene, TCE
epoxide, and dichloroacetyl chloride are proposed as
intermediates that were hydrolyzed to form HC1, CO,
-------
CO2, dichloroacetate, glycolate, and formate within the
ampule reaction environment.
4.4.2 Hydrogen Elimination Initiated TCE
Degradation
An alternative to the oxygen-initiated TCE degradation
mechanism presented above is based on the belief that
the lone hydrogen atom in TCE has an acid character
and is susceptible to elimination by strong nucleophiles
such as hydroxide ions (OH~) (Smith and March, 2001).
For example, passing gas-phase TCE over potassium
or sodium hydroxide (NaOH) at 130°C has long been
known to degrade TCE to yield dichloroacetylene
(DCA) (Delavarenne and Viehe, 1969). Pielichowski
and Popielarz (1984) reported a 70% yield of DCA
after adding TCE to a NaOH solution heated to 70°C.
The elimination of the acidic hydrogen from TCE has
been proposed to form an unstable intermediate, a
trichlorovinyl anion (C13C2"), which reacts spontaneously
to form DCA (C2C12) after loss of a chlorine atom
(Pielichowski and Popielarz, 1984) according to:
CLC=CHC1 + OH
120 C
-> [CLC=CC1-]
120 C
ClC^CCl + Cl- + H.O (4.6)
DCA was identified in all gas phase samples collected
from ampules that contained TCE and were incubated at
120°C, whereas DCA was not detected in ampules with
TCE that were stored at 22°C. Thus the TCE degradation
observed in the ampules was suspected to have involved
hydrogen elimination based on the presence of DCA.
The addition of NaOH to the ampules resulted in an
increase in the rate of TCE degraded along with an
increase in the amount of DCA detected, as compared
to the results from the deionized water ampules, which
was consistent with the proposed hydrogen-elimination
initiation mechanism. In ampules with deionized water,
the elimination of the lone TCE hydrogen atom may
have been initiated by water or hydroxide ions (10~7 M).
Increasing the ampule temperature from 22°C to 120°C
would have increased the hydroxide ion concentration
from 10-7 to 10-6 M (Marshall and Frank, 1981) and
increased the interaction between TCE and water
molecules by lowering the water dielectric strength.
An increase in hydroxide ion concentration would be
expected to increase the rate of hydrogen elimination
from TCE and the increase in TCE and water interaction
may have allowed the electrophilic oxygen atom in
water to act as a nucleophile and eliminate the TCE
hydrogen. However, the mode of hydrogen elimination
from TCE in deionized water is not currently known.
The addition of goethite to the ampules that contained
anoxic water and sand increased the rate of TCE
degradation. While the TCE degradation mechanism
with goethite is unknown, chloroacetylene (HC Cl), a
compound closely related to DCA, has been detected
during the degradation of chloroethenes by zero valent
iron (Arnold and Roberts, 2000). Where zero valent
iron (Fe°) is thought to serve as an electron donor
(Fe° —> Fe2+ + 2e~) during the reductive dechlorination
of TCE, goethite is not expected to serve as an electron
source since the iron in goethite is oxidized (Fe3+).
Instead, goethite may serve as a source of hydroxide ions
through a dissolution reaction:
FeOOH + HO -> Fe3+ + 3OH'
(4.7)
The dissolution of goethite would be expected to
increase the ampule pH with the formation of hydroxide
ions, however, the pH in ampules with goethite
decreased over the 4 day incubation period (Figure 4.10).
Using pH to indicate the presence of hydroxide ions
may not be accurate since pH is the hydronium ion
concentration in the bulk ampule solution rather than
at the goethite-water interface and because the pH
measurement was performed after cooling the ampules
to room temperature (22°C).
Chloroacetylene has also been detected during
TCE degradation by zero valent zinc (Arnold and
Roberts, 1998) and vitamin B12 (Burris et al., 1996).
Dichloroacetylene, in addition to chloroacetylene, have
been detected during the sonolysis of TCE contaminated
water (Drijvers et al., 1996). The presence of
chloroacetylenes under this range of reaction conditions
suggests that the elimination of hydrogen is an important
reaction pathway for the degradation of TCE. The
degradation of the chloroacetylenes intermediates is
equally important because acetylenes can engage in a
wide variety of reactions (Hopf and Witulski, 1995).
Dichloroacetylene (DCA) is known to react explosively
with oxygen to form CO, CO2, and phosgene (COC12)
(Ott et al., 1930), but is also stabilized against oxygen
in the presence of excess TCE (Williams, 1972;
Reichert et al., 1975). Aqueous solutions of DCA
have been prepared at concentrations up to 70 mg/L
and are reported to be stable over a period of 24 hours
at room temperature (Reichert et al., 1983; Arnold
and Roberts, 1998). There are also reports of DCA
spontaneously igniting upon mixing with water (Ott and
Packendroff, 1931; Riemschneider and Klaus, 1961).
While DCA is known to react with oxygen, the amount
of oxygen present in the ampules did not decline over
the experimental period (see Table 4.18, excluding
Ampule 58). DCA was detected in the oxic ampules
with 100 mg/L of TCE and was most likely stabilized
by the excess amount of TCE present. Therefore, DCA
is hypothesized to have been stabilized against reacting
with the oxygen in the oxic ampules.
Acetylene is known to react with water in both the
aqueous and gas phase to form acetaldehyde, which
-------
hydrolyzes to form acetate (Nieuwland and Vogt,
1945). The addition of water to acetylene is rate limited
and requires acidic conditions along with elevated
temperature (Lucchini and Modena, 1990). The pH
decrease in ampules that were incubated at 120°C
(Section 4.6.2) may have provided the acidic and
temperature conditions necessary to cause the hydrolysis
of DCA and produce chloroacetyl chloride (C1H2C2OC1),
the analog of acetaldehyde, according to:
C1OCC1 + H 0+
120°C
C1HCOC1 + H+ (4.8)
Chloroacetyl chloride could then have reacted with
water to form chloroacetate (C1H2C2OO"). For example,
chloroacetate is the reaction product formed when
passing TCE through heated concentrated sulfuric acid
(Kaberdin and Potkin, 1994). Chloroacetate was detected
in the pH 10 ampules which also had the greatest
concentration of DCA. Chloroacetate is known to react
with water (i.e., hydrolyze) to form glycolate (Prager
et al., 2001), and glycolate was detected in the pH 10
ampules. Therefore, DCA could have been hydrolyzed
to DCAA, which was then degraded at 120°C
(Equation 4.3) to the non-chlorinated organic acids,
glycolate and formate, observed in the ampules that
were not amended with NaOH. However, since DCA is
reactive and could potentially interact or react with soil
constituents such as organic carbon, DCA formed during
the in-situ thermal treatment of TCE could degrade to
form non-chlorinated organic acids.
4.4.3 Oven Explosion
The oven explosion is hypothesized to have been
caused by the detonation of DCA in at least one ampule.
Reichert et al. (1975) showed that DCA is stable in the
presence of excess TCE, however, once the ratio of DCA
to TCE exceeds 1:2, a spontaneous reaction occurs with
oxygen. The explosion occurred 4 days after introducing
the sand, sand+1% goethite, and pH 10 ampules into
the oven, and 20 days after incubation of the 1,000
and 100 mg/L ampules began. The explosion scenario
involves the formation of DCA in many ampules and
in excess of TCE in at least one ampule. Some event
then occurred to initiate the rapid DCA decomposition
resulting in the detonation of at least one ampule and
the force from that one ampule exploding is thought to
have caused the other ampules to break open and release
additional DCA into the air resulting in an explosion that
destroyed the oven.
Goethite was initially suspected of having caused the
formation of DCA in an amount in excess of TCE.
However, the goethite ampules contained argon sparged
water and argon gas, meaning that the oxygen content
was limited to approximately 3 umol. The reaction of
3 umol of DCA with oxygen to yield CO, CO2, and
phosgene would have released approximately 1.4 J of
energy based on the heats of reaction (Zhu and Bozzelli,
2002) which is estimated to result in a 1 bar pressure
increase within an ampule. The ampules have a rated
pressure capacity of approximately 14 bar, thus the
estimated 1 bar pressure increase in the ampules with
goethite would be insufficient to cause an ampule to
break open. In addition to the goethite ampules, there
were ampules with 1,000 mg/L of TCE that contained
air sparged water with air in the headspace and had
been in the oven for 24 days. There was approximately
100 umol of oxygen in these oxic ampules and a reaction
with DCA would have released approximately 48 J of
energy or an estimated pressure increase of 47 bar, about
three times greater than the rated pressure capacity of the
ampules.
A significant amount of DCA could have been generated
in the sand+1% goethite or pH 10 ampules, rendering
these ampules potentially explosive. However, the
amount of oxygen present in these ampules was limited
to less than 3 umol making it unlikely that they initiated
the oven explosion. The 1,000 mg/L oxic ampules that
had been incubating for a period of 24 days at 120°C
are more likely to have initiated the explosion due to the
presence of approximately 100 umol of oxygen.
4.4.4 Comparison to Knauss et al. (1999)
Results
The in-situ transformation of TCE to CO2 and chloride
ions has been claimed to occur during the thermal
treatment of subsurface environments contaminated
with TCE according to researchers working in the
Applied Geology and Geophysics Group at the U.S.
Department of Energy, Lawrence Livermore National
Laboratory (LLNL). The LLNL group based their
claim on experimental results obtained by measuring
the disappearance of dissolved-phase TCE along with
the appearance of dissolved phase CO2 and chloride
ions. The experiments were completed over a period of
up to 43 days within a heated, constant pressure, gold-
walled reactor that was completely water filled (Knauss
et al., 1999). The ampule experiment reported herein
was conducted, in part, to independently evaluate the
conclusions reached by the LLNL group.
The results of the gold-walled experiments conducted
by Knauss et al. (1999) were significantly different than
those reported herein. TCE concentrations decreased
to 50% of the initial concentration within 2 days in
the gold-walled reactor operated at 90°C. In contrast,
approximately 80.2% of the initial amount of TCE
(Table 4.15) remained after 30 days of incubation
at 120°C in the oxic ampules. Knauss et al. (1999)
stated that chloride and carbonate were the only TCE
degradation products detected in the gold-walled
experiment. The primary compounds detected in the
-------
ampule experiments included CO and CO2 in the gas
phase, chloride, hydronium, glycolate, and formate in
the aqueous phase. Knauss et al. (1999) analyzed the
aqueous phase from their gold-walled reactors using
a direct infrared spectroscopic method and a high
pressure liquid chromatograph (HP 1090) connected
to a conductivity detector. Details regarding operation
conditions for these methods were not provided, but
if chloride were detected then organic acids, such as
glycolate and formate, should have been detected if
present. However, Gu and Siegrist (1997) reported
that the liquid chromatography method had a detection
limit of 50 mg/L for glycolate, which is at least twice
the initial concentration of TCE used by Knauss et al.
(1999). Therefore, organic acids may have formed in the
Knauss et al., (1999) experiments, but may have been
well below detection limit of the analytical method.
The difference in experimental results between these
two studies may also be associated with the differences
in reactor wall materials. While the gold-walled reactor
is supposedly chemically "inert" (Seyfried et al., 1979),
gold powder has also been used as a catalyst to increase
the rate of reactions. The glass ampules used herein may
not be entirely chemically "inert" either. The use of the
2,500°C flame to seal the ampule may have generated
radical species that caused an increase in the rate of
TCE degradation. For example, there was a 50.9%
reduction in the amount of TCE in Ampule 58 along
with a consumption of dissolved oxygen and a reduction
in the amount of gas phase oxygen by an estimated 50%.
The half-life for TCE degradation in Ampule 58 was
approximately 20 days, similar to the disappearance
rate reported by Knauss et al. (1999) for an initial TCE
concentration of 1,000 mg/L.
There were other significant differences between the
Knauss et al. (1999) gold-walled experiment and the
ampule experiment reported herein. The gold-walled
experiment used a completely water-filled reactor under
a pressure of 10 bar. The ampule experiments were
completed in glass ampules that were partially filled with
water at a pressure of approximately 1.4 bar at 120°C.
The gold-walled experiment involved measuring the
concentration of TCE in a single gold-walled vessel
every day over a period of days. The ampule experiment
used destructive sampling of individually prepared
ampules after incubation periods to determine the TCE
degradation products.
Due to these differences, a direct comparison regarding
the rate of TCE degradation determined by Knauss et
al. (1999) and the results reported herein is not possible.
However, the results of both studies can be used to
conclude that TCE can be degraded in sealed containers
incubated at elevated temperatures for up to 40 days.
4.5 Summary
The results obtained from the four ampule experiments
demonstrate that TCE was degraded within sealed glass
ampules that contained gas, water, and solids. The rates
of TCE degradation in ampules with anoxic water, both
with and without sand, and in oxic water were similar
at 120°C. The degradation rate in ampules with anoxic
water and sand was increased by adding goethite at
120°C. The primary TCE degradation products included
CO and CO2 in the gas phase and chloride, hydronium
ions, formate, glycolate in the aqueous phase, which
represented less than 15% of the initial amount of TCE
initially present when incubated at 22 and 120°C for
periods of up to 40 days. Minor amounts (<1 mg/L)
of dichloroacetic acid (DCAA) were detected in select
ampules, most consistently in ampules that were stored
at 22°C and initially contained 1,000 mg/L TCE along
with oxygen. Dichloroacetylene (DCA) was detected in
minor amounts (i.e., DCA < 1% of TCE) in ampules that
contained TCE and were incubated at 120°C.
Dichlororacetylene, in addition to being a TCE
degradation product, was also thought to represent a
key intermediate. The presence of DCA is proposed to
indicate that the lone hydrogen atom in TCE was being
eliminated by nucleophiles, such as sodium hydroxide,
which increased the rate of TCE degradation and
amount of DCA when added to the ampules as NaOH.
Dichloroacetylene was proposed to be hydrolyzed
to form chlorinated organic acids, such as DCAA,
which were then hydrolyzed at 120°C to form the
non-chlorinated organic acids, glycolate and formate.
However, DCA is a reactive compound that can interact
with the variety of compounds present in soil such as
organic carbon. Therefore, the degradation products
formed during the in-situ treatment of TCE may not
be limited to those found in the ampule experiments
reported herein.
From a practical perspective, these findings suggest
that abiotic TCE degradation during thermal treatment
may vary considerably depending on site conditions
and operational variables, but it is likely that gas phase
recovery of TCE will be the most important process
in thermal remediation at temperatures up to 120°C.
Additional laboratory-scale studies and treatability
tests with a range of soils and soil constituents are
warranted to further elucidate factors controlling abiotic
TCE reaction pathways, byproduct formation, and
reaction rates. Field samples should also be analyzed to
determine if significant concentrations of byproducts are
formed during in situ thermal remediation.
-------
4.6 Quality Assurance Summary for the
Ampule Experiments
These experiments involved placing water with dissolved
phase TCE into glass ampules that were flame sealed and
incubated at 22 and 120°C for up to 40 days. The quality
assurance efforts for these experiments focused on:
1. Assessing system cleanliness (clean controls)
2. Estimating the variability between ampules
(replicate ampules)
3. Demonstrating analysis method performance
relative to water (matrix spike)
4. Determining if contaminants were introduced during
sample storage (storage blanks)
Ampule Cleanliness. Immediately before use, ampules
were rinsed with Nanopure deionized (DI) water,
autoclaved at 121°C for 20 minutes followed by DI
water rinse before being dried in an oven at 240°C for
at least 2 hours. Ampules with DI water and gas alone
were prepared to evaluate the compounds formed in
ampules without TCE. These clean controls also served
to determine if compounds (e.g., O2 and CO2) were
being introduced during sample collection and analysis.
No compounds were detected in the ampules filled with
DI water and gas alone. Therefore, the ampules were
initially clean and no compounds were introduced during
sample collection or analyses.
Variability. Replicate samples were not collected due
to the limited volume of gas and DI water within each
ampule. Instead, variability was assessed by preparing 3
replicate ampules and collecting one sample from each
ampule phase. This approach provided a measure of the
experimental variability which was thought to be more
important than the analytical variability measured by
replicate samples. Experimental variability was low for
these experiments, less than 15% RSD in all cases and
less than 5% RSD for most results. This was expected
as efforts were employed to mix solutions and solids
before loading into ampules. Also, care was taken to
collect and analyze samples using consistent methods
and procedures.
An explosion occurred during the incubation of ampules
which limited the number of replicate ampules to 2 and
eliminated replicates in one case (NaOH ampules, day
4 incubation). While this limited the ability to assess
experimental variability, it did not preclude an estimate
of variability and thus had limited impact on the overall
experimental objectives.
Method Performance. Analytical method performance
was determined using solutions spiked with compounds
obtained from commercial sources. For example,
chloride standards were prepared from a certified
reference solution with the same DI water source used
to fill the ampules. Method performance during analysis
of each sample batch was evaluated using the spiked
solutions and solutions that were free of compounds
(blanks). In all cases, method performance was within
5% of the expected response for spiked solutions and no
compounds were detected in method blanks.
Storage Blanks. Vials with DI water were prepared and
stored with each sample batch. No compounds were
detected in any storage blanks. No storage blanks were
prepared for the gas samples as they were analyzed
immediately after collection.
-------
5.0
References
Addink, R., M. Antonioli, K. Olie,and H.A.Covers.
Reactions of dibenzofuran and 1,2,3,4,7,8 -
hexachlorodibenzo-p-dioxin on municipal waste
incinerator fly ash. Environ. Sci. Technol. 30 (3):
833-836 (1996).
Amama, P.B., K. Itoh, and M. Murabayashi.
Photocatalytic oxidation of trichloroethylene in
humidified atmosphere. J.Mol. Catalysis A: Chem. 176:
165-172 (2001).
Agency for Toxic Substances and Disease Registry
(ATSDR). lexicological Profile for Trichloroethylene,
ATSDR, Division of Toxicology/Toxicology Information
Branch, Atlanta, GA, 1997.
Arnold, W.A. and A.L. Roberts. Pathways of chlorinated
ethylene and chlorinated acetylene reaction with Zn(O).
Environ. Sci. Technol. 32(19): 3017-3025 (1998).
Arnold, W.A. and A.L. Roberts. Pathways and kinetics
of chlorinated ethylene and chlorinated acetylene
reaction with Fe(0) particles. Environ. Sci. Technol.
34(9): 1794-1805 (2000).
Atkins, PW. Physical Chemistry, 6th Edition, W.H.
Freeman and Co., New York, NY, 1998.
Atwater, J.E., J.R. Akse, J.A. McKinnis, and J.O.
Thompson. Aqueous phase heterogeneous catalytic
oxidation of trichloroethylene. Appl. Catalysis B:
Environ. 11(1): L11-L18 (1996).
Baker, R. S. and M. Kuhlman. "A description of the
mechanisms of in-situ thermal destruction (ISTD)
reactions." Presented at the 2nd International Conf.
on Oxidation and Reduction Technologies for Soil
and Groundwater, ORTs-2, Toronto, Ontario, Canada,
November 17-21, 2002. Available at: www.terratherm.
com.
Baulch, D.L., J. Duxbury, S.J. Grant, andD.C.
Montaque. Evaluated kinetic data for high-temperature
reactions, Vol. 4. Homogeneous gas-phase reactions of
halogen-containing and cyanide-containing species. J.
Phys. Chem. Ref Data 10: 1-721, Supplement 1 (1981).
Belsky, A.J., PG. Maiella, andT.B. Brill. Spectroscopy
of hydrothermal reactions 13. Kinetics and mechanisms
of decarboxylation of acetic acid derivatives at
100-260°C under 275 bar. J. Phys. Chem. A 103 (21):
4253-4260 (1999).
Bergmann, J.G. and J.Sanik, Jr. Determination of trace
amounts of chloride in naptha. Anal. Chem. 29 (2):
241-243 (1957).
Beyke, G. and D. Fleming. In situ thermal remediation
of DNAPL and LNAPL using electrical resistance
heating. Remediation Summer, 5-22 (2005).
Bridgeman, O.C. andE.W. Aldrich. Vapor pressure
tables for water. J. Heat Trans. 86: 279-286 (1964).
Brown, R.F "Pyrolytic Methods in Organic Chemistry."
Academic Press, New York, NY, 1980.
Buhler, W., E. Dinjus, HJ. Ederer, and A. Kruse.
Ionic reactions and pyrolysis of glycerol as competing
reaction pathways in near- and supercritical water. J.
Supercritical Fluids 22: 37-53 (2002).
Burris, D.R., C.A. Belcomyn, M.H. Smith, and
A.L. Roberts. Reductive dechlorination of
tetrachloroethylene and trichloroethylene catalyzed by
vitamin B12 in homogeneous and heterogeneous systems.
Environ. Sci. Technol. 30(10): 3047-3052 (1996).
Burt, E.E. and A.H. Rau. The determination of the level
of bicarbonate, carbonate, or carbon dioxide in aqueous
solutions. Drug Develop. Industrial Pharm. 20(19):
2955-2964 (1994).
Butler, E.G. and K.F. Hayes. Factors influencing rates
and products in the transformation of trichloroethylene
by iron sulfide and iron metal. Environ. Sci. Technol. 35:
3884-3891 (2001).
Buxton, G. V, C.L. Greenstock, W.P Helman, and A.B.
Ross. Critical-review of rate constants for reactions
of hydrated electrins, hydrogen-atoms and hydroxyl
radicals ("OH/'O") in aqueous-solutions. J. Phys. Chem.
Ref. Data 17(2): 513-886 (1988).
Cai, H. and F. P. Guengerich. Mechanism of aqueous
decomposition of trichloroethylene oxide. J.Am. Chem.
Soc. 121 (50): 11656-11663 (1999).
Carlisle, PJ. and A.A. Levine. Stability of
chlorohydrocarbons: II. Trichloroethylene. Industrial
Eng. Chem. 24(10): 1164-1168 (1932).
Carroll, S., G. Heron, R. Aines, R. Newmark, A.
Gavaskar, W.-S. Yoon, and K. Greene. "Single well
steam-enhanced extraction and HPO of chlorinated
solvents." Paper 2B-03, A.R. Gavaskar and A.S.C.
Chen (eds.), Remediation of Chlorinated and
Recalcitrant Compounds, Monterey, CA, May, 2004.
Catoire, V, PA. Ariya, H. Niki, and G.W. Harris. FTIR
study of the Cl- and Br-atom initiated oxidation of
trichloroethylene. Int. J. Chem. Kinetics 29(9): 695-704
(1997).
-------
Chang, W.D. and S.M. Senkan. Detailed chemical kinetic
modeling of fuel-rich C2HCl3/O2/Ar flames. Environ. Sci.
Technol. 23(4): 442-450 (1989).
Chuang, S.C. and J.W. Bozzelli. Conversion of
chloroform to HC1 by reaction with hydrogen and water
vapor. Environ. Sci. Technol. 20(6):568-574 (1986).
Davis, E.L. How Heat Can Enhance In-situ Soil and
Aquifer Remediation. EPA/540/S-97/502. Cincinnati,
OH: U.S. Environmental Protection Agency, 1997.
Available at: http://www.epa.gov/ada/pubs/reports.html.
Davis, E.L., personnel communication, 2007.
Delavarenne, S.Y and H.G. Viehe.
1-Halogenoacetylenes, Chapter 10. In Chemistry of
Acetylenes. H.G. Viehe (ed.), Marcel Dekker, New York,
NY 1969.
Dilling, W.L., N.B. Tefertiller, and G.J. Kallos.
Evaporation rates and reactivities of methylene chloride,
chloroform, 1,1,1-trichloroethane, trichloroethylene,
tetrachloroethylene, and other chlorinated compounds
in dilute aqueous solutions. Environ. Sci. Technol. 9(9):
833-838 (1975).
Drijvers, D., R. De Baets., A. De Visscher, and H. Van
Langenhove. Sonolysis of trichloroethylene in aqueous
solution: volatile organic intermediates. Ultrasonics
Sonochem. 3(2): S83-S90 (1996).
Falk, M and A.G. Miller. Infrared spectrum of carbon
dioxide in aqueous solution. Vibrational Spectroscopy 4:
105-108 (1992).
Farnik, M., M. Weimann, and M.A. Suhm. 2003. Acidic
protons before take-off: A comparative jet Fourier
transform infrared study of small HC1- and HBr-solvent
complexes. J. Chem. Physics 118(22): 10120-10136
(2003).
Fossey, J., D. Lefort, and J. Sorba. "Free Radicals in
Organic Chemistry." John Wiley and Sons, Inc., New
York, NY, 1995.
Froese, K.L. and O. Hutzinger. The formation of
chlorinated benzenes and phenols in fly ash catalyzed
reactions of trichloroethylene. Chemosphere 28(11):
1977-1987 (1994).
Gebhart, B., Y Jaluria, R.M. Mahajan, and B. Sammakia.
"Buoyancy-Induced Flows and Transport." Hemisphere
Publishing Company, New York, NY, 1988.
Goodal, A.M. and K.E. Hewlett. The pyrolysis of
chloroalkenes. Part II. Trichloroethylene. J. Chem. Soc.
(JUL):2599-2603 (1954).
Graham, J.L., D.L. Hall, andB. Dellinger. Laboratory
investigation of thermal degradation of a mixture of
hazardous organic compounds. 1. Environ. Sci. Technol.
20(7): 703-710 (1986).
Gu, B. andR.L. Siegrist. Dehalogenation of chlorinated
organic compounds by strong alkalis. J. Environ. Eng.
123(10): 982-987 (1997).
Haag, W.R., and J. Hoigne. Photo-sensitized oxidation
in natural water via 'OH radicals. Chemosphere 14:
1659-1671 (1985).
Haag, W.R., M.D. Johnson, and R. Scofield. Direct
photolysis of trichloroethene in air: Effect of co-
contaminants, toxicity of products, and hydrothermal
treatment of products. Environ. Sci. Technol. 30(2):
414-421 (1996).
Haag, W.R., and T Mill. Effect of a subsurface sediment
on hydrolysis of haloalkanes and epoxides. Environ. Sci.
Technol. 22: 658-663 (1988).
Haderlein, S.B. and K. Pecher. "Pollutant reduction in
heterogeneous Fe(II)-Fe(III) systems." Chapter 17, In
Kinetics and Mechanisms of Reactions at the Mineral/
Water Interface, by D.L. Sparks and T. Grundl (eds.),
ACS Symposium Series Volume 715, Division of
Geochemistry, American Chemical Society, Washington,
DC, 1998.
Halogenated Solvents Industry Alliance (HSIA).
Trichloroethylene White Paper. HSIA, Inc., Washington,
D.C., 2001. Available at: http://www.hsia.org.
Heron, G., T.H. Christensen, and C.G. Enfield. Henry's
Law constant for trichloroethylene between 10 and 95°C.
Environ. Sci. Technol. 32(10): 1433-1437 (1998).
Heron, G., M. VanZutphen, T.H. Christensen, and
C.G. Enfield. Soil heating for enhanced remediation
of chlorinated solvents: A laboratory study on resistive
heating and vapor extraction in a silly, low-permeable
soil contaminated with trichloroethylene. Environ. Sci.
Technol. 32(10): 1474-1481 (1998).
Hopf, H. and B. Witulski. "Functionalized acetylenes
in organic synthesis." Chapter 2, In Modern Acetylene
Chemistry, by PJ. Stang and F Diederich (eds.), VCH,
New York, NY 1995.
Horvath, A.L. Halogentated Hydrocarbons, Solubility-
Miscibility with Water, Marcel Dekker, Inc., New York,
NY, 1982.
Huang, L.Q. and J. J. Pignatello. Improved extraction of
atrazine and metolachlor in field soil samples. J. Assoc.
Anal. Chemists 73: 443-446 (1990).
-------
Hunt, J.R., N. Sitar, and K.S. Udell. Nonaqueous phase
liquid transport and cleanup 1. Analysis of mechanisms.
Water Resour. Res. 24(8): 1247-1258 (1988).
Interagency DNAPL Consortium. Steam injection
treatment of a DNAPL source zone at Launch Complex
34, Cape Canaveral Air Station, Draft Final Innovative
Technology Evaluation Report. Prepared by Battelle,
Columbus, OH, August 20, 2002.
Interagency DNAPL Consortium. Demonstration of
resistive heating treatment of DNAPL source zone at
Launch Complex 34 in Cape Canaveral Air Force
Station, Florida, Final Innovative Technology Evaluation
Report. Prepared by Battelle, Columbus, OH, February
19,2003.
Integrated Water Resources. Deployment of a steam
injection/extraction system at LC34, Cape Canaveral Air
Station. Prepared for MSB-Technology Applications,
Butte, Montana, March 2003.
Jeffers, P.M. and N.L. Wolfe. Homogeneous hydrolysis
rate constants. 2. Additions, corrections and halogen
effects. Environ. Toxicol. Chem. 15(7): 1066-1070
(1996).
Jeffers, P.M., L.M. Ward, L.M. Woytowitch, and N.L.
Wolfe. Homogeneous hydrolysis rate constants for
selected chlorinated methanes, ethanes, ethenes, and
propanes. Environ. Sci. Technol. 23(8): 965-969 (1989).
Kaberdin, R. V. and V.I. Potkin. Trichloroethylene in
organic synthesis. Russian Chem. Rev. 63(8): 641-659
(1994).
Kaslusky, S.F. and K.S. Udell. A theoretical model of
air and steam co-injection to prevent the downward
migration of DNAPLs during steam-enhanced extraction.
J. Contam. Hydrol. 55: 213-232 (2002).
Kim, HJ. and K.Y. Choo. Arrhenius rate parameters for
the thermal decomposition of trichloroethylene. Bull.
Korean Chem. Soc. 4(5): 203-208 (1983).
Kirkbride, F.W Oxidation of Trichloroethylene. United
States Patent no. 2,292,129. Issued August 4, 1942.
Kivinen, A. "Mechanisms of substitution at the COX
group." Chapter 6, In The Chemistry ofAcylHalides, by
S. Patai (ed.), Interscience Publishers, New York, NY,
1972.
Kline, S.A., J.J. Solomon, andB.L. VanDuuren.
Synthesis and reactions of chloroalkene epoxides. J.
Org. Chem. 43(18): 3596-3600 (1978).
Knauss, K.G., S.C. Copenhaver, and R.D. Aines.
Hydrous Pyrolysis/Oxidation Process for In Situ
Destruction of Chlorinated Hydrocarbon and Fuel
Hydrocarbon contaminants in Water and Soil. U. S.
Patent no. 6,127,592. Issued October 3, 2000.
Knauss, K.G., M.J. Dibley, R.N. Leif, D.A. Mew, and
R.D. Aines. Aqueous oxidation of trichloroethylene
(TCE) a kinetic analysis. Appl. Geochem. 14: 531-541
(1999).
Kriegman-King, M.R. and M. Reinhard. Transformation
of carbon tetrachloride in the presence of sulfide, biotite,
and vermiculite. Environ. Sci. Technol. 26:2198-2206
(1992).
Kriegman-King, M.R. and M. Reinhard. Transformation
of carbon tetrachloride by pyrite in aqueous solution.
Environ. Sci. Technol. 28: 692-700 (1994).
Kucher, R.V., R.I. Flyunt, YA. Pazderskii, VI.
Timokhin, I.P Shevchuk, O.I. Makogon, and A. A.
Berlin. Mechanism of liquid-phase oxidation of
trichloroethylene. DokladyPhys. Chem., Proc.Acad. Sci.
USSR (translated) 309(4-6): 953-956 (1990).
LaGrega, M.D., PL. Buckingham, and J.C. Evans.
Hazardous Waste Management, McGraw-Hill, Boston,
MA, 2001.
Lee, W and B. Batchelor. Abiotic reductive
dechlorination of chlorinated ethylenes by soil.
Chemosphere 55(5): 705-713 (2004).
Lee, W. and B. Batchelor. Reductive capacity of natural
reductants. Environ. Sci. Technol. 37(3): 535-541 (2003).
Lucchini, V and G. Modena. Kinetics of acid-catalyzed
hydration of acetylene. Evidence for the presence in
the solution phase of unsubstituted vinyl cation. J. Am.
Chem. Soc. 112(17): 6291-6296 (1990).
Mallingckrodt and Baker, Inc. Material Safety Data
Sheet: Trichloroethylene. 2003a. Available at: www.
jtbaker.com/msds/englishhtml/t4940.htm.
Mallingckrodt and Baker, Inc. Material Safety Data
Sheet: Dichloroacetic acid. 2003b. Available at: www.
j tbaker. com/msds/englishhtml/D2144. htm.
Marshall, W.L. and E.U. Frank. Ion product of water
substance, 0-1000°C, 1-10,000 bars. New international
formulation and its background. J. Phys. Chem. Ref.
Data 10(2): 295-304 (1981).
McCollom, T.M. and J.S. Seewald. Experimental
constraints on the hydrothermal reactivity of organic
acids and acid anions: I. Formic acid and formate.
Geochimica et Cosmochimica Acta 67(19): 3625-3544
(2003).
-------
McDonald, H. J. The vapor pressure and heat of
vaporization of trichloroethylene. J. Phys. Chem. 48(2):
47-50 (1944).
McKinney, L.L., E.H. Uhing, J.L. White, and J.C.
Picken, Jr. Autoxidation products of trichloroethylene.
J. Agricultural Food Chem. 3(5): 413-419 (1955).
McNeill, W.C. "Trichloroethylene." In Kirk-Othmer
Encyclopedia of Chemical Technology, 3rd Edition. M.
Grayson and D. Eckroth (eds.), John Wiley and Sons,
New York, NY, 1978, Volume 5, 745-753.
Mertens, J.A. "Trichloroethylene." In Kirk-Othmer
Concise Encyclopedia of Chemical Technology, 4th
Edition. J. Kroschwitz (ed.), John Wiley and Sons, New
York, NY, 1999, p. 429.
Mertens, R. and C. von Sonntag. Reaction of the OH
radical with tetrachloroethene and trichloroacetaldehye
(hydrate) in oxygen-free aqueous-solution. J. Chem. Soc.
Perkin Trans. 2(10): 2181-2185 (1994).
Mishra, VS., VV Mahajani, and J.B. Joshi. Wet air
oxidation. Industrial Eng. Chem. Res. 34(1): 2-48
(1995).
Moss, G.R Glossary of Terms Used in Physical Organic
Chemistry. International Union of Pure and Applied
Chemistry (IUPAC) 1994. Available at: http://www.
chem.qmul.ac.uk/iupac/gtpoc/.
Mugdan, M. and J. Wimmer. "Manufacture of Oxidation
Products of Trichloroethylene." United States Patent
1,976,265. Issued October 9, 1934.
Mulholland, J.A., A.F. Sarofim, P. Sosothikul, PA.
Monchamp, E.F. Plummer, and A.L. LaFleur. Formation
of perchloroaromatics during trichlororehtylene
pyrolysis. Combustion Flame 89: 103-115 (1992).
Newmark, R.L., R.D. Aines, K. Knauss, R. Leif, M.
Chiarappa, B. Hudson, C. Carrigan, A. Thompson, J.
Richards, C. Eaker, R. Wiedner, and T. Sciarotta. In Situ
Destruction of Contaminants via Hydrous Pyrolysis/
Oxidation: Visalia Field Test. Lawrence Livermore
National Laboratory, UCRL-ID-132671, Livermore, CA,
1998.
Nieuwland, J.A. and R.R. Vogt. The Chemistry of
Acetylene. American Chemical Society Monograph
Series, Reinhold Publishing Corp., New York, NY, 1945.
Nimlos, M.R., WA. Jacoby, D.M. Blake, and
T.A. Milne. Direct mass spectrometric studies of
the destruction of hazardous wastes. 2. Gas-phase
photocatalytic oxidation of trichloroethylene over TiOr
Products and mechanisms. Environ. Sci. Techno!. 27(4):
732-740 (1993).
O'Mara, M.M., L.B. Crider, and R.L. Daniel.
Combustion products from vinyl chloride monomer. Am.
Industrial Hygiene Assoc. J. 32(1): 153-157 (1971).
Ott, E. and K. Packendorff. Dichloroacetylene II. The
influence of the reaction velocity on the stereochemical
course of halogen addition at the acetylene union.
Berichte 6*B: 1324-1329(1931).
Ott, E., W Ottemeyer and K. Packendorff.
Dichloroacetylene. Berichte 63B: 1941-1944 (1930).
Papirer, E., J.-M. Perrin, B. Siffert, G. Philipponneau,
and J.-M. Lamerant. The influence of grinding on the
surface properties of a-Aluminas. J. Colloid Interface
Sci. 156: 104-108 (1993).
Pearson, C.R. and G. McConnell. Chlorinated Cl and C2
hydrocarbons in the marine environment. Proc. Royal
Soc. London. Series B, Biological Science 189(1096):
305-332 (1975).
Pielichowski, J. and R. Popielarz. Trichloroethylene
in organic synthesis: I. A new catalytic synthesis of
dichloroacetylene. Synthesis-Stuttgart 433-434 (1984).
Prager, L, P. Dowideit, H. Langguth, H.P Schuchmann,
and C. von Sonntag. Hydrolytic removal of the
chlorinated products from oxidative free-radical-induced
degradation of chloroethylenes:acid chlorides and
chlorinated acetic acids. J. Chem. Soc. Perkin Trans. 2,
2001(9): 1641-1647 (2001).
Procaccini, C., J.W. Bozzelli, J.P Longwell, A.F.
Sarofim, andK.A. Smith. Formation of chlorinated
aromatics by reactions of Cl', C12, and HC1 with benzene
in the cool-down zone of a combustor. Environ. Sci.
Technol. 37 (8): 1684-1689 (2003).
Prytula, M.T. Bioavailability andMicrobial
Dehalogenation of Chlorinated Benzenes Sorb ed to
Estuarine Sediments. Ph.D. Thesis. Georgia Institute of
Technology, Atlanta, GA, 1998.
Riemschneider, R. andB. Klaus. Halogenated
acetylenes: I. Density and refractive index of
dichloroacetylene. JustusLiebig'sAnnalen der Chemie
640: 1-5 (1961).
Reichert, D., T. Neudecker, U. Spengler, and D.
Henschler. Mutagenicity of dichloroacetylene and
its degradation products trichloroacetyl chloride,
trichloroacryloyl chloride and hexachlorobutadiene.
Mutation Res. 117: 21-29 (1983).
Reichert, D., M. Metzler, and D. Henschler.
Decomposition of the neuro- and nephrotoxic compound
dichloroacetylene in the presence of oxygen: Separation
and identification of novel products. J. Environ. Pathol.
Toxicol. 4(2,3): 525-532 (1980).
-------
Reichert, D., D. Ewald, and D. Henschler. Generation
and inhalation toxicity of dichloroacetylene. Food
Cosmetics Toxicol. 13(5): 511-515 (1975).
Reiser, C., KM. Lussier, C.C. Jensen and J.I. Steinfeld.
Infrared photochemistry of halogenated ethylenes. J. Am.
Chem. Soc. 101(2): 350-357 (1979).
Ryan, T.A., C. Ryan, E.A. Seddon, andK.R. Seddon.
"Phosgene and Related Carbonyl Chlorides."
Monograph 24, Topics in Inorganic and General
Chemistry. R.J.H. Clark (ed.), Elsevier Science, B.V.,
Amsterdam, The Netherlands, 1996.
Sawhney, B.L., J.J. Pignatello, and S.M. Steinberg.
Determination of 1,2-dibromoethane (EDB) in field
soils: Implications for volatile organic compounds. J.
Environ. Qual 17: 149-152 (1988).
Schmidt, R., C. Betz, and A. Farber. LNAPL and
DNAPL behavior during steam injection into the
unsaturated zone. International Association of
Hydrological Sciences 250: 111-117(1998).
Sepiol, J. and R.L. Soulen. Preparation of
monocyclopropene and diiodocyclopropene. J. Organic
Chem. 40(25): 3791-3793 (1975).
Seyfried, W.E., D.R. Janecky, andM.E. Berndt.
"Rocking autoclaves for hydrothermal experiments
II. The flexible reaction-cell system." Chapter 9, In
Hydrothermal Experimental Techniques, by G.C. Ulmer
and H.L. Barnes (eds.), John Wiley and Sons, New York,
NY, 1987.
Seyfried, W.E., PC. Gordon, andF.W. Dickson. Anew
reaction cell for hydrothermal solution equipment. Am.
Mineralogist 64: 646-649 (1979).
Shende, R. and V.V. Mahajani. Kinetics of wet oxidation
of formic acid and acetic acid. Industrial Eng. Chem.
Res. 36(11): 4809-4814 (1997).
Sleep, B.E. and Y Ma. Thermal variation of organic fluid
properties and impact on thermal remediation feasibility.
J. Soil Contain. 6(3): 281-306 (1997).
Smith, M. and J. March. March's Advanced Organic
Chemistry: Reactions, Mechanisms, and Structure, 5th
Edition. John Wiley and Sons, Inc., New York, NY,
2001.
Staudinger, J. and P. V. Roberts. A critical review of
Henry's law constants for environmental applications.
Critical Rev. Environ. Sci. Technol. 26(3): 205-297
(1996).
Stegemeier, G.L and H.J. Vinegar. "In situ thermal
desorption of soils." In Hazardous & Radioactive Waste
Treatment Technologies Handbook, by J. Chang (ed.),
CRC Press, Boca Raton, FL, 2001, 1-37.
Stumm, W and J.J. Morgan. Aquatic Chemistry, 3rd
Edition. John Wiley and Sons, Inc., New York, NY,
1996.
Su, C., and R.W. Puls. Kinetics of trichloroethene
reduction by zerovalent iron and tin: Pretreatment effect,
apparent activation energy, and intermediate products.
Environ. Sci. Technol. 33:163-168 (1999).
Taylor, PH., B. Bellinger, and C.C. Lee. Development
of a thermal stability based ranking of hazardous organic
compound incinerability. Environ. Sci. Technol. 24:
316-328(1990).
Truex, M., T Powell, and K. Lynch. In situ
dechlorination of TCE during aquifer heating. Ground
Water Monitoring and Remediation 27(2): 96-105
(2007).
Udell, K.S. "Thermally enhanced removal of liquid
hydrocarbon contaminants from soils and groundwater."
In Subsurface Restoration, by C.H. Ward et al, (eds.),
Ann Arbor Press, Chelsea, MI, 1997, 251-270.
United States Department of Energy (U.S. DOE).
Pinellas Environmental Restoration Project, Northeast
Site Area A NAPL Remediation Final Report, GJO-2003-
482-TAC. Washington, DC, 2003. Available at: http://
www.gjo.doe.gov/lm/sites/maps/fl/pinellas/pinellas.htm
United States Environmental Protection Agency
(U.S. EPA). In Situ Treatments of Chlorinated
Solvents: Fundamentals and Field Applications,
EPA/542/R-04/010, Washington, DC, 2004. Available
at: www.cluin.org.
United States Environmental Protection Agency (U.S.
EPA). 2003. Common Chemicals Found at Superfund
Sites. Washington, DC, 2003. Available at http://www.
epa.gov/superfund/resources/chemicals.htm.
United States Environmental Protection Agency (U.S.
EPA). Method 8265, Volatile Organic Compounds
in Water, Soil, Soil Gas, and Air by Direct Sampling
Ion Trap Mass Spectrometry (DSITSM). SW-846 Test
Methods for Evaluating Solid Wastes Physical/Chemical
Methods. Revision 0. Washington, DC, 2002. Available
at: http://www.epa.gov/epaoswer/hazwaste/test/
pdfs/8265.pdf
United States Environmental Protection Agency (U.S.
EPA). Compendium of Methods for the Determination
of Toxic Organic Compounds in Ambient Air - Second
Edition, EPA/625/R-96/010b. Washington, DC, 1999.
Available at: www.epa.gov/ttn/amtic/airtox.html.
United States Environmental Protection Agency (U.S.
EPA). National Primary Drinking Water Regulations:
Disinfectants and Disinfection Byproducts; Final Rule.
Federal Register 63: 69389-69476, 1998.
-------
United States Environmental Protection Agency (U.S.
EPA). Method 8260B, Volatile Organic Compounds
by Gas Chromatography/Mass Spectrometry (GC/
MS). SW-846 Test Methods for Evaluating Solid Wastes
Physical/Chemical Methods. Revision 2. Washington,
DC, 1996. Available at: http://www.epa.gov/epaoswer/
hazwaste/test/8_series.htm
United Stated Environmental Protection Agency (U.S.
EPA). Method 552.2: Determination ofHaloacetic
Acids and Dalapon in Drinking Water by Liquid-Liquid
Extraction, Derivatization, and Gas Chromatography
with Electron Capture Detector. Revision 1.0. In
Methods for the Determination of Organic Compounds
in Drinking Water. EPA/600/R-95/131, Supplement III.
Washington, DC, 1995.
Urben, P.G., L. Bretherick, and M.J. Pitt. Bretherick's
Handbook of Reactive Chemical Hazards. 6th Edition.
Butterworth-Heinemann, Boston, MA, 1999.
Verschueren, K. Handbook of Environmental Data on
Organic Chemicals, 2nd ed., Van Nostrand Reinhold,
New York, 1983.
Vinegar, H.J., G.L. Stegemeier, KG. Carl, J.D.
Stevenson, and R. J. Dudley. "In situ desorption of soils
impacted with chlorinated solvents." Paper #99-450
presented at the Air and Waste Management Association
Annual Meeting, St. Louis, MO, 1999.
Werner, J.H. and T.A. Cool. The kinetics of the
combustion of trichloroethylene for low Cl/H ratios.
Combustion Flame 120: 125-142 (2000).
Williams, F.W Determination of dichloroacetylene in
complex atmospheres. Anal. Chemist 44(7): 1317-1321
(1972).
Wu, Y.G. and Y.F. Lin. The oxidation of trichloroethene
with methane: Experimental and kinetic modeling.
Combustion Flame 137(3): 376-402 (2004).
Yang, G. and I.M. Kennedy. The effect of strain on
laminar diffusion flames of chlorinated hydrocarbons.
Combustion Flame 92: 187-196 (1993).
Yasuhara, A. and M. Morita. Formation of chlorinated
compounds inpyrolysis of trichloroethylene.
Chemosphere 21(4-5): 479-486 (1990).
Yokoyama, K., G. Fujisawa, and A. Yokoyama.
The mechanism of unimolecular dissociation of
trichloroethylene CHC1=CC12 in the ground electronic
state. J. Chem. Physics 102(20): 7902-7909 (1995).
Yu, X.Y and J.R. Barker. Hydrogen peroxide photolysis
in acidic aqueous solutions containing chloride ions.
I. Chemical mechanism. J. Phys. Chem. A 107(9):
1313-1324(2003).
Zhang, Y and I.M. Kennedy. Oxidation of
trichloroethene in stagnation flows over inert and
catalytic heated ceramic surfaces. Combustion Flame
131:436-451 (2002).
Zhu, L. and J.W. Bozzelli. Kinetics and mechanisms for
the thermal chlorination of chloroform in the gas phase.
International J. Chem. Kinetics 35(12): 647-660 (2003).
Zhu, L. and J.W. Bozzelli. Thermodynamic properties of
chloroacetylene, dichloroacetylene, ethynyl radical, and
chloroethynyl radical. Chem. Physics Letters 362(5-6):
445-452 (2002).
-------
6.0
APPENDIX A
Detailed Experimental Methods
for Flow-Through Quartz Tube
Reactors
A.1 Quartz Tube Preparation
The quartz tube and associated connectors were prepared
prior to each isothermal experiment by washing in hot
(45°C) tap water with detergent (Versa-Clean, Fisher
Scientific). The tube and connectors were then rinsed
with deionized (DI) water and placed in a drying oven at
200°C for a period of 2 hours. The tube and connectors
were allowed to cool to room temperature and the
experimental apparatus was assembled and rinsed with
approximately 20 mL of dichloromethane (DCM) for
a period of 5 minutes. A 2 mL sample of the DCM
rinse was collected and stored at 4°C until analyzed
to demonstrate the organic-free initial experimental
condition. After collecting the DCM rinse, the
experimental apparatus was disassembled and remained
in the vent hood for a period of 5 minutes to remove the
residual DCM. The apparatus was then re-assembled
after placing the quartz tube within piece of a 1.5 inch
ID galvanized steel pipe that was located in the tube
oven (Model 21100, Barnstead-Thermolyne, Dubuque,
I A). The steel pipe served to minimize heat transfer
between the tube oven lining and quartz tube.
A.2 Quartz Tube Temperature Profile
The temperature profile within the empty quartz tube
from the gas inlet to outlet was determined while the
tube oven was operated at 120°C (Figure A. 1). The
temperature within the quartz tube was measured by
inserting a certified traceable oven thermometer encased
in a vermiculite filled enclosure (Fisher Scientific,
Fair Lawn, NJ) into the heated quartz tube. The oven
thermometer was held at a specific location within
the quartz tube for 5 minutes and removed to read the
temperature value. This temperature measurement
procedure was repeated over the entire length of the
quartz tube from inlet to outlet in approximately
5 cm increments. There was no gas flow during the
temperature profile measurement procedure.
The temperature profile in the quartz tube was
determined after 30, 60, 90 and 120 minutes of heating
with the temperature profiles after heating the tube
for 90 and 120 minutes nearly identical. As shown
in Figure A. 1, the quartz tube is not at a uniform
temperature. TCE entering the quartz tube experiences
increasing temperatures (temperature gradient) with the
tube inlet at 70°C and the maximum tube temperature
(130°C) located approximately 25 cm or about 10 inches
from the tube inlet. This temperature gradient is
similar in length to the one foot wide, 500 to 700°C
high-temperature zone claimed to cause the in-situ
destruction of TCE during thermal conductive heating
(Section 2.4.2). An attempt was made to measure the
temperature profile of the quartz tube with the oven
at 240°C, however, the thermometer enclosure began
to smoke after being inserted into the tube and the
enclosure was not advanced further to avoid causing the
enclosure to catch fire. No attempt was made to measure
the temperature profile of the quartz tube at 420°C; the
high temperature profile is thought to be similar to that
shown in Figure A. 1 with the region approximately
25 cm from the inlet at 420°C and the tube inlet near
ambient temperature since it is located outside the oven.
^ 0-
o
*-* 10
CD
|.H
2
,2 3o^
CO 40-
o
50-
30 minutes
60 minutes
120 minutes
Tube Oven Setpoint
Tube Outlet
40
60 80 100
Temperature (°C)
120 140
Figure A.1
Temperature profile within the quartz tube
heated to 120°C.
A.3 Modified TCE Introduction Method:
Experimental Series 5
The process of introducing TCE into the pre-mix
chamber consisted of initially recording the weight of
a 1 mL gas-tight syringe that contained approximately
0.34 mL (~0.5 g) of neat TCE using an analytical
balance (Model* AG245, Mettler-Toledo, Columbus,
OH) with 0.001 gram readability. The analytical balance
had been checked using an ASTM E617 class 2 certified
traceable 20±0.0001 gram weight (Cat. # 820000.2,
Denver Instruments, Denver, CO) prior to determining
the syringe weight. The syringe needle was the inserted
through a Teflon lined septum affixed with a crimp seal
-------
to a port located on the pre-mix chamber and TCE was
injected at the slow rate of 0.68 mL/hr for a period of
30 minutes. There were no drops of neat TCE visible
at the syringe needle tip, which was located inside the
premix chamber, when using this TCE injection rate so
that TCE entered the quartz tube in the gas phase. The
syringe was removed from the pre-mix chamber and
the final weight recorded using the analytical balance
and the amount of TCE introduced into the apparatus
was determined by the difference in weight between the
initial TCE-filled syringe and the final syringe weight
after the 30 minute TCE injection period. The apparatus
was flushed with TCE-free humidified carrier gas for
45 minutes after removing the syringe to recover as
much of the TCE introduced into the experimental
apparatus as possible.
The carrier gas was humidified by passing through a
mini-bubbler (ACE Glass, Vineland, NJ) filled with
approximately 30 mL of deionized (DI) water prior
to entering the quartz tube (Figure A.2). The DI water
was freshly dispensed from a Nanopure® analytical
deionization system (model D4741, Barnstead
International, Dubuque, IA) with a conductance of
greater than 18 MQ-cm. The amount of water vapor
entering the quartz tube was adjusted by increasing the
temperature of the mini-bubbler and pre-mix chamber
using a resistant-wire based heat tape (McMaster-Carr,
Atlanta, GA) that was wrapped around the mini-bubbler
and pre-mix chamber and connected to a feedback
voltage controller equipped with a K-type thermocouple.
Three inlet temperatures were used including room
temperature (22), 60 and 100°C. The temperatures were
chosen to explore a range of chloride to hydrogen (C1:H)
ratios based on the work presented in Section 2.3.6.
The room temperature inlet condition had a calculated
C1:H ratio of 1 at a carrier-gas flow rate of 85 mL/min
with the TCE liquid influent rate fixed at 0.68 mL/hr.
The 60°C inlet temperature had a calculated C1:H ratio
of 0.28 and the 100°C inlet temperature had a C1:H
ratio of 0.07. The 100°C inlet temperature represented a
condition where the number of hydrogen atoms in water
was approximately 15 times greater than the number of
chlorine atoms in TCE.
A.4 Effluent Trapping Procedures and
Analytical Methods
The gas stream leaving the ice-cooled, DCM filled
trap (Section 3.1.5) was collected in 1.6 L Tedlar®
bags to retain all single-carbon, non-condensable
degradation products (e.g., carbon dioxide). Each bag
was flushed three times with nitrogen gas prior to use.
The Tedlar® bag was removed from the quartz tube
Pressure
Transducer
Pre-Mix Chamber
Wrapped with
Heat Tape
Tube Oven
1 — 1
i
•s
]
\
0
a
Tedlai
0
rBag
Compressed Dry
Carrier Gas
DCM Trap Aniline Trap
Figure A.2 Quartz tube apparatus for the fifth quartz tube experimental series.
-------
effluent stream when full and a gas sample from the bag
was immediately analyzed to determine the amount of
CO and CO2 formed by the degradation of TCE in the
heated quartz tube. The gas sample from the Tedlar®
bag was collected by pulling approximately 60 mL of the
Tedlar® bag contents through a 250 uL gas sample loop
attached to a gas sampling valve heated to 120°C and
located within an insulated box on a Hewlett-Packard
(HP) 6890 Gas Chromatograph (GC). The gas sample
in the 250 uL sample loop was then injected into the
GC inlet that was operated at 8.90 psi in the splitless
mode for 0.75 minutes at 200°C and was connected
to a 30 m by 0.32 mm OD Carboxen-1010 column
(Part* 24246, Supleco, Bellefonte, PA) attached to a
thermal conductivity detector (TCD). Helium was used
as the capillary column carrier-gas at a constant flow
of 2 mL/min and the GC oven was operated at 35°C
for 7 minutes followed by a 40°C/min temperature
ramp to 130°C for 5 minutes. The TCD was operated at
210°C with a helium reference flow of 15 mL/min and
helium makeup flow at 5 mL/min. The Carboxen-1010
column is capable of separating O2, N2, CO, CO2, and
water. However, TCE and other organic compounds
are retained within the carbon molecular sieve based
column; the column was periodically conditioned at
200°C to remove organic compounds. The GC/TCD
was calibrated using serial dilution of an initial 100 mL
volume of certified carbon dioxide (15%), carbon
monoxide (7%), oxygen (5%), and nitrogen (73%)
gas mixture (Cat. No. 23442, Scotty Specialty Gases,
Plumsteadville, PA). The serial dilution was performed
in a 500 mL syringe (Model S-500, Hamilton Company,
Reno, NV) with nitrogen as the dilution gas. At least
three CO/CO2 concentrations were used to calibrate the
GC/TCD response. This technique had a detection limit
of approximately 300 uL/L (ppmv) for CO and 500 uL/L
(ppmv) CO2.
In addition to the DCM trap, a second 40 mL vial filled
with approximately 30 mL of toluene that contained 2%
(wt) aniline was added to the 420°C UZA experiments
to determine the amount of phosgene formed during
the degradation of TCE. Any phosgene present in the
effluent reacted with the aniline to form carbanilide
(1,3-diphenylurea), a stable compound. The toluene/
aniline traps were analyzed by first removing all the
tolulene from the trap by passing nitrogen at 20 mL/min
through the trap while heating the trap to 60°C. The
trap was taken to complete dryness and weighed to
determine the mass of carbanilide formed. A 10 mL
volume of acetonitrile was then added to dissolve the
dry carbanilide and the concentration of carbanilide
was determined by measuring the ultraviolet (UV)
light absorbance at 254 nm. Calibration solutions
were prepared using carbanilide (Sigma-Aldrich,
Milwaukee, WI) in acetonitrile. This second trap and
analysis methods was based on U.S. EPA method TO-6
(U.S. EPA, 1999).
The GC/MS analysis of the DCM trap fluids from
the 420°C experiments identified a number of TCE
degradation products. However, the amounts of
chloroform (CHC13), carbon tetrachloride (CC14),
tetrachloroethylene (PCE), hexachloroethane (C2C16),
hexachlorobutadiene (C4C16), and hexachlorobenzene
(C6C16) were determined by GC/FID analysis. Master
stock solutions (10,000 mg/L) for each of the previous
compounds were prepared in DCM using ACS
grade reagents (Sigma-Aldrich, Milwaukee, WI).
Hexachlorobenzene (HCB) master stock was prepared
by adding HCB solids to iso-octane. The GC/FID
response was determined for each compound using at
least four calibration standards prepared by volumetric
dilution of the master stock at concentrations in the
expected range.
A.5 Quartz Tube Rinse Procedure
The quartz tube was removed from the tube oven after
cooling to room temperature and the interior of the
apparatus was rinsed with DI water and iso-octane
to determine the TCE degradation products that had
formed and condensed or adsorbed onto the quartz
glass surfaces. The apparatus was initially rinsed
with approximately 30 mL of DI water for a period of
5 minutes to collect the water-soluble compounds (i.e.,
chloride) that formed during each isothermal experiment.
The second rinse used 30 mL of iso-octane for a period
of 5 minutes to collect the non-polar TCE degradation
products (i.e., hexachlorobenzene) from the experimental
apparatus.
The chloride content of all water solutions was measured
using a colorimetric method described by Bergmann and
Sanik (1957). The method involved a selective chemical
reaction between free chloride, mercuric thiocyanate
[Hg(SCN)2], and iron (III) ions from ferric ammonium
disulfate [FeNH4(SO4)2] as shown in Equation A. 1:
2C1- + Hg(SCN)2 + 2FeNH4(S04)2 ->
HgCl2 + 2Fe(SCN)2+ + 2NH4(SO4)23
(A.1)
The resulting iron-thiocyanate complex [Fe(SCN)2+]
forms a yellow color that is directly related to the
amount of chloride present in water samples. The
method consisted of placing 2 mL of a water sample
into a 3 mL capacity Suprasil quartz cuvette (Fisher
Scientific, Fair Lawn, NJ). Next, 200 uL of a 9 M nitric
acid solution with 250 mM of ferric ammonium sulfate
was added to the cuvette followed by 200 uL of ethanol
saturated with mercuric thiocyanate. The cuvette was
capped with a Teflon lid, inverted 4 times to mix the
contents, and then placed in a Varian spectrometer
-------
(Model Gary 3E). The light absorption at 460 nm was
measured after 10 minutes with reference to DI water
contained in a second transmission matched cuvette.
Calibration solutions at concentrations of 2, 20, 50, 100,
200, 400, 600, and 1000 uM were prepared in 100 mL
volumetric flasks using a certified 1,000 mg/L chloride
master stock (SPEX CertiPrep, Metuchen, NJ). The
detection limit for this technique was approximately
0.1 mg/L.
The haloacetic acid content of water samples was
determined using procedures based on EPA method
552.2 (U.S. EPA, 1995). This procedure involved 1) pH
adjustment, 2) liquid-liquid extraction, 3) derivatization,
and 4) neutralization followed by GC analysis. The
water-rinse samples were contained in 40 mL glass vials
sealed with Teflon lined septa affixed with a screw caps.
The pH of each water rinse sample was adjusted to less
than 0.5 by adding 1.5 mL of concentrated sulfuric acid
(H2SO4) to convert any carboxylates present into the acid
form. The pH adjustment was followed by adding 5 mL
of methyl-tert butyl ether (MTBE) to the 40 mL vials
which were then resealed and hand-shaken for 2 minutes
to extract the haloacetic acids. Approximately 3 mL of
the MTBE was then transferred from each 40 mL vial
to 14 mL glass vials using a Pasteur pipette. One mL of
acidic methanol (10% H2SO4) was added to each 14 mL
vial, which were sealed with a Teflon lined septa affixed
with a screw cap and then placed in an oven at 50°C for
a period of 2 hours to convert the carboxylic acids to
their derivatized, methyl ester form. After cooling the
14 mL vials to room temperature, the MTBE extract and
acid methanol mixture was neutralized by adding 2 mL
of saturated sodium bicarbonate solution. Two 1 mL
samples of the MTBE extracts were transferred from the
14 mL vials into 2 mL autosampler vials and the internal
standard, 1,2,3-trichloropropane, was added to each
2 mL vial. The vials were then sealed with Teflon lined
septa affixed with aluminum crimps.
The analysis of the MTBE extracts consisted of using
an automatic liquid sampler (HP7683) to inject 1 uL
of sample into a GC (HP6890) equipped with a 30 m
by 0.32 mm OD HP-1 capillary column connected
to an electron capture detector (BCD). The GC inlet
was operated at 7.00 psi in the splitless mode for
0.5 minutes at 200°C with helium as the column carrier
gas at a constant flowrate of 2 mL/min. The GC oven
was operated at 35°C for 21 minutes followed by an
ll°C/min temperature ramp to 136°C for 3 minutes,
and a final temperature ramp of 20°C/min to 230°C
for 3 minutes. The BCD was operated at 250°C with a
nitrogen gas makeup flow of 60 mL/min. Dichloroacetic
acid calibration standards at concentrations of 12,
50, 100, and 400 ug/L were prepared from a 60 mg/L
primary dilution standard made from ACS grade
dichloroacetic acid (Sigma-Aldrich, Milwaukee, WI).
The calibration samples were processed with each
sample batch along with at least two uncontaminated
water samples including freshly dispensed DI water and
a storage blank.
The GC/ECD chromatograms from the analysis of
the MTBE extracts collected from the 420°C UZA
experiments contained peaks that eluted at times
different than dichloroacetic acid. A GC (HP 6890)
equipped with a 30 m by 0.32 mm DB-5ms column
connected to a mass select detector (MSD, HP5973)
was used to identify the compounds associated with
the unknown peaks. The GC operating conditions
were identical to the GC/ECD method given above.
Compounds were identified using software (ChemStation
ver. D.00.00.38, Agilent Technologies, Palo Alto, CA)
that matched their mass spectra with reference mass
spectra in the NIST/EPA/NIH Mass Spectral Library
(NIST02). The unknown compound was identified
when the mass spectrum fit with a matching NIST02
library spectrum with quality of greater than 70. The
ChemStation software rated the mass spectrum match
on a scale of 0 to 100 whereas the Varian software used
a scale of from 0 to 1,000. The iso-octane rinse was
initially analyzed using the Varian GC/MS to determine
the identity of TCE degradation products. The amount
of each degradation product was then determined using
GC/FID analysis.
-------
7.0
APPENDIX B
Change in Dissolved Oxygen
Dissolved oxygen (DO) was determined using a
colorimetric method that reports DO concentration
(mg/L) in a range between two values (e.g., 6 to
8 mg/L). An estimate of the average DO concentration
range for each set of ampules, as a function of time,
was made based on the colorimetric analysis result for
each vial (Table B.I). The DO concentration in the oxic
ampules (Batches 2 and 3) remained constant (6-8 mg/L)
or increased by one increment (8-10 mg/L) over the
30-day incubation period, with the exception of Ampule
58 (1,000 mg/L TCE). The DO concentration in Ampule
58 decreased from 6-8 mg/L to 0.6-0.8 mg/L after
incubation at 120°C for 20 days. This result suggests that
oxygen was not consumed during ampule incubation at
120°C, with the exception of Ampule 58 where oxygen
was consumed in conjunction with a greater than 50%
decrease in TCE content after 20 days at 120°C.
Table B.I Dissolved Oxygen (DO) Concentration Range for Anoxic and Oxic Ampules with 1,000 and 100 mg/L
of TCE (Batches 1-4)
Ampule Temperature
(°C)
Initial
120
22
120 and 22 (Controls)
120
22
120 and 22 (Controls)
120
22
120 and 22 (Controls)
120
22
120 and 22 (Controls)
Time (days)
0
10
20
30
41
1,000 mg/L anoxic
(1)
0.2 to 0.3
4 to 5
3 to 5
Ito3
Ito3
2 to 3
3 to 4
Lost
4 to 5
5 to 6
NA
1,000 mg/L
oxic (2)
6 to 8
6 to 8
6 to 8
6 to 8
8 to 10*
8 to 10
8 to 10
Lost
8 to 10
8 to 10
NA
100 mg/L anoxic
(4)
0.2 to 0.3
2 to 3
2 to 3
2 to 3
NA
0.8 to 1.0
0.8 to 1.0
0.8 to 1.0
0.8 to 1.0
0.8 to 1.0
0.8 to 1.0
100 mg/L
oxic (3)
6 to 8
6 to 8
6 to 8
6 to 8
8 to 10
8 to 10
8 to 10
6 to 8
6 to 8
6 to 8
NA
(#) - Batch number
Controls - TCE-free ampules
* Excludes Ampule 58 results, DO between 0.6 and 0.8 mg/L after 20 days at 120°C.
Lost -Ampules destroyed in oven explosion
NA - No ampules analyzed for DO at time interval
-------
The DO concentration tended to increase in anoxic
ampules with TCE and without TCE (TCE-free). The
DO concentration increased in anoxic ampules that
contained 1,000 mg/L of TCE (Batch 1) from an initial
value of 0.2-0.3 mg/L to final values up to 4-5 mg/L after
30 days at 22°C and 120°C. The DO concentration also
increased in TCE-free controls for the Batch 1 ampules
from 0.2-0.3 mg/L to 5-6 mg/L over the course of the
incubation period. Anoxic ampules with 100 mg/L
of TCE and without TCE (Batch 4) exhibited slight
increases in oxygen content on day 10 (2-3 mg/L),
followed by declines on day 30 (0.8-1.0 mg/L).
The DO concentrations of ampules that contained
Ottawa sand (Batch 6), Ottawa sand+1% goethite
(Batch 7), and amended with NaOH (Batch 5) are
provided in Table B.2. All of the sampled ampules
exhibited a slight increase in dissolved oxygen
concentration, ranging from the initial anoxic condition
(0.2-0.3 mg/L) up to 1-2 mg/L for the TCE-free control
ampules containing Ottawa sand+1% goethite (Batch 6).
However, the observed increases in dissolved oxygen
concentration in ampules containing solids and NaOH
(Batches 5-7) were less than in the anoxic ampules
containing 1,000 mg/L TCE (Batch 1, Table B.I). For
Batch 6 and 7 ampules, solids filled the ampule neck
thus preventing the measurement of dissolved oxygen
using the procedures illustrated in Figure 4.2. Hence,
the DO concentration for these ampules had to be made
within the main ampule body to gain access to sand-free
water, and for this reason, DO was measured in ampules
that contained solids but not TCE where the CHEMets
reagents may have led to interferences during subsequent
analyses for TCE degradation products.
Table B.2 Dissolved Oxygen (DO) Concentration Range for Anoxic Ampules that Contained Solids (Ottawa
sand and Ottawa Sand+1% Goethite) or NaOH (pH 10) (Batches 5-7)
Ampule
Temperature (°C)
Initial
120
120 (Controls)
120
120 (Controls)
22
120 (Controls)
22 (Controls)
Time (days)
0
10
30
40
Batch 6
Sand
DO (mg/L)
0.2 to 0.3
NA
0.8 to 1.0
NA
0.6 to 0.8
NA
0.3 to 0.4
0.8 to 1.0
Time
(days)
0
1
2
3
Batch 7
Sand+l%Goethite
DO (mg/L)
0.2 to 0.3
NA
Ito2
NA
Ito2
NA
Ito2
Ito2
Batch 5
pHIO
DO (mg/L)
0.2 to 0.3
0.8 to 1.0
0.4 to 0.6
0.8 to 1.0
0.4 to 0.5
0.8 to 1.0
0.4 to 0.6
0.8 to 1.0
(#) - Batch number
Controls - TCE-free ampules
NA - Did not measure DO in solids filled ampules that contained TCE
-------
8.0
APPENDIX C
C.1 Rates of TCE Degradation
Previous studies performed by Knauss et al. (1999)
and Jeffers and Wolfe (1996) reported the rate of TCE
disappearance based on the first-order reaction rate
model. Thus, for comparison purposes, the rate of TCE
disappearance is reported in Table C. 1 based on a fit of
the natural log of TCE concentration vs. incubation time,
and a fit of the natural log of total carbon and chlorine
degradation products versus incubation time. The rate
for ampules with 1,000 mg/L of TCE (Batches 1 and 2)
are not presented as these results were from only 2 time
periods.
While the rate of TCE disappearance from the anoxic
(Batch 4) ampules could not be determined using the
TCE concentration results, there was a measurable
increase in the amount of carbon with incubation time
whereas the rate of chlorine production did not fit a first-
order rate model. The rates of TCE disappearance from
the oxic (Batch 3) ampules, based on the decrease in the
amount of TCE and the increase in carbon degradation
products were similar; however, the uncertainty in the
first order fit based on the carbon degradation products
makes this similarity statistically insignificant. The rate
of chlorine production in the oxic ampules did not fit a
Table C.I Rate of TCE Disappearance from the 100 mg/L Ampules at 120°C
Basis
Rate
Anoxic (4)
Oxic (3)
Sand (6)
Sand+
l%Goethite (7)
NaOH (5)
Knauss et al.
(1999)
Jeffers and Wolfe
(1996)
Decrease in TCE
Rate (I/day) Half-Life
xlOOO (days)
No Change
3.4±0.8
R2 = 0.89 2U1
6.2±2.6
R2 = 0.74
53.4±20.1
R2 = 0.78
34.0±21.9
R2 = 0.45 M
-606 1.1
-0.8 858
Increase in Carbon
Rate (I/day) Half-Life
xlOOO (days)
2.6±0.8
R2 = 0.83 262
2.8±1.2
R2 = 0.72 248
2.SH.O
R2 = 0.80
23.3±6.3
R2 = 0.87
39.9±11.9
R2 = 0.79
Not Reported
Not Reported
Increase in Chloride
Rate (I/day) Half-Life
xlOOO (days)
1.2±1.3
R2 = 0.30
1.8±1.6
R2 = 0.37
3'5±L2 200
R2 = 0.81 2UU
15.8±3.2
R2 = 0.92
42.0±8.1
R2 = 0.90
Not Reported
Not Reported
-------
first-order rate model, similar to the results for the anoxic
ampules. Note that the anoxic and oxic ampule results
span different incubation systems, necessitated by the
convection oven explosion, thus the results for these
two series are not strictly comparable for all incubation
times.
The rate of degradation based on carbon detected as
degradation products was similar within an order-of-
magnitude between the anoxic (Batch 4), oxic (Batch 3),
and anoxic ampules that contained sand (Batch 6)
suggesting that oxygen and sand had no influence
on the rate of TCE disappearance from the ampules.
The addition of goethite increased the rate of TCE
disappearance by an order of magnitude, demonstrating
that a commonly found mineral could have a significant
impact on the rate of TCE disappearance during thermal
treatment. The rate of TCE disappearance reported by
Knauss et al. (1999), reported in Table C. 1 assuming an
initial TCE concentration of 100 mg/L and a temperature
of 120°C, was an order-of-magnitude greater than the
rate measured using the ampules with goethite and two
orders-of-magnitude greater than ampules with oxic and
anoxic water. The rate of TCE disappearance reported
by Jeffers and Wolfe (1996), assuming a temperature of
120°C, was approximately an order-of-magnitude less
than that determined for the anoxic and oxic ampules,
and approximately 3 orders-of-magnitude less than the
rate reported by Knauss et al. (1999).
These rates are not necessarily comparable since they
were obtained using different experimental systems. The
most significant difference between the experimental
systems was that Knauss et al. (1999) used a gold
walled reactor that was completely water filled (no gas
phase present) while Jeffers and Wolfe (1996) used
Pyrex tubing (0.3 mL) with gas phase present, and the
work completed herein used borosilicate glass ampules
(50 mL) with a gas phase present.
C.2 Methods used to Determine TCE
Dedradation Rate Parameters in
Ampule Experiments
Three methods were evaluated for determining the
rate of TCE degradation. The first method consisted
of plotting the natural log of the average number of
moles of TCE (Tables 4.15 and 41.6) normalized by the
average initial moles of TCE as a function of incubation
time as described by the following equation:
In
( molesTCE |
I initial moles TCE I
or In
TCE
'TCE,
(C.I)
Figures C.I through C.4 show plots of Equation C.I for
all four ampule experiments along with the first order
decay coefficient as determined from the slope of a linear
regression analysis. There was no significant change in
the amount of TCE in the anoxic ampules (Figure C.I)
and thus no rate of TCE loss was determined for the
anoxic ampules based on TCE content. There was an
apparent loss of TCE from the other ampules at 22°C
and 120°C based on the data shown in Figures C.2
through C.4. However, the apparent loss of TCE from
the ampules incubated at 120°C may not have been
entirely attributable to degradation alone. No reaction
products were detected (i.e., chloride or organic acids)
in the companion ampules stored at 22°C while a rate
of TCE loss was apparent in these ampules. Thus, a
second method for interpreting the rate of TCE loss
from ampules incubated at 120°C involved adding
the apparent amount of TCE lost from the ampules,
based on the amount of TCE detected in the ampules
stored at 22°C, to the amount of TCE detected in the
120°C ampules. Non-degradative loss of TCE was
thought to occur after cooling the 120°C ampules to
room temperature (22°C), which was potentially due
to sorption to ampule walls or solids, when present.
Equation C.2 corrects for the loss of TCE from the 22°C
ampules:
(TCEO-[TCE 25°-TCE 120° c]"
TCP
I l L,±10
(C.2)
The loss of TCE after correcting for the apparent loss
from the 22°C ampules is shown in Figures C.5 through
C.8. The first order rates shown in these figures are
reported in Table C. 1.
The final method used to determine the rate of TCE
degradation was based on the amount of reaction
products detected. The sum of carbon represented by
CO, CO2, formate, and glycolate were subtracted from
the average initial amount of TCE according to the
following equation:
Z \Carbon\
In
TCEn
(C.3)
The rate of TCE degradation was also determined based
on the amount of chloride detected using Equation C.4:
In
TCE0-
Z[Chloride]
TCEn
(C.4)
Figures C.9 through C.12 show plots of Equations C.2,
C.3, and C.4 for the ampule results.
-------
0.2
0.1
-•- ln(TCE/TCEo) 120°C
..Q... InfTCEA-CEJ 25°C
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
• ln(TCE/TCEJ 120°C
O ln(TCE/TCEtl)250C
k = -68.1x10rV
k = -107.9x10^ f'
10
20 30
Time (day)
40
1 2
Time (day)
Figure C.I The change in anoxic ampule TCE content Figure C.4 The change in anoxic ampule TCE
as a function of incubation time.
InfrCEfTCEJ 120°C
InCTCE/TCEJ 25°C
k = -3.4x1O'er1
-0.2
Time (day)
Figure C.2 The change in oxic ampule TCE content as
a function of incubation time.
content, ampules with Ottawa sand and 1%
Goethite, as a function of incubation time.
0.2
0.1
"E" o.o
-0.1
•- ln(TCE/TCE0)120°C
..Q.. ln(TCE/TCEJ 25°C
^_ ln((TCE 120°C + [TCE^ -TCE 2?C])/TCE0)
10
20 30
Time (day)
40
Figure C.5 The change in anoxic ampule TCE content
as a function of incubation time.
• InCTCE/TCEJ 120°C
0 InCTCE/TCEJ 25"C
20 30 40
Time (day)
Figure C.3 The change in anoxic ampule TCE content,
ampules with Ottawa sand, as a function of
incubation time.
x,
c
O InCTCE/TCEJ 25t
A ln(CTCE 120°C + [TCE.-TCE 2ffC])/TCE0)
k = -3.4x1
0 10 20
Time (day)
Figure C.6 The change in oxic ampule TCE content as
a function of incubation time.
-------
-0.4
-0.5
-0.6
• ln(TCE/TCEJ 120°C
O ln{TCE/TCE0)25°C
A ln(CTCE 120"C + [TC^-TCE 25°C]yTCEJ
k =-6.2x10-V
ft
0 10 20 30 40
Time (days)
Figure C.7 The change in anoxic ampule TCE content,
ampules with Ottawa sand, as a function of
incubation time.
0.1
0.0
-0.1
-0.2
• InCTCE/TCEJ 120°C
O InCTCEfTCEJ 25°C
i ln(CTCE 120*C + [TCE.-TCE 25°C]VTCE0)
k = -53.4x1 ffV1
Time (day)
Figure C.8 The change in anoxic ampule TCE
content, ampules with Ottawa sand and 1%
Goethite, as a function of incubation time.
o.o
X
• ln((TCE120°C + [TCE0-TCE25°C])/TCE0)
O ln(TCE0-SCarton/2XTCE0)
T ln(TCE0-EChlorine/3)n'CE0)
k = -2.1x10Jd'1
10 20 30
Time (day)
40
Figure C.9 The change in anoxic ampule TCE content
based on detected reaction products.
o.o
x
= -0.1
• ln((TCE 120°C
O InCTC
-,-— ln(TCE0-EChlorine/3)/TCEJ
10 20
Time (day)
Figure C.10 The change in oxic ampule TCE content
based on detected reaction products.
x
-0.1
-0.4
ln((TCE120°C + |TCE0-TCE25t])/TCE0)
ln(TCE0-5:Carbon/2VTCE0)
--- k = -2.1x10r3d'1
20 30
Time (day)
Figure C.ll The change in anoxic ampule TCE content,
ampules with Ottawa sand, based on
detected reaction products.
o.o
ln((TCE 120t + [TCE0-TCE 25°C])/TCE0)
o
» ln(TCE0-ZChlorine/3)n-CE0)
k = -15.7x10'V
1 2
Time (day)
Figure C.12 The change in anoxic ampule TCE
content, ampules with Ottawa sand and
1% Goethite, based on detected reaction
products.
-------
-------
--•:.• jncy
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
-------
------- |