&EPA
United States
Environmental Protection
Agency
tnvironmental Research
Laboratory
Athens GA 30605
EPA-600/7-78-074
May 1978
Research and Development
Environmental Pathways
of Selected Chemicals
in Freshwater
Systems: Part II.
Laboratory Studies
Interagency
Energy-Environment
Research
and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-78-074
May 1978
ENVIRONMENTAL PATHWAYS OF SELECTED CHEMICALS
IN FRESHWATER SYSTEMS
Part II: Laboratory Studies
by
J. H. Smith, W. R. Mabey, N. Bohonos,
B. R. Holt, S. S. Lee, T.-W. Chou,
D. C. Bomberger, and T. Mill
i SRI International
! Menlo Park, California 94025
Contract No. 68-03-2227
Project Officer
George Baughman
Environmental Processes Branch
Environmental Research Laboratory
Athens, Georgia 30605
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GEORGIA 30605
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory,
U.S. Environmental Protection Agency, Athens, Georgia, and approved for publi-
cation. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does men-
tion of trade names or commercial products constitute endorsement or recommend-
ation for use.
ii
-------
FOREWORD
Environmental protection efforts are increasingly directed towards pre-
vention of adverse health and ecological effects associated with specific com-
pounds of natural or human origin. As part of this Laboratory's research on
the occurrence, movement, transformation, impact, and control of environmental
contaminants, the Environmental Processes Branch studies the microbiological,
chemical, and physico-chemical processes that control the transport, trans-
formation, and impact of pollutants in soil and water.
Delineation of the environmental pathways followed by potentially harm-
ful chemicals in freshwater systems is a key element in predicting the effects
of pollutants before extensive damage occurs. Based on concepts developed
over a number of years at this Laboratory, the extramural work reported here
integrates independent transformation and transport processes with hydrologic
parameters in a computer model that provides information on environmental ex-
posure in many kinds of aquatic environments.
David W. Duttweiler
Director
Environmental Research Laboratory
Athens, Georgia
iii
-------
ABSTRACT
Environmental exposure assessment models and laboratory procedures for
predicting the pathways of potentially harmful chemicals in freshwater en-
vironments were described in Part I of this report (EPA-600/7-77-113). Pro-
cedures were developed for measuring the rates of volatilization, photolysis,
oxidation, hydrolysis, and biotransformations as well as the sorption parti-
tion coefficients on natural sediments and on a mixture of four bacteria. The
results were integrated with a simple computer model to predict the pathways
of chemicals in ponds, rivers, and lakes. This second part of the project
report describes the successful application of these procedures to 11 chemi-
cals of environmental interest. The chemicals were £-cresol, benzfa]anthra-
cene, benzo[a]pyrene, quinoline, benzo[f]quinoline, 9H-carbazole, 7H-dibenzo-
[c,g]carbazole, benzo[b]thiophene, and dibenzothiophene, which might be found
in the effluents of plants using or processing fossil fuels, and methyl para-
thion and mirex, which are agricultural pesticides.
This report was submitted in partial fulfillment of Contract No. 68-03-
2227 by SRI International under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period from June 30, 1975, to
June 30, 1977.
-------
CONTENTS
Foreword ........ iii
Abstract iv
Figures vii
Tables xiii
Abbreviations and Symbols xxiii
Acknowledgments xxv
1. Introduction 1
2. Conclusions 2
3. Recommendations ..; 5
4. Laboratory Investigation of p-Cresol 6
4.1 Synopsis 6
4.2 Background 6
4.3 Environmental Assessment 9
4.4 Physical Properties 18
4.5 Chemical Transformation 20
4.6 Biodegradation 27
5. Laboratory Investigation of Benz[a]anthracene 39
5.1 Synopsis 39
5.2 Background. 39
5.3 Environmental Assessment 42
5.4 Physical Properties 46
5.5 Chemical Transformation 57
5.6 Biodegradation. 62
6. Laboratory Investigation of Benzo[a]pyrene 64
6.1 Synopsis 64
6.2 Background 64
6.3 Environmental Assessment 67
6.4 Physical Properties 75
6.5 Chemical Transformation 81
6.6 Biodegradation 90
7. Laboratory Investigation of Quinoline .. 93
7.1 Synopsis 93
7.2 Background 93
7.3 Environmental Assessment - 96
7.4 Physical Properties 104
7.5 Chemical Transformation 1Q6
7.6 Biodegradation 112
8. Laboratory Investigation of Benzo[f]quinoline 122
8.1 Synopsis. 122
8.2 Background 122
8.3 Environmental Assessment 124
-------
8.4 Physical Properties 126
8.5 Chemical Transformation 134
8.6 Biodegradation 138
9. Laboratory Investigation of 9H-Carbazole 14j
9.1 Synopsis 143
9.2 Background 143
9.3 Environmental Assessment 145
9.4 Physical Properties 154
9.5 Chemical Transformation 156
9.6 Biodegradation 159
10. Laboratory Investigation of 7H-Dibenzo[c,g]carbazole 167
10.1 Synopsis 167
10.2 Background 167
10.3 Environmental Assessment 171
10.4 Physical Properties 177
10.5 Chemical Transformation 182
10.6 Biodegradation 189
11. Laboratory Investigation of Benzo[b]thiophene 191
11.1 Synopsis 191
11.2 Background 192
11.3 Environmental Assessment 193
11.4 Physical Properties 202
11.5 Chemical Transformation 209
11.6 Biodegradation 212
12. Laboratory Investigation of Dibenzothiophene 216
12.1 Synopsis 216
12.2 Background 216
12.3 Environmental Assessment 219
12.4 Physical Properties 227
12.5 Chemical Transformation 231
12.6 Biodegradation , 234
13. Laboratory Investigation of Methyl Parathion 248
13.1 Synopsis 248
13.2 Background 249
13.3 Environmental Assessment 252
13.4 Physical Properties 260
13.5 Chemical Transformation 267
13.6 Biodegradation 278
14. Laboratory Investigation of Mirex 294
14.1 Synopsis 294
14.2 Background . 294
14.3 Environmental Assessment 296
14.4 Physical Properties 300
14.5 Chemical Transformation 307
14.6 Biodegradation 312
15. References 316
Appendices
A. Input Data for Nine-Compartment Model 333
B. Experimental Procedures 376
C. Natural Water Sources 404
vi
-------
FIGURES
Number Page
4.1 Persistence of £-Cresol in a Two-Compartment Pond System ... 12
4.2 Persistence of £-Cresol in a Partially Mixed River System . . 14
4.3 Persistence of ja-Cresol in a Partially Mixed Eutrophic
Lake ..... ........................ 15
4.4 Persistence of _p_-Cresol in a Partially Mixed Oligotrophic
Lake ............ ................. 16
4.5 Annual Variation of Photolysis Half-Life f or _g-Cresol .... 24
4.6 Biodegradation of ja-Cresol in Eutrophic Pond Water ...... 28
4.7 Viable Cell Counts in 0, 3, and 5 yg ml £-Cresol
Biodegradation by .Batch Fermentation with a Low-Level
Inoculum .................. ......... 30
4.8 Viable Cell Counts in 0, 10, and 20 yg ml"1 p-Cresol
Biodegradation by Batch Fermentation with a Low-Level
Inoculum ........................... 31
4.9 £-Cresol Biodegradation by Batch Fermentation with a
Low-Level Inoculum ...................... 32
4.10 Lineweaver-Burk Plot of Data from _p_-Cresol Biodegradation
in a Batch Fermentation Study ................ 34
4.11 Lineweaver-Burk Plot of Data from j>-Cresol Biodegradation
in a Continuous Fermentation .......... ....... 37
5.1 Persistence of Benz [a] anthracene in a Partially Mixed
Oligotrophic Lake ...................... 47
5.2 Persistence of Benz [a] anthracene in a Partially Mixed
Eutrophic Lake .............. .......... 48
5.3 Persistence of Benz [a] anthracene in a Two-Compartment
Pond System ......................... 49
5.4 Persistence of Benz [a] anthracene in a Partially Mixed
River System ......................... 50
vii
-------
FIGURES (Continued)
Number
Page
5.5 Sorption Isotherms of Benz[a]anthracene on Coyote Creek
and Des Moines River Sediments 54
5.6 Sorption and Desorption Isotherms of Benz[a]anthracene
(Searsville Sediment) 55
5.7 Seasonal and Daily Variation of Photolysis Half-Life
of Benz[a]anthracene 60
6.1 Persistence of Benzo[a]pyrene in a Two-Compartment Pond
System 71
6.2 Persistence of Benzo[a]pyrene in a Partially Mixed
River System 72
6.3 Persistence of Benzo[a]pyrene in a Eutrophic Lake 73
6.4 Persistence of Benzo[a]pyrene in an Oligotrophic Lake .... 74
6.5 Sorption Isotherms of Benzo[a]pyrene ........ 78
6.6 Seasonal and Daily Variation of Photolysis Half-Life of
Benzo [a]pyrene 84
7.1 Persistence of Quinoline in a Two-Compartment Pond System . . 100
7.2 Persistence of Quinoline in a Partially Mixed River System . . 101
7.3 Persistence of Quinoline in a Partially Mixed Eutrophic
Lake 102
7.4 Persistence of Quinoline in a Partially Mixed Oligotrophic
Lake 103
7.5 Direct Photolysis of Quinoline at pH 6.9 and pH 4.4 1Q9
7.6 Annual Variation of Photolysis Half-Life of Quinoline
7.7 Cascade Batch Fermentations with a Freshly Developed
Degrading System Transferred in Fresh Pond Water Media ....
7.8 Viable Counts in Quinoline Degradation by Batch Fermentations
with Low-Level Inocula ........ .... 115
7.9 Quinoline Biodegradation by Batch Fermentations with
Low-Level Inocula 116
viii
-------
FIGURES (Continued)
Number Page
7.10 Lineweaver-Burk Plot for Data from Quinoline Biodegradation
in a Batch Fermentations with Low-Level Inocula ....... 117
7.11 Quinoline Utilization in Cascade Batch Fermentations 120
8.1 Persistence of Benzo[f]quinoline in a Two-Compartment
Pond System 128
8.2 Persistence of Benzo[f]quinoline in a Partially Mixed
River System 129
8.3 Persistence of Benzo[f]quinoline in a Eutrophic Lake .... 130
8.4 Persistence of Benzo[f]quinoline in an Oligotrophic Lake. . . 131
8.5 Seasonal and Daily Variation of Photolysis Half-Life of
Benzo[f]quinoline 137
8.6 Benzo[f]quinoline Biodegradation in Batch Fermentations
with High Cell Counts 141
9.1 Persistence of 9H-Carbazole in a Partially Mixed Two-
Compartment Pond System 150
9.2 Persistence of 9H-Carbazole in a Partially Mixed
River System 151
9.3 Persistence of 9H-Carbazole in a Eutrophic Lake 152
9.4 Persistence of 9H-Carbazole in an Oligotrophic Lake 153
9.5 Seasonal and Daily Variation of Photolysis Half-Life
of 9H-Carbazole 160
9.6 9H-Carbazole Biodegradation in Batch Fermentation with
High Cell Counts (Experiment 1) 163
9.7 9H-Carbazole Biodegradation in Batch Fermentation with
High Cell Counts (Experiment 2) 164
10.1 Persistence of 7H-Dibenzo[c,g]carbazole in a Partially
Mixed Two-Compartment Pond System 173
10.2 Persistence of 7H-Dibenzo[c,g]carbazole in a Partially
Mixed River System 174
ix
-------
FIGURES (Continued)
Number Page
10.3 Persistence of 7H-Dibenzo[c,g]carbazole in a Eutrophic
Lake 175
10.4 Persistence of 7H-Dibenzo[c,g]carbazole in an
Oligotrophic Lake 176
10.5 Sorption Isotherms of 7H-Dibenzo[c,g]carbazole ... 182
10.6 Seasonal and Daily Variation of Photolysis Half-Life
for 7H-Dibenzo[c,g]carbazole 188
11.1 Persistence of Benzo[b]thiophene in a Partially Mixed
Two-Compartment Pond System 198
11.2 Persistence of Benzo[b]thiophene in a Partially Mixed
River System 199
11.3 Persistence of Benzo[b]thiophene in a Partially Mixed
Eutrophic Lake 200
11.4 Persistence of Benzo[b]thiophene in a Partially Mixed
Oligotrophic Lake 201
11.5 Volatilization of Benzo[b]thiophene from Aqueous Solution. . . 204
11.6 Annual Variation of Photolysis Half-Life for
Benzo[b]thiophene 211
12.1 Persistence of Dibenzothiophene in a Two-Compartment
Pond System 223
12.2 Persistence of Dibenzothiophene in a Partially Mixed
River System 224
12.3 Persistence of Dibenzothiophene in a Eutrophic Lake 225
12.4 Persistence of Dibenzothiophene in an Oligotrophic Lake. . . . 226
12.5 Volatilization of Dibenzothiophene from Aqueous Solution . . . 229
12.6 Annual Variation of Photolysis Half-Life of
Dibenzothiophene 233
12.7 Dibenzothiophene Biodegradation in Batch Fermentation with
High Cell Counts 238
-------
FIGURES (Continued)
Number Page
12.8 Metabolism of Dibenzothiophene 239
12.9 Mass Spectrum of Compound A, Dibenzothiophene-5-oxide 240
12.10 Mass Spectra of Compound B, Bis-trimethylsilyl Ether 241
12.11 Mass Spectra of Compound C, Bis-trimethylsilyl
Derivative, Isomer 1 242
12.12 Mass Spectra of Compound D, Bis-trimethylsilyl
Derivative, Isomer 2 243
12.13 Mass Spectrum of Compound E, Trimethylsilyl Ether 244
12.14 Mass Spectrum of Compound F, Trimethylsilyl Ether 245
12.15 Mass Spectrum of Compound G 246
13.1 Persistence of Methyl Parathion in a Two-Compartment
Pond System 256
13.2 Persistence of Methyl Parathion in a Partially Mixed
River System 257
13.3 Persistence of Methyl Parathion in a Eutrophic Lake 258
13.4 Persistence of Methyl Parathion in an Oligotrophic Lake. . . . 259
13.5 Concentration of Methyl Parathion in Supernatant . . 264
13.6 Desorption of Methyl Parathion from Coyote Creek Sediment. . . 266
r
13.7 Annual Variation of Photolysis Half-Life of Methyl
Parathion 270
13.8 pH Rate Profile for Methyl Parathion at 40°C 274
13.9 Cell Counts and MP Concentrations in Batch Fermentations
with Low Count Inocula of Rested Cells 282
13.10 Methyl Parathion Utilization in Batch Fermentations with
Low Count Inocula of Rested Cells 283
13.11 Methyl Parathion Degradation in Batch Fermentations with
Low Inoculum of Rested Cells 285
xi
-------
FIGURES (Concluded)
Number Page
13.12 Lineweaver-Burk Plot of Specific Growth Rates and
Substrate Concentrations for Continuous Feed Chemostat
Fermentations 287
13.13 Cascade Batch Fermentation of Methyl Parathion 289
13.14 Batch Fermentation with a High Population of HP-Degrading
Organisms 290
13.15 Mass Spectrum of the Trimethylsilyl Derivative of
£-nitrophenol 293
14.1 Persistence of Mirex in a Two-Compartment Pond System . . . 301
14.2 Persistence of Mirex in a Partially Mixed Lake 302
14.3 Persistence of Mirex in a Partially Mixed River System. . . 303
xii
-------
TABLES
Number Page
2.1 Summary of Equilibria and Rate Processes at 25°C ........ 3
2.2 Summary of Environmental Assessments .............. 4
4.1 Physical Properties of j>-Cresol ................. 7
4.2 Summary of j>-Cresol Laboratory Data .............. 10
4.3 Transformation and Transport of _o-Cresol Predicted by
the One-Compartment Model .................... H
4.4 Distribution of j>-Cresol in Various Aquatic Systems at
Steady State .......................... 13
4.5 Absorption Spectra of _p_-Cresol in Pure Water .......... 19
4.6 Volatilization Rate Constants for £-Cresol ........... 20
4.7 p-Cresol Sorption on Sediments ................. 21
4.8 Rate Constants for Photolysis of 1.0 yg ml _p_-Cresol ...... 22
4.9 Free Radical Oxidation of 0.72 yg ml"* £-Cresol at
50°C and 100 hours ....................... 25
4.10 Yield of Cells in a p-Cresol Batch Fermentation with a
Low-Level Inoculum .... ................... 33
4.11 First-Order Rate Constants in a £-Cresol Batch Fermentation
with a Low-Level Inoculum ................ .... 35
4.12 Kinetics of a jD-Cresol Biodegradation in a Continuous
Fermentation .......................... 36
4.13 Comparison of Kinetic Data Obtained on Biodegradation
of p-Cresol by Two Procedures ....... ........... 38
5.1 Physical Properties of Benz[a]anthracene ......... . . . 40
5.2 Summary of Benz[a]anthracene Laboratory Data .......... 42
5.3 Transformation and Transport of Benz [a] anthracene
Predicted by the One-Compartment Model ............. 43
xiii
-------
TABLES (Continued)
Number Page
1 i • • • ' - » .1^^—
5.4 Distribution of Benz[a]anthracene in Various Aquatic
Systems at Steady State 45
5.5 Absorption Spectrum of Benz[a]anthracene in 50%
Acetonitrile/50% Water at pH 7 52
5.6 Benz[a]anthracene Sorption on Sediments . 53
5.7 Benz[a]anthracene Sorption and Desorption on a Mixed
Bacterial Population 57
5.8 Benz[a]anthracene Sorption by Viable and Heat-Killed
Bacteria 57
5.9 Rate Constants for Photolysis of Benz[a]anthracene
With Air 58
6.1 Physical Properties of Benzo[a]pyrene .. 65
6.2 Summary of Benzo[a]pyrene Laboratory Data 67
6.3 Transformation and Transport of Benzo[a]pyrene Predicted
by the One-Compartment Model 68
6.4 Distribution of Benzo[a]pyrene in Various Aquatic Systems
at Steady State 70
6.5 Absorption Spectrum of Benzo[a]pyrene in 20% Acetonitrile/
80% Water at pH 6 . 77
6.6 Benzo[a]pyrene Sorption on Coyote Creek Sediments . 80
6.7 Benzo[a]pyrene Sorption and Desorption on a Mixed
Bacterial Population 81
6.8 Rate Constants for Photolysis of Benzo[a]pyrene 82
6.9 Product Yields and Material Balance for Benzo[a]pyrene
Photolysis and Free-Radical Oxidation 86
6.10 Rate Constants for Singlet Oxygen Addition to
Benzo[a]pyrene and Benz[ajanthracene 89
7.1 Physical Properties of Quinoline 95
7.2 Summary of Quinoline Laboratory Data 96
xiv
-------
TABLES (Continued)
Number Page
7.3 Transformation and Transport of Quinoline Predicted by
the One-Compartment Model 97
7.4 Distribution of Quinoline in Various Aquatic Systems
at Steady State 99
7.5 Absorption Spectra of Quinoline in Water at pH 4.7
and pH 6.8 105
7.6 Quinoline Sorption on Sediments 107
I
7.7 Rate Constants for Photolysis of Quinoline 110
7.8 Quinoline Biodegradations in Batch Fermentations with
Low-Level Inocula ..... . 118
7.9 Quinoline Degradations in Cascade Batch Fermentations .... 121
8.1 Physical Properties of Benzoff jquinoline 123
8.2 Summary of Benzo[fJquinoline Laboratory Data 124
8.3 Transformation and Transport of BenzoffJquinoline
Predicted by the One-Compartment Model 125
8.4 Distribution of Benzo[fJquinoline in Various Aquatic
Systems at Steady State 127
8.5 Absorption Spectra of BenzoffJquinoline in Water 132
8.6 Sorption of Benzo[f]quinoline on Coyote Creek Sediment. . . . 133
8.7 Biosorption of Benzo[f]quinoline 134
8.8 Rate Constants for Photolysis of Benzo[fJquinoline 135
8.9 Development of BenzoffJquinoline Biodegrading
Enrichment Cultures 140
8.10 Viable Cell Counts During Pseudo-First-Order Biodegradation
Studies with Benzoff Jquinoline 142
9.1 Physical Properties of 9H-Carbazole 144
9.2 Summary of 9H-Carbazole Laboratory Data 145
xv
-------
TABLES (Continued)
Number Page
9.3 Transformation and Transport of 9H-Carbazole Predicted
by the One-Compartment Model 146
9.4 Distribution of 9H-Carbazole in Various Aquatic Systems
at Steady State 148
9.5 Absorption Spectrum of 9H-Carbazole in 26% Acetonitrile/
74% Water at pH 7 155
9.6 Sorption of 9H-Carbazole on Sediments 157
9.7 Rate Constants for Photolysis of 1.0 yg ml"1 9H-Carbazole. . . 158
9.8 Development of 9H-Carbazole Biodegradation
Enrichment Cultures 161
9.9 Cell Counts During Kinetic Studies of the
Biodegradation of 9H-Carbazole 162
9.10 Rate Constants for Biodegradatinn of 9H-Carbazole 165
10.1 Physical Properties of 7H-Dibenzo[c,g]carbazole 168
10.2 Summary of 7H-Dibenzo[c,g]carbazole Laboratory Data 169
10.3 Transformation and Transport of 7H-Dibenzo[c,g]carbazole
Predicted by the One-Compartment Model 170
10.4 Distribution of 7H-Dibenzo[c,g]carbazole in Various
Aquatic Systems at Steady State 171
10.5 Absorption Spectrum of 7H-Dibenzo[c,g]carbazole in
5% Acetonitrile/95% Water at pH 4.5 178
10.6 Absorption Spectra of 7H-Dibenzo[c,g]carbazole in
50% Acetonitrile/50% Water 179
10.7 Sorption of 7H-Dibenzo[c,g]carhazole on Sediments 181
10.8 7H-Dibenzo[c,g]carbazole Sorption and Desorption with
Mixed Bacterial Population 184
10.9 Rate Constants for Photolysis of 7H-Dibenzo[c,g]carbazole. . . 186
11.1 Physical Properties of Benzofb]thiophene 192
xvi
-------
TABLES (Continued)
Number Page
11.2 Summary of Benzo[b]thiophene Laboratory Data 194
11.3 Transformation and Transport of Benzo[b]thiophene
Predicted by the One-Compartment Model 195
11.4 Distribution of Benzo[b]thiophene in Various Aquatic
Systems at Steady State 197
11.5 Absorption Spectra of Benzofb]thiophene in Water
at pH 6.6 203
I
11.6 VolatilizationiRate Data for Benzo[b]thiophene 205
11.7 Sorption of Benzofb]thiophene on Sediments 207
11.8 Biosorption of Benzo[b]thiophene by Mixed Bacterial
Cultures 208
11.9 Rate Constants for Photolysis of 1.0 vg ml~l Benzofb]thiophene. 209
12.1 Physical Properties of Dibenzothiophene 217
12.2 Summary of Laboratory Data for Dibenzothiophene 220
12.3 Transformation and Transport of Dibenzothiophene Predicted
by the One-Compartment Model 220
12.4 Distribution of Dibenzothiophene in Various Aquatic
Systems at Steady State 222
12.5 Absorption Spectrum of Dibenzothiophene in 30% Acetonitrile/
70% Water at pH 7 227
12.6 Volatilization Data for Dibenzothiophene 228
12.7 Dibenzothiophene Sorption on Coyote Creek Sediment 230
12.8 Biosorption of Dibenzothiophene 231
12.9 Rate Constants for Photolysis of 0.5 yg ml~l
Dibenzothiophene 232
12.10 Development of Dibenzothiophene Biodegrading
Enrichment Cultures 236
12.11 Viable Cell Counts During Pseudo-First-Order Biodegradation
Kinetic Studies with Dibenzothiophene 237
xvii
-------
TABLES (Continued)
Number Page
13.1 Physical Properties of Methyl Parathion 249
13.2 Summary of Methyl Parathion Laboratory Data 252
13.3 Transformation and Transport of Methyl Parathion Predicted
by the One-Compartment Model 253
13.4 Distribution of Methyl Parathion in Various Aquatic
Systems at Steady State 255
13.5 Absorption Spectrum of Methyl Parathion in Water 261
13.6 Methyl Parathion Sorption on Sediments 263
13.7 Sorption Partition Coefficients of Methyl Parathion on
a Mixed Bacterial Population 267
13.8 Rate Constants for Photolysis of 26 yg ml"1
Methyl Parathion 268
13.9 Free Radical Oxidation of Methyl Parathion at 50°C. 269
13.10 First-Order Rate Constants kfl for Hydrolysis of Methyl
Parathion as a Function of pH 272
13.11 First-Order Rate Constants kg for Hydrolysis of Methyl
Parathion in Natural Waters 273
13.12 Temperature Dependence of Rate Constants for Neutral (k^)
and Base-Catalyzed (kg) Hydrolysis of Methyl Parathion. ... 275
13.13 Product Balance for Hydrolysis and Photolysis of Methyl
Parathion 276
13.14 Cell Counts in Batch Fermentations with Low Count
Inocula of Rested Cells 280
13.15 Methyl Parathion Utilization in Batch Fermentations
with Low Count Inocula of Rested Cells 281
13.16 Summary of Data and Kinetic Constants in Batch Fermentations
with Low Count Inocula of Rested Cells 284
13.17 Data From Continuous Feed Chemostat Fermentations 286
13.18 Summary of Kinetic Constants for Methyl Parathion
Biodegradation 292
xviii
-------
TABLES (Continued)
Number Page
14.1 Physical Properties of Mirex 295
14.2 Summary of Mirex Laboratory Data 296
14.3 Transformation and Transport of Mirex Predicted by
the One-Compartment Model. 297
14.4 Distribution of Mirex in Various Aquatic Systems at
Steady State 299
14.5 Sorption of Mirex on Coyote Creek Sediment 306
14.6 Mirex Sorption and Desorption on a Mixed Bacterial
Population 307
14.7 Aerobic Mirex Enrichment Studies with Aeration Effluents . . . 313
A.I Nine-Compartment Computer Model Inputs for £-Cresol: Pond
Simulation 333
A.2 Nine-Compartment Computer Model Inputs for jj-Cresol: River
Simulation 334
A.3 Nine-Compartment Computer Model Inputs for p-Cresol:
Eutrophic Lake Simulation 335
A. 4 Nine-Compartment Computer Model Inputs for j3-Cresol:
Oligotrophic Lake Simulation 336
A.5 Nine-Compartment Computer Model Inputs for Benz[a]anthracene :
Pond Simulation 337
A.6 Nine-Compartment Computer Model Inputs for Benz[a]anthracene:
River Simulation 338
A.7 Nine-Compartment Computer Model Inputs for Benz[a]anthracene :
Eutrophic Lake Simulation 339
A.8 Nine-Compartment Computer Model Inputs for Benz[a]anthracene:
Oligotrophic Lake Simulation 340
A.9 Nine-Compartment Computer Model Inputs for Benzo[a]pyrene:
Pond Simulation 341
A.10 Nine-Compartment Computer Model Inputs for Benzo [a.Jpyrene :
River Simulation 342
xix
-------
TABLES (Continued)
Number Page
A.11 Nine-Compartment Computer Model Inputs for Benzcfajpyrene:
Eutrophic Lake Simulation 343
A.12 Nine-Compartment Computer Model Inputs for Benzo[a]pyrene:
Oligotrophic Lake Simulation ... 344
A.13 Nine-Compartment Computer Model Inputs for Quinoline: Pond
Simulation 345
A.14 Nine-Compartment Computer Model Inputs for Quinoline:
River Simulation 346
A.15 Nine-Compartment Computer Model Inputs for Quinoline:
Eutrophic Lake Simulation 347
A.16 Nine-Compartment Computer Model Inputs for Quinoline:
Oligotrophic Lake Simulation 348
A.17 Nine-Compartment Computer Model Inputs for BenzoffJquinoline :
Pond Simulation 349
A.18 Nine-Compartment Computer Model Inputs for Benzc(f]quinoline:
River Simulation 350
A.19 Nine-Compartment Computer Model Inputs for Benzo[f]quinoline:
Eutrophic Lake Simulation 351
A.20 Nine-Compartment Computer Model Inputs for Benzo[fjquinoline:
Oligotrophic Lake Simulation . 352
A.21 Nine-Compartment Computer Model Inputs for Carbazole:
Pond Simulation 353
A.22 Nine-Compartment Computer Model Inputs for Carbazole:
River Simulation 354
A.23 Nine-Compartment Computer Model Inputs for Carbazole:
Eutrophic Lake Simulation 355
A.24 Nine-Compartment Computer Model Inputs for Carbazole:
Oligotrophic Lake Simulation 356
A.25 Nine-Compartment Computer Model Inputs for 7H-Dibenzo[c,g]-
carbazole: Pond Simulation 357
A.26 Nine-Compartment Computer Model Inputs for 7H-Dibenzo[c,g]-
carbazole : River Simulation 358
xx
-------
TABLES (Continued)
Number Page
A.27 Nine-Compartment Computer Model Inputs for 7H-Dibenzo[c,g]-
carbazole: Eutrophic Lake Simulation . . 359
A.28 Nine-Compartment Computer Model Inputs for 7H-Dibenzo[c,g]-
carbazole: Oligotrophic Lake Simulation ... 360
A.29 Nine-Compartment Computer Model Inputs for Benzo[b]thiophene :
Pond Simulation 361
A.30 Nine-Compartment Computer Model Inputs for Benzo[b]thiophene :
River Simulation 362
I •
A.31 Nine-Compartment ;Computer Model Inputs for Benzo[bjthiophene:
Eutrophic Lake Simulation 363
A.32 Nine-Compartment Computer Model Inputs for Benzo[b]thiophene:
Oligotrophic Lake Simulation 364
A.33 Nine-Compartment Computer Model Inputs for Dibenzothiophene:
Pond Simulation ..... 365
A.34 Nine-Compartment Computer Model Inputs for Dibenzothiophene:
River Simulation . . 366
A.35 Nine-Compartment Computer Model Inputs for Dibenzothiophene :
Eutrophic Lake Simulation 367
A.36 Nine-Compartment Computer Model Inputs for Dibenzothiophene:
Oligotrophic Lake Simulation 368
A. 37 Nine-Compartment Computer Model Inputs for Methyl Parathion:
Pond Simulation 369
A.38 Nine-Compartment Computer Model Inputs for Methyl Parathion:
River Simulation 370
A.39 Nine-Compartment Computer Model Inputs for Methyl Parathion:
Eutrophic Lake Simulation 371
A.40 Nine-Compartment Computer Model Inputs for Methyl Parathion:
Oligotrophic Lake Simulation 372
A.41 Nine-Compartment Computer Model Inputs for Mirex: Pond
Simulation 373
A.42 Nine-Compartment Computer Model Inputs for Mirex: River
Simulation
xxi
-------
TABLES (Concluded)
Number Page
A.43 Nine-Compartment Computer Model Inputs for Mirex: Eutrophic
and Oligotrophic Lake Simulations 375
B.I Nominal Wavelengths and Wavelength Intervals for UV and
Visible Absorption Spectra 377
B.2 Sources and Characteristics of Ca-Montmorillonite Clay and
Natural Sediments 379
B.3 Recommended Experimental Plan for Isotherm Measurements .... 380
B.4 Some Water Quality Parameters for 0.22-um Filtered Water
From Searsville Pond, Coyote Creek, and Lake Tahoe 386
B.5 UV Absorbance of 0.22-ym Filtered Natural Waters 388
xxii
-------
ABBREVIATIONS AND SYMBOLS
AA 4,4'-Azobis(4-cyanovaleric acid)
H Henry's law constant
I Light flux (photons time"1 liter"1)
A
K Partition coefficient for sorption on biota or sorption on sediments
K Partition coefficient
P
K Concentration of substrate at which u = % u
s. m
M Moles liter"1
>L. Mass of chemical in biota
Dl . I ,
Mj . Total mass of substrate in aqueous phase (&) of compartment i before sorp-
tion
M' . Total mass of substrate in aqueous phase (&) of compartment i after sorp-
tion
M Mass of suspended sediment
S •
M . Total mass of substrate in the suspended sediment of compartment i before
sorption
M' . Total mass of substrate in the suspended sediment of compartment i after
sorption
M Mass of water
w
P Vapor pressure of pure substrate (torr)
R Gas constant
*
S Substrate concentration (mass per unit volume)
[S] Substrate concentration (moles per liter)
S Substrate concentration in the aqueous phase of compartment i
T Temperature (°K)
V. Volume of compartment i
X Bacterial mass or cell count (cells ml"1)
*
Specific units of mass, volume, and time used varied.
a.
Biomass in kinetic studies was generally in cell counts and in units of mass
in biosorptions.
xxiii
-------
X.^ Microbial population in compartment i (cells ml"1)
X
Y Biomass or cell yield per mass of substrate utilized' (cells ug"1 sub-
strate)
Z Solar radiance intensity (photons cm"3 sec"1 nm~l)
e Efficiency of production of R02• from AA
k Rate constant for acid-catalyzed hydrolysis (M~l sec"1)
o.
k Rate constant for base-catalyzed hydrolysis (M~ sec" )
k^j Rate constant for neutral hydrolysis (sec"1)
k First-order rate constant for light absorption by chemical (sec~l)
3.
k u/Y = Rate constant for biodegradation (mg cell hr )
k? Pseudo-first order rate constant for biodegradation (hr )
k, 2 Second-order rate constants for biodegradation (ml cell"1 hr"1)
k. Rate constant for hydrolysis (see"1)
— 1
k. Rate constant for decomposition of AA (sec )
k Oxygen reaeration rate (time"1)
k Rate constant for oxidation (M~l time"1)
ox
k Rate constant for photolysis (time"1)
k Rate constant for volatilization (ug ml"1 time"1)
v
m, . Mass of biota
r Biodegradation rate (ug ml"1 time"1)
r. Hydrolysis rate (ug ml"1 time"1)
r Oxidation rate (ug ml"1 time"1)
r Photolysis rate (ug ml"1 time"1)
r Volatilization rate (ug ml"1 time"1)
t, Half-life (time)
e Absorption coefficient (M cm"1)
A Wavelength (nm)
n absorpl
u Specific growth rate (hr"1)
u = Maximum specific
max
Reaction quantum yield
A Wavelength of an absorption maximum (nm)
max
u u = Maximum specific growth rate (hr"1)
m max t- a
xxiv
-------
ACKNOWLEDGMENTS
The assistance and counsel of the Project Officer for this work, Mr.
George Baughnjan of the Environmental Research Laboratory, Athens, Georgia,
has been invaluable in this work, as were the frequent discussions with his
staff: Drs. Richard Zepp, Lee Wolfe, David Brown, Charles Steen, and Ms. Doris
Paris, as well as Drs. James Hill and Ray Lassiter, who are also located at
the Athens, Georgia, laboratory.
The contributions of the SRI International staff members who carried out
the laboratory work are gratefully acknowledged. The following persons parti-
cipated in the following tasks :
Environmental Assessment: T. Peyton, B. Suta, E. C. Waters
Physical Transport: D. Haynes, B. Kingsley, D. Stivers, and M. Zinnecker
Chemical Degradation: D. Hendry, A. Baraze, B. Lan, and H. Richardson
Biodegradation: R. Spanggord, E. Shingai, H. G. Shan, S. Sorenson, G.
Shepherd, R. Langley, D. Donaldson, and D. Brajkovich
Substantial assistance in report preparation was provided by C. Reeds.
xxv
-------
1. INTRODUCTION
The objective of this study was to develop objective, we11-documented
laboratory and assessment procedures that can be used to predict the pathways
of potentially harmful chemicals in freshwater environments. Laboratory pro-
cedures were used to measure the sorption partition coefficients on sediments
and biomass and the rate constants for volatilization, oxidation, hydrolysis,
photolysis, and microbial transformations, using adapted cultures, under con-
ditions that are extrapolatable to environmental conditions. The results of
the laboratory procedures were integrated with a simple computer model to pre-
dict the pathways for the chemicals in ponds, rivers, and lakes. These
procedures are described ip detail in Part I of this report (EPA-600/7-77-113).
The results of applying these procedures to eleven chemicals are described
in this part of the report. The chemicals were p-cresol, benz[a]anthracene,
benzo[a]pyrene, quinoline, benzo[f]quinoline, 9H-carbazole, 7H-dibenzo[c,g]-
carbazole, benzo[b]thiophene, and dibenzothiophene, which might be found in the
effluents of plants using or processing fossil fuels, and methyl parathion and
mirex, which are agricultural pesticides.
-------
2. CONCLUSIONS
The environmental assessment models and the laboratory procedures described
in Part I of this report have been used successfully to predict the environ-
mental pathways of eleven organic chemicals. Nine of the chemicals are
representative of those that would be found in the effluents of facilities
utilizing fossil fuels. Two pesticides were also tested. The results of
these studies are summarized in Tables 2.1 and 2.2.
-------
TABLE 2.1. SUMMARY OF EQUILIBRIA AND RATE PROCESSES AT 2S°C
Chemical
£-Cresol
Benz[ a] anthracene
Benzo[ a] pyrene
Qulnoline
Benzo[ f ] qulnollne
9H-Carbazole
7H-Dlbenzo[ c , g] carbazole
Benzo[ b] thlophene
Dlbenzothiophene
Methyl parathion
Mirex1
Rate law
Solubility in
water at 22° C
1.8b mg ml"1
5.7 ng ml"1
1.2 ng ml"1
6.11b mg ml"1
76 ng «al"1
1.0 ng "I"1
63 ng ml"1
130 ng ml"1
1.1 ng ml"1
50b ng ml"1
70 pg ml"1
—
Sorption
partition
coefficient*
<10
26,000
76,000
11
1,300
175
20,500
59
1,380
50
460,000
--
Volatilization
ratio, k*/ky
(X 103)
1.2 ± 0.3
1.6 ± 0.6
3.6 ± 0,6
2.7 ± 0.4
0.22 ± 0.09
<0.1
0.1
380 ± 20
120 ± 40
<0.4
63 ± 18
4 ><>>
V
Solar photolysis
rate constant, kD
(sec'M P
3.3 x 10""c
3.3 x 10"«d
3.6 x 10"*d
3.5 x 10"7C
3.7 x 10"«d
1.9 x 10"*d
5.5 x 10"«d
5.7 x 10"6C
1.5 x 10"8C
9.8 x 10"7C
5 x 10"aj
kp[S]
Photolysis
quantum
vield. »
0.079
3.3 x 10~a
8.9 x 10"*
3.3 x 10"*
0.014
7.6 x 10"3
2.8 x 10'3
O.lf
5 x 10"*
1.7 x 10"«
<1.0f
--
Oxidation rate
constant, kox
(M"1 sec"1)
20
5,000
1,900
<3
<3
29
830
83
<3
<3
«1
kox[ROa.][S]
Biodegradation
rate constant, kb2
(ml cell"1 hr"1)
5.2 x 10"7
Oe
Oe
3.1 x 10"8
3.6 x 10"8
5.0 x 10"7
Oe
g
5.3 x 10"7
2.4 x 10"7
oe
kb2[B][SI
*Value for a natural sediment collected in Coyote Creek, California.
Literature value.
Rate constant averaged over 24-hr day, June 21.
Instantaneous rate constant for noon, June 21.
Biodegradation was not observed during our enrichment procedures.
Assumed value.
8Blodegradatlon was observed only in the presence of naphthalene; there is no kinetic expression that can be used to express this type of
biodegradation.
''Rate constant for neutral hydrolysis at 25°C is 9 x 10"8 sec"1 .
State constant for neutral hydrolysis is less than 1 x 10~10 sec"1.
^Measured rate constant for a 6-month period.
-------
TABLE 2.2. SUMMARY OF ENVIRONMENTAL ASSESSMENTS
Chemical
j> -Cresol
Benz [a ]-
anthracene
Benzo [a]-
pyrene
Qulnoline
Benzo [f ]-
quinoline
9H-Carbazole
7H-Dibenzo[c,g]-
carbazole
Benzo [b]-
thiophene
Dibenzo-
thiophene
Methyl
parathion
Mirex
River
Major
pathway(s)
Biodegradation
Sorption
Sorption,
photolysis
Biodegradation
Photolysis
Photolysis,
biodegradation
Photolysis,
sorption
Volatilization
Biodegradation
Biodegradation
Sorption,
volatilization
•Half-
lifea
(hr)
0.55
0.55
0.48
0.28
0.5
0.5
0.36
0.5
0.5
0.6
0.83
Percent
sorbedb
0.1
71
83
0.1
12
2.0
66
0.5
12
0.5
97
Pond
Major
pathway (s)
Biodegradation
Sorption
Sorption,
photolysis
Biodegradation
Photolysis,
biodegradation
Photolysis,
biodegradation
Photolysis,
sorption
Volatilization
Biodegradation,
sorption
Biodegradation
Sorption
Half-
life3
(hr)
12
22
7.3
0.5
6.9
10
1.5
19
13
27.3
420
Percent
sorbedb
0.3
88
93
0.3
42
5.6
82
1.5
42
1.5
99
Eutrophic
Major
pathway (s)
Biodegradation
Sorption
Sorption,
photolysis
Biodegradation
Photolysis,
biodegradation
Photolysis,
biodegradation
Photolysis,
sorption
Volatilization
photolysis,
biodegradation
Biodegradation
Biodegradation
Sorption
lake
Half-
life3
(hr)
12
22
7.4
0.5
7.0
10
1.5
19
13
28.3
1480
Percent
sorbedb
0.05
55
71
0.05
6.5
1
50
0.2
6.5
0.25
94
Oligotrophlc lake
Half-
Major life*
Pathway(s) (hr)
Biodegradation, 2400
photolysis
Sorption 8
Sorption, 1.
photolysis
Biodegradation, 600
photolysis
Photolysis 1.
Photolysis, 3
biodegradation
Photolysis, 0.
sorption
Volatilization, 140
photolysis,
biodegradation
Photolysis, 140
Volatilization
Sorption, 152
photolysis,
hydrolysis,
biodegradation
Sorption 1480
Percent
sorbedb
0.05
55
5 71
0.05
4 6.5
1
5 50
0.2
6.5
2.25
94
Net half-life for all processes including dilution; predicted by the one-compartment model.
Percent sorbed in bottom sediment; from the compartment nearest the source; predicted by the nine-compartment computer model.
-------
3. RECOMMENDATIONS
General recommendations for improving this environmental exposure assess-
ment method were given in Section 3 of Part I of this report. In addition,
we offer the following recommendations.
1. These procedures should be verified by ecosystem and field
studies, using several of the compounds that were predicted
in this report to be persistent.
2. The methodology
the strategies
developed here should be used to improve
for setting effluent guidelines.
-------
4. LABORATORY INVESTIGATION OF p_-CRESOL
4.1 SYNOPSIS
The laboratory data suggest that biodegradation is the major pathway
for £-cresol in eutrophic systems. In oligotrophic systems, both biodegra-
dation and photolysis may be important. Volatilization, sorption onto
suspended sediments, and oxidation are not significant processes compared
with biodegradation and photolysis.
Limited product identification studies were carried out. Two dimeric
products of ji»-cresol were found in small amounts in oxidation experiments.
Biodegradation products containing an aromatic ring were not observed; if
they were formed, they were degraded by the mixed culture systems at a
rate comparable to the degradation rate of p-cresol at the low concentra-
tions of j>-cresol tested.
The nine-compartment environmental exposure model predicted the fol-
lowing steady-state concentrations of £-cresol in solution, suspended
solids, and sediments near point sources in the presence of a continuous
discharge of 1 pg ml"1 (1 ppm) jj-cresol:
Suspended
Half-life* Solution solids Sediments
(hr) (yg ml"1) (pg g-1) (pg g-1) -
River 0-55 0.980
Pond 12 0.017
Eutrophic 12 0.037
lake
Oligotrophic 2400 0.0221
lake
9.80
0.17
0.37
2.20
9.65
0.17
0.10
0.94
4.2 BACKGROUND
Phenols enter the environment from diverse biological sources, as by-
products from the petroleum and coking industries and as a result of the
use of creosote as a wood preservative. Consequently, they may be widely
^Predicted by one-compartment model for all processes including dilution.
-------
encountered in low concentrations in natural waters. p_-Cresol was selected
for investigation as being representative of the general class of phenolic
compounds. jg-Cresol is a constituent of creosote, and the use of creosote
probably represents the highest localized concentrations of p_-cresol in the
environment.
Production of creosote and other coal tar derivatives is concentrated
in the Great Lake and Middle Atlantic States with a secondary locus in central
Alabama. These products are used throughout the nation, with major concen-
trations in the Southeast and Pacific Northwest. Approximately 84 plants
produce creosote and some 200 use it for wood preservation. Effluents from
most of these plants receive biological treatment. The prolonged use of
creosote without noticeable adverse effects suggests that releases to the
environment are relatively nonhazardous. The physical properties of £-cresol
available in the literature are listed in Table 4.1.
TABLE 4.1. PHYSICAL PROPERTIES OF p-CRESOL
Structure
Molecular weight
Melting point, °C
(CRC Handbook, 54th Ed.)
Boiling point, °C, 760 torr
(CRC Handbook, 54th Ed.)
Vapor pressure, 25°C
Solubility in water, 25°C
(Lange, 1973)
201.9°C
0.108 torr
1.8 mg ml'1
1 yg ml"1 (1 ppm) _p_-cresol in water = 9.26 x 10 M
^ Volatilization of £-cresol was not expected to be an important
environmental fate. The estimated Henry's law constant is 36 torr and the
volatilization half-life from a lake is 167 days, as calculated from
equations (5) and (6) of Mackay and Wolkoff (1973).
No literature reference to sorption of p-cresol, phenols, or other
cresols on naturally occurring sediments was found in the literature.
Cowan and White (1962) report data that fit a Freundlich isotherm of phenols
and cresols on complexes of aliphatic amines and bentonite clay.
-------
Phenols are known to oxidize readily under a variety of conditions to
give quinones, polyhydroxybenzenes and eventually, carboxylic acids (Wiberg,
1965; Waters, 1964). Dimeric and polymeric products are also obtained under
some oxidizing conditions. Chemical oxidations of p_-cresol have been repor-
ted to give the oxidative coupling products Pummerer's ketone (I), 2,2'-
dihydroxy-4,4'-dimethylbiphenyl (II), and the trimer III (Musso, 1967).
OH OH OH OH OH
CH
CH
CH3
CH2
II
III
Direct photolysis of p_-cresol in water at 254 nm gave compound II,
2-hydroxy-3,4'-dimethyldiphenyl ether (IV) and 4-methylcatechol (V) (Joschek
and Miller, 1966). The analytical method in this experiment provided only
qualitative identification for these products and could not identify any
nonphenolic products.
IV v
Kinetics of free radical oxidations of phenols, including p_-cresol,
have been summarized by Howard (1972) and include rate constants for reactions
of alkylperoxyl (R02") and alkoxyl (RO-) radicals. Land and Porter (1963)
showed that dimerization of _p_-methylphenoxy radicals proceeds at nearly
diffusion-controlled rate and mostly by £,o^-coupling to give II. Neither
this information nor the limited reports on the direct and photosensitized
oxidation of p_-cresol (Leaver, 1971; Pfoertner and Boese, 1970) or autoxida-
tion of phenolate anion in aqueous alkali (Kirso and Gubergrits, 1972) pro-
vide any information useful in assessing the rates of chemical transformation
of _p_-cresol in aqueous environments.
Many reports have been published on the biological degradation of
phenols, cresols, xylenols, and other "phenolics." A comprehensive review
of this subject or of the degradative pathways is not warranted here, but
some observations will be reported.
Landa et al. (1953) observed that an Escherichia coli cultivated from
Moldavian water degraded phenol at 50-100 mg liter"1 hr~ x, which was two to
five times as fast as the degradation of _p_-cresol. An Oospora culture
degraded phenol, cresols, and xylenol, but at 1/3 to 1/.7 the rate of the
£_._ coli. Phillips and Hinshelwood (1953) adapted Aerobacter aerogenes to
grow on phenol and alkylated phenols. Oka (1962) found that phenols were
absorbed on _E. coli and Staphylococcus aureus at about 22 mM kg"1 cells.
This is not an unexpected observation when inhibition of the bacteria was
occurring. Decomposition of phenol, c>-cresol, toluene, benzene, and other
products by a Pseudomonas and an Achromobacter was reported by Glaus (1964).
-------
The list of types of organisms that will degrade phenolic substances
has been expanded, and there have been many reports on the biological puri-
fication of coke oven effluents by various procedures. Of particular interest
is the publication by Kostenbader and Flecksteiner (1968) in which they repor-
ted 99.9% oxidation of 1300 pounds of phenol in a plant processing 112,000
gallons of effluents per day. Kaplin et al. (1968), in a study carried out
on samples of 5 liters of river water containing 50 ml of effluent from a
coking plant, observed that phenol decomposed immediately but that cresols
required two days of adaptation. The decomposition of p_-cresol could be
represented by the equation C/CO = e~(kt)n where n = 1 and k is the reaction
rate constant = 0.188.
Dagley and Patel (1957) found that with a strain of Pseudomonas, degra-
dation of jp_-cresol was as follows: _p_-cresol -> p_-hydroxymethylphenol ->
p_-hydroxybenzaldehyde
tiydroxybenzoic acid -> protocatechuic acid -*• ring
cleavage. Bayly et al. (1966), in the same laboratories, then described a
Pseudomonas oxidation of pj-cresol to 4-methyl-catechol, which was then cleaved
between carbons 2 and 3. These observations and studies with other phenols
indicate a multiplicity of degradative pathways, depending on the organisms
and compounds.
Although it was apparent that p_-cresol was subject to microbial degra-
dation, our objectives in environmental assessment of the fate of p_-cresol
as a pollutant were to examine different types of waters for their
capacities to develop biodegrading systems, to determine with one culture
mixture the kinetic rate constants that could be applicable in an assessment
model along with rate constants obtained by other disciplines, to determine
any major metabolites and, if warranted, to conduct sorption studies.
4.3 ENVIRONMENTAL ASSESSMENT
4.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigations of the
transformation and transport processes for p_-cresol are summarized in Table
4.2.
s
4.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved £-cresol calculated for individual trans-
formation or removal processes following an acute discharge, such as a
spill, are cited in Table 4.3. Biotransformation (biodegradation) is
clearly the dominant transformational pathway in eutrophic waters, but is
four times slower than photolysis in oligotrophic waters. Dilution is
important compared with biotransformation only in rivers. Sorption and
volatilization are unimportant in all waters.
*
The units of k are not specified in the abstract of the reference.
-------
TABLE 4.2. SUMMARY OF p-CRESOL LABORATORY DATA
Process
Sorption equilibrium
Partition coefficient
Sorption
Photolysis0
Oxidation
Hydrolysis
S = K S
s p w
Rate expression
Volatilization13 k [C]
kp[C] =
k [R02-][C]
OX
NA
Biodegradation k, „ =
max
YK
K = 9.1 ± 6.5'
P
Rate constant at 25°C
= (1.2 ± 0.3) x 10~9 '
k = 6.8 x 10~7 sec"1
P
k =20 M~x sec"1
ox
k, , = 5.2 x 10~7 ml cell"1 hr"1
b/
On Coyote Creek Sediment.
tj
See discussion in Part I, Section 5.3 of the final report.
"Overcast weather in April.
4.3.3 Persistence
£-Cresol is not persistent, as the term is usually applied with respect
to pesticides. Half-lives of p_-cresol in oligotrophic waters are predicted
to be about three months, but are half a day or less in eutrophic waters.
Desorption of £-cresol from sediments is unlikely to be significant as a
source of pollution over the long term.
4.3.4 Mass and Concentration Distributions Calculated Using Computer Models
The pseudo-first-order rate constants used in these simulations are
presented in Appendix A. The distributions of mass and concentration of
jD-cresol expected at steady state during chronic discharge to each of four
types of water bodies are given in Table 4.4.
Steady-state concentrations were not reached in all compartments of
any of the simulations because of the delays in transport of the p_-cresol
10
-------
TABLE 4.3. TRANSFORMATION AND TRANSPORT OF £-CRESOL PREDICTED
BY THE ONE-COMPARTMENT MODEL
Eutrophic Eutrophic Oligotrophic
Process River pond lake lake
Photolysis, half-life (hr)a 4,800 > 10,000 > 10,000 2,400
Oxidation, half-life (hr)
Volatilization, half-life (hr)
Hydrolysis, half-life (hr)
Biotransformation, half-life (hr) 12 12 12 > 10,000
Half-life for all processes, 12 12 12 2,400
except dilution (hr) I
Half-life for all processes 0.55 12 12 2,400
including dilution
Amount jg-cresol sorbed 1 3 0.5 0.5
(mg/m3)b
Percentage _p_-cresol sorbed 0.1% 0.3% 0.05% 0.05%
Estimates are the average photolysis rates on a summer day at 40° latitude.
Photolysis rates in midwinter are at least three times slower.
1 yg ml"1 of p-cresol in aqueous phase and partition coefficient of 10 are
assumed.
solution between compartments in the more complex simulations and the large
storage capacity of the sediments and larger aqueous compartments. However,
extrapolation of the curves suggests that steady-state conditions would be
reached in the simulations within days or weeks in the aqueous compartments,
although months might be required for attainment of steady state in the sedi-
ments. These patterns are clearly seen in the pond simulation (Figure 4.1)
in which steady state in the aqueous compartment is nearly reached within
4 days, while the mass of £-cresol in the sediments at the end of 4 days is
roughly 3% of the mass expected at steady state (Table 4.4). Steady states
are obtained 40 times faster in the river simulation, where dilution is
extremely rapid (Figure 4.2), but are obtained 100 to 1000 times more
slowly in the stratified lake simulations, in which transport into the
deeper waters and sediments is relatively slow (Figures 4.3 to 4.4).
The concentrations of £-cresol observed at steady state varied signi-
ficantly among simulations and among compartments within each of the
11
-------
simulations (Table 4.4). Concentrations on the sedimented and suspended
solids were generally 10 times higher than in solution, and occasionally
as little as 3 times higher. Aqueous concentrations are highest in the
river simulation since removal from solution is slow relative to the import
and export of p_-cresol from the simulated river segments. Indeed, under the
assumptions of the simulation in which the stream volume remains constant,
the concentration of ja-cresol decreases only 1% within each kilometer-long
segment. In contrast, in the eutrophic lake simulation in which the flows
of the jo-cresol solution are much slower, there is a 3- to 10-fold difference
in concentrations of dissolved p_-cresol between adjacent compartments in the
surface waters, and nearly a 150-fold decrease in concentration between
the surface and bottom waters in the lake center.
2 x 10"
10
-5
150
200
250
TIME - hours
FIGURE 4.1 PERSISTENCE OF p-CRESOL IN A TWO-COMPARTMENT POND SYSTEM
12
-------
TABLE 4.4. DISTRIBUTION OF 2~CRESOL IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(Input concentrations of 1 ug ml"1 p-cresol)
Pond
Mass Cone.
(kg) (pg g"1)
Compartment 1
(surface water)
Solution 3.40 x 10"1 1.70 x 10"2
Suspended solids 9.55 x 10"3 1.70 x 10~l
Compartment 2
(surface water)
Solution
Suspended solids — —
Compartment 3
(surface water)
Solution
Suspended solids
Compartment 5
(bottom water)
Solution
Suspended solids
Compartments 7-9
(sediment)3
Solution 4.25 x 10"2 1.70 x 10~2
Solids 1.14 x 1Q-1 1.70 x lO"1
River
Mass
(kg)
2.94 x 102
2.94 x 10"1
2.89 x 102
2.89 x 10"'
2.85 x 102
2.85 x 10'1
—
—
2.41
1.95 x 102
Cone.
(MB g"')
9.80 x 10"1
9.80
9.66 x 10"1
9.66
9.50 x 10"1
9.50
—
—
9.65 x 10"1
9.65
Eutrophic lake
Mass
(kg)
9.44
4.72 x 10~3
6.02
3.01 x 10~3
2.49 x 10"1
1.24 x 10"*
7.22 x ID"2
3.61 x IS"5
8.75 x 10"2
1.30
Cone.
(ug g-1)
~3777
3.77
2.40
2.40
9.99
9.99
2.89
2.89
9.58
9.58
x 10~2
x 10" *
x 10-3
x 10-2
x 10"*
x 10-3
x 10-5
x 10"*
x ID'3
x 10- 2
Oligotrophic lake
Mass
(kg)
5.32 x
2.66 x
3.55 x
1.60 x
3.00 x
1.50 x
5.25 x
2.62 x
8.29 x
1.29 x
10
10"2
102
10" '
10
10~2
10
ID'2
10- l
10
Cone.
2.20 x 10"1
2.20
1.28 x 1Q-1
1.28
1.20 x 1Q-1
1.20
2.10 x 10"2
2.10 x 10"1
9.48 x 10"2
9.48 x 10"'
aThe amounts given for solid and solution phases in the sediment compartments are estimated from the sorption partition coefficient
for suspended solids and may be overestimated because it was assumed that biodegradation of sorbed material does not occur.
-------
a.
i
cc
o
a I
01
o
z
o
o
5 x 10"
456
TIME - hours
FIGURE 4.2 PERSISTENCE OF p-CRESOL IN A PARTIALLY MIXED RIVER SYSTEM
14
-------
5 x
E
01
a.
i
ui
5
o-\
z
o
tr
01
u
o
o
100
200
300 400
TIME - hours
500
600
700 720
FIGURE 4.3 PERSISTENCE OF p CRESOL IN A PARTIALLY MIXED EUTROPHIC LAKE
15
-------
5 x 10'1
SOLUTION
— SEDIMENTS
10
100
200
300 400
TIME - hours
500
600
700 720
FIGURE 4.4 PERSISTENCE OF p-CRESOL IN A PARTIALLY MIXED OLIGOTROPHIC LAKE
16
-------
Concentrations in the oligotrophic lake simulation are at least 3 to
10 times those of the eutrophic lake even though the higher of the two
photolysis rate constants is assumed. Use of the lower rate constant would
increase the difference between the two lake types. Changes in concentration
between compartments of the oligotrophic lake are also about 3- to 10-fold
in the surface compartments of the oligotrophic lake simulation and about
5-fold between the upper and lower water compartments of the lake center.
However, these differences would decrease if the lower rate constant were
used, and degradation would be correspondingly slower relative to dispersal
by currents.
Once discharges stop, the concentrations of £-cresol in solution
decline rapidly by five to six orders of magnitude to new steady states in
the river and eutrophic lake simulations. Concentrations in the pond appear
to be declining similarly, although the simulation was stopped prematurely.
Concentrations of jp_-cresol -on the sediments decline slowly in all simulations
because of the slow scouring of sediments.
Patterns of variation for suspended solids are similar to patterns
of dissolved jv-cresol, but the concentrations are roughly 10 times those of
dissolved p-cresol in the same, compartments, except in the river simulation.
Here concentrations in solution and on suspended solids are about the same.
For simplicity, the data for suspended solids are presented graphically
only for the pond simulation.
In summary, the steady-state concentrations resulting from an inflow
of solution containing 1 yg ml"1 of £-cresol are very nearly 1 yg ml"1 in
the initial segments of the simulated stream, roughly 0.1 yg ml 1 in the
oligotrophic lake, about 0.04 yg ml"1 in the eutrophic lake, and approxi-
mately 0.02 yg ml"1 in the pond. The river can, of course, accommodate a
much larger mass of p-cresol without a significant change in concentration
due to rapid dilution, but when contaminated by a steady discharge, the
concentration downstream changes very slowly indeed, resulting in the con-
tamination of very large volumes of water.
4.3.5 Discussion
Biodegradation is likely to be the dominant removal mechanism, account-
ing for at least 90% of the removal of £-cresol from solution, even if allow-
ance is made for the variability in the importance of photolysis and sorption
as a function of variations in turbidity, water color, or temperature. While
photolysis may vary as much as tenfold as the pH rises to pH 9 (the upper
limits of the values commonly encountered in natural waters), it is unlikely
that these high pHs will occur in clear lakes. Hence, biodegradation may be
dominant even in these waters.
Hence, the concentration of j>-cresol, given constant inflows, should
exhibit considerable seasonal variation, with lowest concentrations occurring
in the summer months (assuming a constant volume of the water body). Conse-
quently, excluding the effects of changing stream volume, the hypothetical
discharge of about 1 metric ton per hour assumed in our simulation would
lead to concentrations between 10 and 1000 ng ml'1 over a 100-km stream segment
17
-------
in the summer months, and roughly the same concentration range over a 1000-
km stream segment in the winter. These decreases in concentration may indeed
occur due to dilution with inflowing, unpolluted waters, but are unlikely to
occur within most streams through transformation.
The risks associated with large discharges of £-cresol in lakes are
higher because the potential for dilution is much lower due to slower rates
of mixing. However, if the same assumption of a 1 yg ml"1 discharge is made,
the concentrations of £-cresol that can be expected in the various compart-
ments of eutrophic and oligotrophic lakes are lower than in the river simula-
tion as long as inputs of solution are of reasonable size. Hence, while the
concentrations of p-cresol in lakes should be relatively low, the concentra-
tions in shallow waters, particularly under ice cover, when both photolysis
and biodegradation are suppressed, may rise to levels approximating those of
the stream simulation.
The biota exposed to the highest concentrations will tend to be
organisms that live or feed in streams or shallow waters of lakes. At least
some of these organisms are likely to be exposed to toxic concentrations
under the conditions assumed in the simulations, since LD50s for fish and
invertebrates are generally in the range of 5 to 15 ppm (McKee and Wolf,
1963). Hazards to humans and other mammals appear to be low (von Rumker
et al., 1975; Christensen and Luginbyhl, 1975), since the LDsos for the rat
and mouse are 207 and 344 ppm, respectively, for orally ingested p-cresol.
4.4 PHYSICAL PROPERTIES
4.4.1 Solubility in Water
The solubility of £-cresol in pure water is reported in the literature
to be 1.8 mg ml"1 (Lange, 1973).
4.4.2 Absorption Spectra
The ultraviolet (UV) absorption spectra of 4.19 x 10"'* M (45.3 yg ml~l)
p_-cresol in pure water at pH 5, 7, and 9 were measured on a Gary Model 11
spectrophotometer. The absorption coefficients at wavelength intervals from
297.5 to 330 are reported in Table 4.5. The absorption spectrum of £-cresol
depends on pH because _p_-cresol (pK = 10.2) dissociates at higher pH and
the anion is a more strongly absorbing species.
4.4.3 Volatilization Rate
The volatilization rate of £-cresol was measured using the method of
Hill et al. (1976), which is described in detail in Part I, Section 5.3 of
this report. The volatilization half-life of p_-cresol in aqueous solution
was about 160 hours at a moderate stirring rate. The rate constants were
calculated from two experiments as shown in Table 4.6. Although these are
measurable volatilization rates, additional measurements were not made
because biodegradation was so fast.
18
-------
TABLE 4.5. ABSORPTION SPECTRA OF p-CRESOL IN PURE WATER3
Center of
wavelength
interval" (nm)
Average absorption
coefficient0 (M"1 cm"*
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330
pH 5.1 pH 7.0 pH 8.9
14 18 193
3.8 7.2 173
2.4 3.8 150
2 2 124
i 0 2 92.6
2 63.5
1 40.3
0 24.1
13.0
6.2
3.8
0
a£-Cresol concentration is 4.19x10 " M (45.3 pg ml 1).
The wavelength intervals are given in Appendix B, Table B.I.
°The absorption coefficient at 313.0 nm is 28.6 at pH 8.9.
4.4.4 Sorption on Clay and Sediments
The partition coefficients of £-cresol an Ca-contmorillonite clay
and on Coyote Creek sediment were measured. The data were fitted to a
Freundlich isotherm; from the literature data (Cowan and White, 1962), we
assumed that n was equal to 1 (see Part I, Section 5.4.2.3).
The experiments consisted of duplicate flasks at one p_-cresol concen-
tration and two clay loadings (for the calcium clay), and at two £-cresol con-
19
-------
TABLE 4.6. VOLATILIZATION RATE CONSTANTS
_ _ FOR p-CRESOL _
Oxygen p_-Cresol
reaeration volatilization
0
rate, k,
(hr"1)
rate, k
(hr"1)
Ratio
kg/kg
Experiment 1 3.55 ± 0.24 (3.38 ± 0.67) x 10 3 (0.95 ± 0.25) x 10 3
Experiment 2 2.97 ± 0.21 (4.50 ± 0.39) x 10~3 (1.5 ± 0.24) x 10~3
Average (1.2 ± .3) x 10~3
centration and two sediment loadings (for the Coyote Creek sediment). The pH
was adjusted to 6.0 ± 0.2 with dilute hydrochloric acid. The equilibration
time for the initial experiments was 18 hours.
The initial experiments with natural sediments suggested that the
partition coefficient Kp was between 50 and 300. However, we suspected that
biodegradation might be occurring, which would lead to erroneously high Kp
values. Therefore, to minimize biological activity, all solutions were
sparged with nitrogen before the jj-cresol was added and the p_-cresol was
exposed to the sediment for only 1 hour. The 1-hour exposure was selected
because the cascade experiments using natural water, showed that the time
required for acclimation to p^-cresol was about 6 hours (Section 4.6). While
it would have been possible to sterilize the samples by autoclaving or adding
ethylene oxide, we chose not to do this because the nature of the sediment
might have been altered. The results are summarized in Table 4.7.
The studies described in Section 4.6 showed that biodegradation of
£-cresol is very fast. Therefore, we did not pursue the sorption studies.
4.4.5 Biosorption
Because jg-cresol so readily biodegraded (Section 4.6),
studies were not undertaken.
biosorption
4.5 CHEMICAL TRANSFORMATION
4.5.1 Photolysis Rate
Data for the photolysis of £-cresol in pure water and in water con-
taining 9.5 ug ml"1 humic acid are given in Table 4.8. The initial concen-
trations of £-cresol in these experiments ranged from 0.80 to 0.60 Mg ml"1.
With one exception, solutions maintained in the dark as controls during the
time period of the photolyses showed less than 2% loss of p-cresol if the
glassware and water were sterilized before use. The rate constants in Table
20
-------
TABLE 4.7. p-CRESOL SORPTION ON SEDIMENTS
Total
organic
carbon
Sediments (percent)
Ca-
Montmorillonite 0.06
clay
sorption
Coyote
Creek 1.4
sorption
p-Cresol p_-Cresol
Sediment concentration concentration
Q
concentration in supernatant on sediment Partition coefficient , K
(yg ml~l) (yg ml~l)a (yg ml'1)13 ao = ° ao * ° P
0.30 0.12 ± 0.00 2.14 ± 4.62 16 ± 14 -490 ± 740
._—
3.0 0.12 + 0.00 1.55 ± 0.29 (a = 60 ± 87)
o
4.2 4.3 + 0.1 0.0 ± 0.2 -1.7 + 2.4 9.1 ± 6.5
12 4.1 ± 0.1 8.0 ± 6.2 (aQ = -41 ± 62)
7.8 7.3 ± 0.3 56 ± 31
24 7.6 ± 0.1 6.5 ± 6.0
Concentration measured in supernatant with population standard deviation.
'Concentration on sediment calculated from supernatant concentration with population standard deviation.
Calculated by linear least squares method; see Appendix B, Section B.I.4. for description of regressions.
Limits are 95% confidence limits.
1-hour equilibration time.
-------
4.8 were calculated assuming that the reactions were first order in £-cresol,
which was found for photolyses beyond two half-lives using the borosilicate
filter system. The quantum yield for p-cresol disappearance, 0.079, was
calculated from the rate constant kp measured for the direct photolysis of
jg-cresol at 313 nm.
The higher photolysis rate of _p_-cresol in the experiment using the
borosilicate glass (immersion well only) filter system is consistent with
the greater light intensity transmitted by this filter compared with the
313-nm filter system. From the data for photolyses conducted in April sun-
light, the measured half-lives of jD-cresol are 35 days in pure water and
about 3 days in solutions containing humic acid.
The half-life for direct photolysis of £-cresol in sunlight as a
function of the time of the year was calculated by the procedure of Zepp and
Cline (1977) using the quantum yield of 0.079 and the measured UV absorption
spectrum of ja-cresol (Section 4.4.2). These data are plotted in Figure 4.5.
TABLE 4.8. RATE CONSTANTS FOR PHOTOLYSIS OF 1.0 ug ml"1 p-CRESOLa
Irradiation
source
313 nm
Pyrex filter
only
Solar, in
April
Ratio of k_ with
Extent of Rate constant humic acid to that
Solution reaction (%) k x 106 sec"1 in pure water
Pure water 28 0.72 ± 0.04b'C
2.6
Pure water with
9.5 ug ml~' humic 40 1.84 ± 0.04
acid
Pure water 81 11.7 ± 1.6
4.8
Pure water with
9.5 iig ml'1 humic 93 56 ± 1
acid
Pure water 32 0.68 ± 0.12 >e
12
Pure water with , ,
9.5 us ml"1 humic 93 8.0 ± 0.2 '
acid
' 1.0 ug ml~' p-cresol in water = 9.2 10 * M.
Standard deviation.
Quantum yield for £-cresol disappearance was 0.079.
Calculated assuming 8 hours of sunlight per day; weather was mostly overcast during the two-week
reaction period. To obtain average ratio constant for full calendar day (24 hours), divide rate
constant by three.
PHalf-life of 35 days.
fHalf-life of 3.0 days.
22
-------
Figure 4.5 shows that the computer-calculated half-life for direct
solar photolysis of £-cresol in April is about 400 days, based on a UY spec-
trum with a cutoff at 315 nm (see Table 4.5). However, a hand calculation
including the absorption tail beyond 315 nm estimates a half-life of about
200 days. The experimentally measured half-life for the solar photolysis of
£-cresol (Table 4.8) is about 35 days. The difference between the calculated
and measured half-lives cannot be explained by a pH effect on the absorption
coefficient, since the measured half-life was obtained in a solution that was
pH 6.3 and the calculation is based on a UV spectrum measured in pH 7.0
solution (Table 4.5).
*
Since some loss of £-cresol (7% in 8 days ) was found in the dark
control solution for the direct photolysis in sunlight, some biodegradation
may have occurred in spite of the precautions taken to maintain sterile condi-
tions. If it is assumed
that biodegradation was occurring simultaneously
with the direct photolysis and that both processes are first order in p-cresol
(and that the organism cokmt remains constant), subtracting out the contribu-
tion from biodegradation gives a corrected photolysis half-life of about 70
days. This value is still a factor of three less than the calculated 200 day
half-life. It is important to note that the computer calculation method has
been shown to be quite reliable for other compounds where the tail of the
absorption spectrum does not contribute significantly to the absorption rate
constant, k .
a
The data in Table 4.8 show that the presence of humic acid in solution
accelerates the photolyses of £-cresol by factors of 2 to 12 over those in
pure water. Humic acid absorbs essentially all the light that leads to trans-
formation of _p_-cresol in these photolyses; the absorbance of humic acid at
313 nm is about 0.6, compared with 10~5 calculated for 1 jig ml"1 jv-cresol,
assuming a 1—cm effective path length in the reaction tube. The difference
in the rate factors for the three experiments is certainly due in part to
the different spectral distributions of the light sources used and to the
capacity of humic acid to absorb light to beyond 400 nm. It may also be due
to a wavelength dependence for the quantum efficiency for the sensitized
photolyses. Wavelength dependence has been observed for photolyses of para-
thion in natural waters containing humic substances (N. L. Wolfe, private
communication, 1976).
The acceleration in rate by humic acid may be due to a sensitized
photolysis in which humic acid serves to transfer excitation energy to
_p_—cresol; humic acid could act as a photoinitiator, a triplet diradical for
example, which initiates a radical oxidation process. Other possible roles
for humic acid, such as a screening agent or quencher, are apparently not as
important in this system. The data do not permit any distinction between
these possible reaction pathways.
The actual elapsed time was 8.0 days, corresponding to 64 sunlight hours.
Using k = 6.81 x 10~7 sec"1 (Table 4.8), the calculated loss of £-cresol is
15%. P
23
-------
KJ
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MONTH OF YEAR
SA-4396-38
FIGURE 4.5 ANNUAL VARIATION OF PHOTOLYSIS HALF-LIFE FOR p-CRESOL
-------
4.5.2 Oxidation Rate
The susceptibility of £-cresol to free radical oxidation was examined
using the oxidation reaction initiated by 4,4'-azo bis(4-cyanovaleric acid)
(AA) (see Appendix B for discussion and procedure). Oxidation reactions
with 1.0 x lO"*1 M AA were carried out both in pure water and in pure water
that contained 9.5 u-g ml"1 humic acid; the data are presented in Table 4.9.
TABLE 4.9. FREE RADICAL OXIDATION OF 0.72 yg ml"1
p-CRESOL AT 50°C AND 100 HOURS3
Solution
Initial cone.
_p_-cresol
I (yg ml'1)13
Final cone.
£-cresol
(yg ml"1)13
% Reaction
Water with 10 * M AA
Water with
9.5 g ml"1 humic
acid and 10~A M AA
0.72
0.73
0.58 (0.71)
0.53 (0.72)'
20
27
Based on one analysis of reaction at 100 hours.
1.0 yg ml"1 £-cresol in water = 9.2 x 10~6 M.
f+
Control solution without AA initiator maintained under identical reac-
tion conditions.
From the data in Table 4.9, we obtained a rate constant term (kox[R02-J)
of 6.0 x 10~7 sec"1 for oxidation of £-cresol in water under the reaction
conditions. Under these conditions (see Part I, Section 6.3), this corresponds
second-order rate constant k,,,, of 180 M"1 sec"1,
to a
be 20 M"1 sec"
At 25°C, kox would then
and with the assumption that [R02- ] = 10~9
M in the environ-
ment, the half-life of _p_-cresol toward oxidation is about 1 year (first-order
rate constant of 2 x 10~8 sec"1).
This value of kox is considerably less than ^10** M J sec"1 observed
for a number of phenols in organic media (Howard, 1972). The difference may
be due to the hydrogen bonding ability of water, which could reduce the reac-
tivity of the phenolic hydrogen towards abstraction. Alternatively, because
of the low radical concentrations, the difference may be due to a more complex
reaction sequence than anticipated.
Another important observation is the ineffectiveness of 9 yg ml"1 humic
acid in the radical oxidation process. Phenolic OH, usually present in humic
acids, would be expected to compete with £-cresol for R02- and inhibit the
rate of _p-cresol oxidation. The simplest explanation for lack of a retarding
25
-------
effect is that phenolic OH in humic acid is internally bound (H-bonds or
metal complex) and is unreactive or inaccessible to R02-.
In a separate study of the AA-initiated oxidation of cumene (i-propyl-
benzene) in water from Coyote Creek, cumene was oxidized readilyy indicating
that dissolved natural organics did not scavenge all the radicals (Mill et
al., 1977). Additional studies of oxidations in natural waters, supported by
NSF-RANN are now under way.
4.5.3 Hydrolysis Rate
jj-Cresol contains no functional groups that are hydrolyzable; therefore,
no hydrolysis studies were carried out.
4.5.4 Products from Chemical Transformations
Although oxidation and photolysis both showed transformation of
j)-cresol, only limited studies of these processes were conducted. Biodegrada-
tion was obviously the dominant transformation process, as was evident by the
biodegradation problems that interferred with the initial chemical transfor-
mation studies (Section 4.6). Detailed product studies would have required
significant experimental work and therefore were not undertaken.
Information on products was not available from kinetic experiments
since some products could not be extracted and others would not elute in the
GLPC work-up used for analysis of the £-cresol solutions. One exploratory
experiment to analyze for products from oxidation of 7 x 10~5 M p_-cresol
(8 ug ml"1) containing 10~A M AA was carried out and analyzed by high pressure
liquid chromatography (HPLC). A photoinitiated free radical oxidation was
induced by photolyzing AA with 366 nm light (where only AA absorbs). The only
products identified were Pummerer's ketone (I) and 2,2'-dihydroxybiphenyl (II);
these were identified by comparing retention time with those of samples pre-
pared by literature procedures (Haynes et al., 1956).
These two products accounted for only about 25% of the reacted ja-cresol, and
after an initial photoperiod, the biphenyl (II) began to decrease with longer
irradiation.
More polar products were evident in the HPLC traces but could not be
separated for certain identification. These may be the more polar hydroxylated
and ring cleavage products and [R02%] adducts of p_-cresol.
26
-------
4.6 BIODEGRADATION
4.6.1 Development of Biodegrading Cultures
£-Cresol was one of the first and also the most readily degraded of the
compounds studied. For these reasons, there was considerable experimentation
with the enrichment and kinetic techniques. Enrichments were first initiated in
the 9-liter bottle fermentors containing 4 liters of water, £-cresol, and 1 liter
of sterile 0.05% NHitN03 or (NHtt)2SOit and 0.2% KHaPO^-KaHPOi* mixture that
maintained the pH of the fermentation at approximately 7.0. When it became
obvious that enrichment £-cresol biodegrading systems could be developed
readily, initial studies with water samples were sometimes conducted in
Erlenmeyer or Fernbach flasks and the water samples were diluted as much as
100%. Incubations were at 25°C except for the Lake Tahoe water samples,
which were incubated at 15°p. Initial £-cresol concentrations were 3 yg ml"1
or 10 yg ml"1. For some ancillary studies, the initial levels were as high
as 1000 yg ml"1. ;
\
A very simple, rapid preliminary test to monitor p-cresol biodegrada-
tion consisted of adding 1 drop of 10% aqueous H3POi, to a 1- or 2-ml sample,
adding 2 ml ethyl acetate, vortexing these contents in a test tube, removing
the clear solvent layer, and determining the UV absorbance at 260, 280 and
300 nm. The differences in absorptions at 280 nm and those at 260 and 300 nm
were compared with those obtained with standard p_-cresol solutions. These
wavelengths were chosen because £-cresol has an absorption maximum at 280 nm,
aromatic metabolites are expected to absorb at 260 nm, and neither should
absorb significantly at 300 nm.
We were unable to obtain water samples from two western coking plants
or a wood-creosoting plant. Enrichment biodegrading systems were readily
obtained with the water samples from Coyote Creek, the pond near Searsville
Lake, aeration effluents from the Palo Alto and South San Francisco sewage
plants, the aeration effluent from the Shell Oil Refinery wastewater treat-
ment plant, and Lake Tahoe.
There were invariably induction or lag periods before £-cresol bio-
degradation began. In one series of experiments with pond water diluted to
contain 45% pond water with 3 and 10 yg ml"1 £-cresol, no significant bio-
degradation occurred until 5.5 and 6 hours incubation of the respective
concentrations of pollutant, and degradation was completed at 7 and 8.5
hours, respectively (Figure 4.6). With a 40% v/v dilution of aeration
effluent from the Palo Alto sewage plant, total degradation at the 3 and
10 yg ml"1 £-cresol levels occurred at <4 and 6 hours, respectively. In one
experiment with 75% pond water and extreme levels of 100, 300, and 1000 yg ml
£-cresol, the respective degradations and times were 100% at <24 hours, 100%
at approximately 32 hours, and 30% at 72 hours. Undoubtedly, at levels
approaching 1000 yg ml"1 £-cresol, many organisms were totally inhibited
or killed and some were retarded in growth. No significant degradation was
observed in the 1000 pg ml"1 flasks until they had been incubated for 48
hours.
27
-------
0
4 6
TIME (hr)
8
10
SA-4396-32
FIGURE 4.6 BIODEGRADATION OF £-CRESOL IN EUTROPHIC POND WATER
Initial concentrations of £-cresol were 3 and 10 fig ml"1. Degradation
was monitored by uv (O) and gc (D) assays.
28
-------
With 80% Lake Tahoe water and with 80% Lake Tahoe water that had been
mixed with sediment and the sediment was allowed to settle, there was total
decomposition of 10 yg ml"1 p_-cresol in <144 and <96 hours, respectively.
The respective bacterial counts for these Lake Tahoe samples were 12 and 61
cells ml"1. We believe that these counts were too low and that other pro-
cedures such as MPN should have been used.
In subtransferring initial culture mixtures, no problems were encoun-
tered in using p-cresol as the only carbon source. There was also no diffi-
culty in developing the pond water culture mixtures to the level that they
decomposed 600 yg ml"1 p_-cresol in 24 hours. Culture mixtures from other
enrichments were not subtransferred in media containing more than 100 yg ml"1
p-cresol. In fact, as will be discussed later, because of the rapid degrada-
tion of _p_-cresol at lower levels, there could be many organisms that grow on
release or autolysis products or other bacteria that are primary or secondary
utilizers of p_-cresol.
A multitransferrec
£-cresol enrichment culture was streaked out on
basal salts/p_-cresol agar, and many types of cells were observed. Seven
very obvious types were selected from a larger number of variants, and after
transfers, they were tested for growth rates on 600 yg ml"1 p_-cresol media by
optical density measurements at 600 nm. All seven isolates had comparable
growth rates.
4.6.2 Biodegradation Kinetics
Two types of kinetic studies were conducted: batch fermentations with
low-level inocula and continuous fermentations in chemostats.
Batch Fermentations with Low-Level Inocula—Figures 4.7, 4.8, and 4.9
present the data from a batch fermentation with a low-level of inoculum of
p_-cresol decomposing organisms developed from a eutrophic pond. Figures 4.7
and 4.8 are plots of viable counts in flasks with initial p_-cresol levels of
0, 3, 5, and 0, 10, 20 yg ml"1, respectively. The increase in counts in the
control flask with no p_-cresol is attributed to the fact that the viable
organisms were metabolizing and excreting products, and dead cells or cells
that were dying in this mixed culture were autolysing and releasing products
that other organisms could use for survival and multiplication. The possi-
bility also exists that trace quantities of metabolizable organic matter may
be introduced in the distilled water or from the atmosphere (Kayser et al.,
1975). Because ]3-cresol is so readily metabolized, increases in cell counts
in media with this compound are most likely due to good survival and growth
of £-cresol utilizers (primary and secondary).
The specific growth rate (y) in each medium was determined by measuring
the slope of the curves in the logarithmic phase of growth. For the media
initially containing 3, 5, 10, and 20 yg ml"1 p_-cresol,y values were 0.58,
Most Probable Number, as defined in "Standard Methods for Examination of
Water and Wastewater," 13th edition, p. 635.
29
-------
0
8 12
TIME (hr)
SA-4396-33
FIGURE 4.7 VIABLE CELL COUNTS IN 0, 3, and 5/^ mll1 p-CRESOL
BIODEGRADATION BY BATCH FERMENTATION WITH A
LOW-LEVEL INOCULUM
Initial £-cresol concentrations: 0 (O), 3, (D) and 5 (A)/^g ml'1.
30
-------
0
8 12
TIME (hr)
SA-4396-34
FIGURE 4.8 VIABLE CELL COUNTS IN 0, 10, AND 20 ^g mMp-CRESOL
BIODEGRADATION BY BATCH FERMENTATION WITH A
LOW-LEVEL INOCULUM
Initial £-cresol concentration: 0 (A), 10 (O) and 20 (Q) fig ml'1.
31
-------
2 -
0
6 8 10
TIME (hr)
14 16
SA-4396-35
FIGURE 4.9 p-CRESOL BIODEGRADATION BY BATCH FERMENTATION WITH
A LOW-LEVEL INOCULUM
Initial p-cresol concentration: 3 (D), 5 (A), 10 (O) and 20 (®) fjg ml 1.
32
-------
0.61, 0.63, and 0.67 hr 1, respectively. Figure 4.9 shows that most of the
£-cresol has been utilized by 14 hours, even though Figures 4.7 and 4.8 indi-
cate significant increases in cell counts between 14 and 16 hours.
Table 4.10 gives the calculated cell yield (Y) values for cell growths
in media with different initial £-cresol concentrations and at different
periods in the fermentations. The average Y values obtained from the time
significant utilization of ]3-cresol could be measured up to 14 hours of
incubation were used in subsequent calculations of rate constants. Figure
4.10 is a Lineweaver-Burk plot developed by the least squares method. The
values of Ks and ym, determined by the intercept of the ordinate axis and
the slope of the curve in Figure 4.10, were calculated as 0.69 yg ml"1 and
0.69 hr"1, respectively.
TABLE 4.10. YIELD OF CELLS IN A p_-CRESOL BATCH FERMENTATION
! WITH A LOW-LEVEL INOCULUM3
1
SQ = 3 yg ml"1
AS (yg ml"1)
X (cells x 10~6
Y (cells x 10~6
SQ = 5 yg ml"1
AS (yg ml"1)
X (cells x 10~6
Y (cells x 10~6
SQ = 10 yg ml"1
AS (yg ml"1)
X (cells x 10~6
Y (cells x 10~6
S = 20 yg ml"1
AS (yg ml"1)
X (cells x 10~6
Y (cells x 10~6
I ••
10 hr
—
ml"1)
yg ip-cresol) —
—
ml"1)
yg ^-cresol) —
—
ml"1)
yg ip-cresol) —
1.5
ml"1) 2.0
yg ^-cresol) 1.3
Incubation
12 hr
1.4
2.5
1.8
2.5
3.5
1.4
3.2
3.7
1.2
7.0
6.5
0.9
time
14 hr
2.7
7.4
2.7
4.6
9.0
2.0
8.6
12
1.4
19.3
25
1.3
Average
yield
2.2
1.7
1.3
1.2
a9.8 x 103 cells ml l of a biodegrading system developed from eutrophic
pond water. SQ is initial concentration of jg-cresol, AS is difference
in concentration after incubation, X is bacterial cell count, and Y is
the cell yield.
33
-------
Ks = 0.69 //g ml'1
- 0-69
8 12
S (fig mH p-cresol)
16
20
SA-4396-36
FIGURE 4.10 LINEWEAVER-BURK PLOT OF DATA FROM
p-CRESOL BIODEGRADATION IN A BATCH
FERMENTATION STUDY
34
-------
Table 4.11 presents the calculated Y and first-order rate constant
values obtained in this batch fermentation study with low levels of
inoculum of a £-cresol degrading system. The second-order rate constant
kb2 = kbl/Ks equais 6.7 x 10~7 ml cell""1 hr"1.
TABLE 4.11. FIRST-ORDER RATE CONSTANTS (kb) IN A £-CRESOL BATCH
FERMENTATION WITH A LOW-LEVEL INOCULUM3
Initial cone. S Cell yield (Y)
(pg ml"1) (cells vig"1)
3 2.2 x 106
5 ; 1.7 x 106
10 '; 1.3 x 106
20 1.2 x 106
Average (1.6 ± 0.5) x 106
kb = ym/Y
(MB cell"1 hr"1)
3.1 x 10~7
4.1 x 10~7
5.3 x 10~7
5.8 x 10~7
(4.6 ± 1.2) x 10~7
9.8 x 103 cells ml"1 of a biodegrading system developed from eutrophic
pond water.
Biodegradation in a p-Cresol Continuous Feed Fermentation in a Chemo-
stat—Table 4.12 presents data from a fermentation in which a basal inorganic
salts/100-pg ml"1 p_-cresol medium was continuously fed to a fermentation in
a chemostat. The Lineweaver-Burk plot determined by the least squares method
is presented in Figure 4.11. From these data, Ks was calculated as 0.99
yg ml"1 and ^m was calculated as 0.56 hr"1. The second-order rate constant
kt,9 is 3.7 x 10~7 ml cell"1 hr"1.
The um, KS, Y, kb, and kb2 values obtained by the two procedures are
compared in Table 4.13.
4.6.3 Metabolites
The p_-cresol degradation took place so rapidly that no metabolites
were observed.
4.6.4 Discussion
Enrichment culture systems that would biodegrade £-cresol were readily
obtained from all.waters after relatively short induction periods. It was
also relatively easy to develop culture systems that could decompose high con-
centrations of p_-cresol when it was the only carbon source. Under microscopic
35
-------
TABLE 4.12. KINETICS OF A £-CRESOL BIODEGRADATION IN A CONTINUOUS FERMENTATION
Specific Outflow
growth rate cone. (S)
GO (hr-1) (ug ml-1)
0.56
0.53
0.50
0.48
0.47
0.45
0.44
0.42
0.40
0.38
46.4
26.9
18.0
14.9
9.72
5.84
1.76
1.24
0.76
0.26
S/u
(calcu-
lated)
82.9
5008
36.0
31.1
20.7
13.0
4.0
2.95
1.90
0.68
Microbial
count (X)
(cells ml"1
4.0 x 107
9.0 x 107
1.8 x 108
2.0 x 108
2.8 x 108
2.6 x 108
3.2 x 108
Average
k - um/v
Cell yield (Y) *T> m/ x
) (cells ug-1 MP) (ug cell"1 hr"1)
7.4 x 106 7.6 x 10~7
1.1 x 106 5.1 x 10"7
2.0 x 106 2.8 x 10"7
2.1 x 106 2.7 x 10"7
2.9 x 106 1.9 x 10~7
2.6 x 106 2.2 x 10~7
3.2 x 106 1.75 x 10~7
(2.1 ±0.9) x 105 (3.4 ± 2.2) xlO~
examination and in a plating-out study it was obvious that many types of cul-
tures were present after many passages on liquid media. It is not known how
many of these were primary utilizers of £-cresol, or secondary utilizers that
grew on metabolites, or how many grew on excretion products associated with
living or dying p-cresol metabolizers.
The agreement in the um, Ks, Y, k^, and k{,2 values obtained by the two
kinetic procedures (Table 4.13) is excellent for mixed and even pure cultures.
There were difficulties with film formation (Atkinson and Fowler, 1974) in the
continuous feed studies. Also, these kinetic* studies were conducted at such
different periods that there could have been changes in proportions of differ-
ent strains in the microbial mixture, particularly in the prolonged chemostat
fermentations. These values cannot be compared with those reported for phenol
by Yang and Humphrey (1975) because, although ja-cresol and phenol have similar
structures, they would obviously have different metabolic processes and
because Yang and Humphrey's studies with phenol were conducted at levels that
were inhibitory to the microorganisms they used.
No attempts were made to determine optimal degradation conditions or to
develop or select strains or mixtures of strains with high degradation rates.
The development of degrading culture mixtures that are more exclusively asso-
36
-------
20 30 40
p-CRESOL S - //g ml'1
SA-4396-37
FIGURE 4.11 LINEWEAVER-BURK PLOT OF DATA FROM
p-CRESOL BIODEGRADATION IN A CONTINUOUS
FERMENTATION
37
-------
TABLE 4.13. COMPARISON OF KINETIC DATA OBTAINED ON BIODEGRADATION
OF p-CRESOL BY TWO PROCEDURES
Um Ks Y kb kb?
(hr"1) (yg ml"1) (cells yig"1) (yg cell"1 hr~x) (ml cell~r hr"1)
Batch fermen-
tation with 0.69 0.69 1.6 x 106 4.6 x 10~7 6.7 x 10~7
low inoculum
chemostat 0.56 0.99 2.1 x 106 3.7 x 10~7 3.7 x 10'
Continuous
chemostat
fermentation
Average 0.62 0.84 1.8 x 106 4.2 x 10~7 5.2 x 10"
ciated with the primary and secondary utilization of p-cresol would have
required very frequent, short-interval transfers before all p_-cresol had
degraded, but this would have been impossible without continuous 24-hour
monitoring. This could be avoided, however, by using a continuous fermenta-
tion in which a desired level is maintained, similar to the phenol-stat
chemostat used by Yang and Humphrey (1975).
During the course of our p_-cresol degradations, no breakdown products
were detected. This suggests that, with our complex mixuture of cultures, the
growth limiting step was the primary attack on p_-cresol. However, we were
working with very low levels of compound and frequently close to the limits
of detection.
Undoubtedly, if single isolates had been obtained, if the fermentations
had been appropriately regulated, and if higher levels of p_-cresol had been
used, metabolites would have been observed with characteristics of those
reported by Dagley andPatel (1957), and Bayly et al. (1966). The mixed cul-
ture fermentations are, however, more representative of conditions in nature.
38
-------
5. LABORATORY INVESTIGATION OF BENZ[a]ANTHRACENE
5.1 SYNOPSIS
The results of the laboratory investigations suggest that benz[a]anthracene
(BA) will accumulate in the sediment and biota portions of an aquatic environ-
ment. A small amount of BA will be dissolved and will be transformed by photol-
ysis and, to a lesser extent, by oxidation. One primary photolysis product is
the 7,12-quinone, which is photostable at 366 nm. Other products are also
formed as a result of photolysis in the presence of both water and oxygen.
Biotransformation of BA;was not observed in our enrichment cultures; however,
reports in the literature suggest that biotransformation of BA might occur
under environmental conditions.
The nine-compartment environmental exposure assessment model predicted the
following steady-state concentrations of BA in solution, suspended solids,
and sediments near point sources in the presence of a continuous discharge of
1 ng ml-1 (1 ppb) BA.
^ Suspended
Half-life Solution solids Sediments
(hr) (ng ml *) (ng g"1) (ng g"1)
River 0.55 2.6 x 10"1 6.6 x 103 6.1 x 103
Pond 22 3.2 x 10"3 7.2 x 101 7.2 x 101
Eutrophic
2
lake 22 3.2 x 10~2 8.0 x 102 2.7 x 10
Oligotrophic „
lake 1.8 x 10~2 4.4 x 102 1.2 x 102
5.2 BACKGROUND
Benz[a]anthracene (BA) is a polynuclear aromatic hydrocarbon (PAH) that is
known to be carcinogenic to animals and probably also to humans. Excellent
reviews on the occurrence of BA and other PAH in various environments are
available (Radding et al., 1976; Harrison et al., 1975; National Academy of
Sciences, 1972; Andelman and Suess, 1970).
Predicted by one-compartment model.
'Average concentration in sediments.
39
-------
Benz[a]anthracene (BA) is ubiquitous in the environment. Typical sources
are engine exhaust, cigarette smoke, coal-tar pitch, and the soot and smoke
of industrial and domestic origin. High concentrations of BA have been found
in air and in soil near industrial centers and transportation routes. BA
concentrations of 1 to 23 pg ml"1 have been found in drinking waters, with
4.3 to 185 pg ml"1 found in surface waters. Concentrations as high as
31.4 ng ml"1 have been found in sewage water, and concentrations ranging from
25 to 10,360 pg ml"1 have been found in industrial effluents and waters contam-
inated by bituminous substances (IARC, 1973).
The general physical properties of BA that were found in the literature
are given in Table 5.1.
TABLE 5.1. PHYSICAL PROPERTIES OF BENZ[a]ANTHRACENE
11 12
ooo
876
Molecular weight
Melting point (°C)
Boiling point at 760 torr (°C)
Vapor pressure at 20°C (torr)
(Pupp et al., 1974)
Solubility in water at 27°C (ng ml"1)
(Davis and Parke, 1942)
1.00 ng ml"1 (1 ppb) BA in water = 4.38 x 10
,-9
228.28
155-7 (160)(162)(167)
435 (sublimes)
5 x 10~9
11
M
The oxidation and photochemistry of PAH were reviewed by SRI as part of a
task report for EPA-OTS (Radding et al., 1976). McGinnes and Snoeyink (1974)
have reported the direct photolysis of BA; their data indicate that the half-
life for photolysis in sunlight is less than a day. They tentatively identified
one BA photoproduct as benz[a]anthracene-7,12-quinone and a second major prod-
uct only as a complex organic acid.
Microbial oxidations of benzo[a]pyrene, benz[a]anthracene, and dibenz[a]-
anthracene have been reported by Sisler and Zobell (1947), Fedoseeva et al.
(1968), and Poglazova et al. (1967a,b; 1968). However, little was known about
the metabolic products of these compounds until recently.
Gibson et al. (1975), using enrichment techniques with biphenyl as a sole
carbon source, isolated a Beijerinckia from a polluted stream. This organism
40
-------
completely degraded biphenyl, but a mutant obtained from this parent could
only oxidize biphenyl to cis-2,3-dihydroxy-l-phenylcyclohexa-4.6-diene when the
mutant was grown in an inorganic-salt medium containing succinate (0.2%) and
biphenyl (0.1%) and the washed cells were incubated with biphenyl.
Under these growth conditions, succinate provided all the carbon for growth.
These two strains have been very useful in metabolic studies with such chemi-
cally related compounds as BA, benzo[a]pyrene, and dibenzothiophene (Gibson, 1975;
Gibson et al., 1975). The mutant converted BA to four dihydrodiols (Gibson et
al., 1975; Gibson, 1975), and the parent strain could oxidize this compound to
cause ring fission and form organic acids that were not identified (Gibson, 1975).
This excellent work from Gibson's laboratories indicates the value of using
chemically related products for development of enrichment cultures and the
utility of this procedure,in transformation or degradation studies with more
complex products.
In totally different ^tudies with sediments from an oil-polluted creek,
Walker and Colwell (1975) concluded that the susceptibility of aromatic components
to degradation decreases with increase in the number of rings, i.e., in order
from monoaromatics through diaromatics, triaromatics, tetraaromatics, and penta-
aromatics.
Wodzinski and Johnson (1968) reported that the generation times of the
bacteria isolated on aromatic hydrocarbon were related to the solubilities of
the aromatic hydrocarbons on which they were grown. The more insoluble aromatic
hydrocarbons resulted in slower growth.
Wodzinski and Bertolini (1972) and Wodzinski and Coyle (1974) showed that
bacteria utilize naphthalene and phenanthrene in the dissolved state and that the
growth rates were independent of the amount of solid, undissolved hydrocarbon
presented, which suggests that bacteria do not utilize hydrocarbons directly at
the surface of a solid particle suspended in the medium. They predicted that
very insoluble aromatic hydrocarbons would not be utilized readily.
Wodzinski and Larocca (1977) reported that bacteria can also utilize liquid
state aromatic hydrocarbons and solid aromatic hydrocarbons dissolved in liquid
hydrocarbons at the water-hydrocarbon interfaces, but not at the solid particle
surface.
A personal communication from J. J. Perry (1977) indicated that a number of
fungi could transform benzo[a]pyrene to complex mixtures of organic acids. A
comparable situation could exist in the biodegradation of BA.
These results suggest that biodegradation of BA might occur under certain
conditions. However, there have been no reports on the development of BA-biode-
grading enrichment systems or on the rate of metabolism of BA in aquatic environ-
ments.
We anticipated that photolysis and sorption onto sediments and biota would
probably be important pathways for BA. The literature data on these processes
41
-------
were, however, insufficient to allow us to quantitatively estimate the importance
of these processes in relation to biodegradation, volatilization, and oxidation.
Screening studies were therefore conducted for all of these processes.
5.3 ENVIRONMENTAL ASSESSMENT
5.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigations of the
transformation and transport processes for BA are summarized in Table 5.2.
TABLE 5.2. SUMMARY OF BENZ[a]ANTHRACENE LABORATORY DATA
Process Sorption equilibrium Partition coefficient
Sorption S = K S K = 26,200 4- 1,700*
s p w p — '
Rate expression Rate constant at 25°C
b 0
Volatilization k [BA] k = (0.0016 + 0.0006)k
Photolysis0 k [BA] = $ (EZ.. e.) [BA] k = (5.91 + 0.07) x 10~5 sec'1
p A A p
Oxidation k [ROV] [BA] k = 5.0 x 103 M"1 sec"1
ox 2J ox
Hydrolysis NA k, = 0
Biodegradation k^_ = y k = 0
o
On Coyote Creek sediment.
See discussion in Part I, Section 5.3 of the final report.
c
Midday on a cloudy day in early March.
Biodegradation was not observed during our enrichment procedures.
5.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved BA calculated for individual transformation
or removal processes following a spill are listed in Table 5.3. Although these
half-lives vary among water bodies, photolysis and sorption consistently domi-
nate the transformation and removal processes. Loss by oxidation is about 4
times slower than loss by photolysis, and loss by volatilization is 20 to 100
times slower than photolysis.
42
-------
As discussed in detail in Section 5.5, the half-life of photolysis as
measured at noon was approximately 3 hours. Allowing for variations in light
intensity during a 24-hour cycle, the average photolysis rate constant is
estimated to be 6.93 x 10 l per hour in oligotrophic waters, giving an esti-
mated half-life of 40 hours.
TABLE 5.3. TRANSFORMATION AND TRANSPORT OF BENZ[a]ANTHRACENE
PREDICTED BY THE ONE-COMPARTMENT MODEL
Process
Photolysis, half-life
}
(hr)
Oxidation, half-life3 (hr)
Stream '
20
38
Eutrophic
pond
50
38
Eutrophic
lake
50
38
Oligotrophic
lake
10
33
Volatilization,
half-life (hr)
Hydrolysis,
half-life (hr)
Biodegradation,
half-life (hr)
Half-life for all
processes, except
dilution (hr)
Half-life for all
processes, including
>1000
NA
Slow
13
>1000
NA
Slow
22
>1000
NA
Slow
22
>1000
NA
Slow
dilution (hr)
Amount BA sorbed0 (mg m~3)
Percentage BA sorbed
0.55
2.5
71%
22
7.5
88%
22
1.25
55%
8
1.25
55%
-9
Based on peroxy radical concentration of 10~y M.
31 ng ml"1 BA is assumed in the solution phase.
-------
The oxidation rate of BA may also vary during a 24-hour period, since
the formation of the peroxy radical, which is at least partially responsible for
the oxidation of BA, may be governed by light intensity. Hence the oxidation
half-life of 38 hours estimated from the laboratory data, when adjusted to
account for diurnal variations, gives an estimated half-life of about 160 hours
in an aquatic environment.
Sorption half-lives have not been measured, but the times needed to reach
equilibrium are generally less than half an hour for those compounds for which
the rates have been measured.
The partition coefficient of BA was measured as 2.5 x 1014. As shown
in Table 5.3, ranges of 55% to 88% BA are sorbed on the solid phase. This
means that less than half of the total BA in the water body is available for
photolytic and oxidative reactions.
Half-lives are estimated to range from 10 to 50 hours in rivers, ponds,
and lakes as wholes, but the half-life of BA in specific river segments should
be an order of magnitude less, the precise value depending on the flow rates
and geometry of that river segment.
5.3.3 Persistence
The persistence of concentrations that approximate the initial concentra-
tions following an acute discharge should be largely a function of the dilution '
rate in fast flowing streams. Persistence of relatively high concentrations in
lakes and ponds is expected to be determined by rates of sorption and photolysis
within the aqueous phase. In both cases, "high" concentrations should be tran-
sient, lasting only a few hours. However, low concentrations of BA caused by
desorption from contaminated sediments can be expected indefinitely.
BA may be more persistent in eutrophic water than in oligotrophic waters,
since the major transformation pathway, photolysis, would be less effective in
such waters, as indicated in Table 5.3. Moreover, since the sediment loading
is typically higher in eutrophic waters, the amount of BA sorbed to solids
should be greater than in oligotrophic waters, creating a larger reservoir of
BA that is subject to later desorption. This pattern is borne out by the more
elaborate simulations reported below.
5.3.4 Mass and Concentration Distributions Calculated Using Computer Models
The pseudo-first-order rate constants used in these simulations are
presented in Appendix A. The distributions of mass and concentration of
BA expected at steady state during chronic discharge to each of four types of
water bodies (Table 5.4) are typical of most of the chemicals studied to date.
At most, 20%, and generally less than 0.5%, of the BA is in solution under
steady-state conditions. The steady-state concentrations of BA in solution or
sorbed to suspended solids differ about 100-fold between compartments in each
of the multicompartment simulations, but are generally at least 100 times less
than in the inflowing waters, with the highest concentrations occurring near
the source.
44
-------
TABI.K 5.4. DISTRIBUTION OK BENZ[a]ANTHRACENE IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(Input concentrations of 1 ng ml"' henzlajanthraeene)
—
Compartment
1
1'ond
Mass Cone.
(kg) _(ng g'1)
River
Mass
(kg)
Cone.
("8 f
>-')
(surface water)
Solution
Suspended
Compartment
sol ids
2
6.30 x IO-5 3.2 x IO-3
4.4 x 10-'' 72.0
7
1
.93
.98
x 10-"
x 10-'
2.64 x
6.60 x
10-'
IO3
(surface water)
Solution
Suspended
solids
—
7
1
.35
.83
x 10
x 10-'
2.45 x
6.10 x
io-'
IO3
Eutrophic Lake
Mass
(kg)
7.98 x Ur3
9.98 x 10~3
2.74 x 10~?
3.42 x IQ~2
Cone.
(ng g-1)
3. 19 x 10~2
7.98 x IO2
1.09 x 10"2
2.73 x IO2
Oligotrophlc Lake
Mass
(kg)
4.40 x IO-3
5.50 x 10~3
5.18 x 10~3
6.48 x 10"3
Cone.
_ (ng g~')
1.76 x 10"2
4.4 x IO2
2.07 x 10~3
51.8
Compartment 3
(surface water)
Solution
Suspended sol ids
6.81 x ID"2 2.27 x 10"' 1.79 x 10"3 7.16 x 10~3 2.47 x 10-" 9.83 x 10''1
1.70 x 10-' 5.66 x IO3 2.24 x 10~3 1.79 x IO2 3.09 x 10-" 24.7
Compartment 5
(bottom water)
Solution — — — — 4.93 x IO-3 1.97 x lO'1* 3.47 x !Q-b 1.38 x 10~J
Suspended solids — — — — 6.17 x UT3 49.3 4.33 x 10'5 3.46
Compartments 7-9
(sediment)3
SolutIon
Sol ids
8.0 x 10-' 3.2 x 10~3 1.83 x IQ-2 2.65 x 10"' 4.99 x 10~b 5.71 x 10"3 2.33 x 10"b 2.66 x 10~3
4.9 x~10-'' 72.0 1.24 x 10Z 6.12 x IO3 3.63 2.66 x IO2 1.59 1.16 x IO2
'4Tlie iimouiits given for solid and solution phases In the sediment compartments are estimated from the sorption partition coefficient
for suspended solids and may he overestimated because it was assumed that biodegradation of sorbed material does not occur.
-------
The most interesting patterns occur during the approaches to steady
states before and just after the simulated chronic discharges are stopped
in the oligotrophic and eutrophic lake simulations (Figures 5.1 and 5.2).
After discharge stops in the oligotrophic lake simulation, concentrations of
dissolved BA decline rapidly by one to two orders of magnitude, reaching new
steady states in each aqueous compartment in a week to ten days. New steady
states are approached more slowly in the eutrophic lake simulation, apparently
requiring at least two weeks, and possibly one or more months. In addition,
the range of concentrations between compartments is an order of magnitude smaller
in the eutrophic than in the oligotrophic lake, and the concentrations of dis-
solved BA are two to four times higher. The net effects are higher concentra-
tions and much slower loss of BA from solution in eutrophic waters and a
wider distribution of relatively high concentrations.
Changes in the concentrations of suspended sediments are not shown in
Figures 5.1 and 5.2 but can be inferred from the changes in the concentrations
of dissolved BA, which they parallel, as seen in the pond simulation shown in
Figure 5.3, which closely resembles compartment 1 of the eutrophic lake simu-
lation.
Changes in concentrations of dissolved BA are most dramatic in the river
simulation (Figure 5.4) where the high rates of dilution result in very low
concentrations of dissolved BA in individual compartments and minimize the
difference in concentration between compartments. As observed in other river
simulations, "the zone of peak concentration of dissolved BA moves downstream
as increasing concentrations reach the successive downstream compartments.
While this effect is accentuated by the absence of tributaries in the model
stream, this pattern of downstream shift in peak concentrations might occur in
actual streams and could obscure the location of sources of contaminants.
5.3.5 Discussion
The significance of the concentrations simulated in this study cannot
be readily appraised with the available toxicological data, which are based on
experiments with somewhat higher concentrations than we used. Although acute
exposures to concentrations of 1 ng ml"1 are apparently not toxic to fish
(McKee and Wolf, 1963) and are within the range reported for some drinking
waters, BA is a carcinogen (IARC, 1973), and as such, must be assumed to be
hazardous at these low concentrations until demonstrated to be otherwise.
5.4 PHYSICAL PROPERTIES
5.4.1 Solubility in Water
Two reports of the solubility of BA in water were found in the literature.
Davis and Parke (1942) give the solubility as 11 ng ml""1 at 27°C and Klevens
(1950) gives a value of 10 ng ml"1 at 25°C. We also measured the solubility
of BA in water using the method described in Section 5 of Part I of this re-
port and obtained a value of 5.7 + 0.5 ng ml-1 [2.5 (+ 0.2) x 10~8 M] at 20.0
+ 0.1°C based on replicate analyses. Some of the solutions were centrifuged
after equilibration to make sure they were free of suspended solid BA.
46
-------
SEDIMENTS
SOLUTION
100
200
300 400
TIME - hours
500
600
700 720
FIGURE 5.1 PERSISTENCE OF BENZ [a]ANTHRACENE IN A PARTIALLY MIXED
OLIGOTROPHIC LAKE
47
-------
SEDIMENTS
SOLUTION
100
200
300 400
TIME - hours
500
600
700 720
FIGURE 5.2 PERSISTENCE OF BENZ [a]ANTHRACENE IN A PARTIALLY MIXED EUTROPHIC LAKE
48
-------
100
10
c
I
LU
H
LU
O
<
cc
I
N
2
H
O
<
0-
Z
LU
O
O
O
10'1
1-2
10'
-L-'-J-I— '-_
SUSPENDED SOLIDS
I I I I
IIIIIE
SEDIMENTS
=~ I
~ I
-I
I
1
DISCHARGE STOPPED
SOLUTION
50
100 150
TIME - hours
200
250
FIGURE 5.3 PERSISTENCE OF BENZ [a] ANTHRACENE IN A TWO-COMPARTMENT POND SYSTEM
49
-------
Ul
o
10 —
E
o>
c
I
LU
z
LU
O
tr
i
z
N 10'
O _
o
cc
z
111
u
§
o
10"'
10'=
10"
COMPARTMENT
I I
SEDIMENTS
DISCHARGE STOPPED
X
SEDIMENTS
SOLUTION
SOLUTION
10
TIME - hours
FIGURE 5.4 PERSISTENCE OF BENZ [a] ANTHRACENE IN A PARTIALLY MIXED RIVER SYSTEM
-------
5.4.2 Absorption Spectrum
BA absorbs light strongly in the solar region to 390 nm, then weakly
to 490 nm. The absorption spectrum was measured in 50% acetonitrile/50% water.
The high percentage of acetonitrile was necessary to give a solution concen-
tration that was adequate for measurement on the Gary 15 spectrophotometer.
The solutions were tested for cloudiness with collimated beam of light to make
sure that they did not contain undissolved BA. Since BA contains no functional
groups, the effect of changes in pH on the absorption spectrum was not investi-
gated. All measurements were made at pH 7. The absorption spectra for BA at
wavelength intervals from 297.5 to 500 nm and at wavelengths of 313 and 366 nm
are reported in Table 5.5.
5.4.3 Volatilization Rate
The volatilization rate for BA was measured by the method of Hill et al.
(1976), which is described in detail in Part I, Section 5.3.3. The volatili-
zation half-life of BA In an aqueous solution was about 89 hours under the
experimental conditions used for this measurement. The rate constants for BA
volatilization and oxygen reaeration were measured to be:
* 0 i
Oxygen reaeration rate: k =4.92+1.09 hr"1
BA volatilization rate: kM = 0.00789 + 0.00093 hr"1
v —
Ratio: kBA/k° = 0.00160 + 0.00057
v v —
5.4.4 Sorption on Clay and Sediments
The sorption partition coefficients of BA were measured on three natural
sediments: Coyote Creek, Des Moines, and Searsville Pond. In addition, a
desorption partition coefficient was measured for the Searsville Pond sediment.
Care was taken to assure that the samples were protected from light at all
times to minimize photooxidation of the BA. The initial concentration of BA
in,each sample was always less than the saturation concentration of 5.7 yg ml"1
(2.5 x 10~5 M). Each experiment consisted of triplicate analyses of replicate
flasks at two BA concentrations and two sediment loadings plus suitable blanks,
as described in Appendix B. Periodic checks were made for sorption of BA
onto the glassware; since none of the glassware extracts contained BA, we
presumed that no sorption onto the glassware took place.
The data, summarized in Table 5.6 and Figures 5.5 and 5.6, were fitted
to a Freundlich isotherm, where n ~ 1, using several statistical procedures.
This value represents the average and standard deviation of six reaeration
measurements taken over the period of the experiment.
51
-------
TABLE 5.5. ABSORPTION SPECTRUM OF BENZ[a]ANTHRACENE
IN 50% ACETONITRILE/50% WATER AT pH 7a
Center of Average absorption
wavelength interval (nm) coefficient (M"1 cm )
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330.0
340.0
350.0
360.0
370.0
380.0
390.0
400.0
410.0
420.0
430.0
440.0
450.0
460.0
470.0
480.0
490.0
500.0
7930
7070
5880
3790
3200
3480
3900
4200
4170
4120
4800
5450
5390
4850
3350
1560
662
417
17.2
18.1
18.1
13.6
33.6
15.4
11.8
36.3
8.2
1.8
0
•q 1
Benz[a]anthracene concentration was 25.16 pg ml"1
(1.102 x 10""4 M).
The wavelength intervals are given in Appendix B, Table B.I.
°The absorption coefficients at 313.0 and 366.0 are 4010
and 2410 M"1 cm"1, respectively.
The column labeled LLS, a = 0 gives the results of using the linear least
squares (LLS) equations* for the equation
S = K S (5.1)
s p w
where S and S are the substrate concentrations on sediments and supernatant,
respectively, and K is the sorption partition coefficient.
* The regression equations are given in Appendix B, Section B.I.
The Hewlett-Packard Model 65 calculator was programmed to carry out these
calculations.
52
-------
TABLE 5.6. BENZ[a]ANTHRACENE SORPTION ON SEDIMENTS
- - - - - - - — - -
Total
organic
carbon
Sediment (percent)
Coyote 1.4
Creek
sorption
Des Moines 0.6
sorption
Searsville 3.8
Pond
sorption
Searsville 3.8
Pond
desorption
Sediment
concentration
(Mg ml'1)
19
35
19
50
20
60
20
60
16
32
16
32
31
63
31
63%
BA
concentration
in supernatant3
(ng ml-1)
2.7 + 0.4
2.6+0.2
2.0 + 0.3
1.4 + 0.2
4.4 + 0.2
3.3 + 0.2
3.0 + 0.3
2.1 + 0.3
3.5 + 0.3
2.8 + 0.2
2.0 +0.5
1.6 + 0.2
1.4 + 0.1
1.3 + 0.1
1.0 + 0.2
0.8 + 0.3
BA
concentration
on sediment"
(ng g-1 x 10~3)
82 + 5
~
63 + 6
39 + 15
33 + 2
51 + 10
23 + 3
30 + 4
16 + 2
75 + 25
55 + 11
50 + 6
33 + 4
29 + 10
24 + 6
13 + 5
22 + 3
Partition coefficient, K (x 10~3)d
P
n r LLS NLLS
Recovery & = Q & ^ Q
(percent) o o w s w
91 29.3 + 4.7 27.3 + 5.3 - 26.2 + 1.7
—
103 (a0 - 4.3 + 27)
87
95
101 9.8 + 2.0 15.0 + 6.6 8.75 + 0.55 8.30 + 0.70
87 (a0 = -18 + 43)
103
88
97 20.0 +5.1 15.0 + 20.0 20.6 + 0.9 20.7 + 1.0
92 (a0 = 14 + 52)
92
89
97 19.0 + 5.3 13.0 + 23.0 - 20.6 + 1.3
81 (a0 = 7.7 + 26)
89
103
Concentration measured in supernatant with population standard deviation.
Concentration measured on sediment with population standard deviation.
Based on using blank flasks to determine total material added to each flask.
LLS » linear least squares; NLLS « nonlinear least squares. See text for description of regressions. Limits are 95% confidence limit.
-------
tn
120
100
Z x
LU
~
01
a
vs
ulo
CO j
u. 5
O O
I- ai
O Q
Z LU
O <"
0 Z
o
80
60
40
20
AVERAGE OF TWO FLASKS WITH
POPULATION STANDARD DEVIATIONS
DES MOINES RIVER: AVERAGE OF TWO FLASKS WITH
POPULATION STANDARD DEVIATIONS
26.000
Kp = a300
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
CONCENTRATION OF BENZ[a]ANTHRACENE IN SUPERNATANT AT EQUILIBRIUM - agent'1 (ppbl
4.5
FIGURE 5.5 SORPTION ISOTHERMS OF BENZ[a]ANTHRACENE ON COYOTE CREEK AND
DES MOINES RIVER SEDIMENTS.
Kp was estimated by nonlinear least squares using supernatant and sediment
concentrations from sorption experiment
-------
U1
LTI
UJ X
^
C »
I en
I- C
I '
2 S
UJ CD
CO ~J
< t-
cc z
O Q
Z liJ
O «"
0 Z
o
120 —
8
80
60
20
1
1
J_
SORPTION EXPERIMENT: AVERAGE OF TWO FLASKS WITH
POPULATION STANDARD DEVIATIONS
DESORPTION EXPERIMENT: AVERAGE OF TWO FLASKS WITH
POPULATION STANDARD DEVIATIONS
0.5 1.0 1.5 2.0 2.5 3.0 3.5
CONCENTRATION OF BENZ[a]ANTHRACENE IN SUPERNATANT AT EQUILIBRIUM - ng ml ° (ppb)
4.0
FIGURE 5.6 SORPTION AND DESORPTION ISOTHERMS OF BENZiaJANTHRACENE
(SEARSVILLE SEDIMENT).
Kp was estimated by nonlinear least squares using supernatant and
sediment concentrations from sorption experiment.
-------
The column labeled LLS, a £ 0 gives the results of using linear least squares
equations for the equation
S = K S + a (5.2)
s p w o
where a is the intercept. In both cases, the average S and S for each flask
were used as input data to the regression equations. These calculations show
that the intercept of the regression equation is the origin within experimental
error. The K values in the columns labeled NLLS were calculated using the
nonlinear least squares (NLLS) techniques described in Appendix B, Section
B.I.4. The Kp values in the column labeled Sw only were calculated from
supernatant concentrations only. The "best" estimates of Kp are the values
reported in the last column, labeled Sg and S . These estimates were ob-
tained from the measurements of Ss and Sw.
Several conclusions can be made from these results. First, the nonlinear
least squares technique gives the best estimate of K , since the 95% confidence
limits are less than the LLS method. Second, the estimates of K are virtually
identical when calculated from S data only or when calculated from values of
S and S . Third, the LLS method, where a =0, also gives a good estimate,
since the range of K overlaps the NLLS estimates. However, the 95% confidence
limits are much wide? for the LLS, a0 = 0, method than for the NLLS methods
because the error due to flask effects is included in the LLS methods.
The results from the desorption isotherm measurement on Searsville Pond
sediment are summarized in Table 5.6 and Figure 5.6. These data show that the
value of K is, within experimental error, the same for sorption and desorption.
Therefore, for a sorption time of 12 hr sorption of BA is reversible on this
sediment.
The correlation of K with sediment organic content is poor (r2 = 0.45).
However, only three points are available and the correlation might improve if
K were measured on more sediments.
P
5.4.5 Biosorption
Because of the low solubility of BA, biosorption studies were conducted
at BA concentrations of 12.8 and 7.5 ng ml"1. The density of the mixture of
the four bacterial cultures used in sorption studies was reduced to an optical
density of 0.2. The results from the one-hour sorption and three-hour desorp-
tions are presented in Table 5.7. The cell dry weights were equivalent to
96 yg ml"1. The adjustments for sorption of BA on glassware in Table 5.7 were
3.4% and 1.6% for sorption studies with 12.8 and 7.5 mg ml"1 BA, respectively.
The corresponding adjustments for glassware sorption in the desorption were
2.6% and 1.7%, respectively.
*
Calculated using the Hewlett-Packard Model 65 calculator and the Stat-Pac I
routines.
56
-------
TABLE 5.7. BENZ[a]ANTHRACENE SORPTION AND DESORPTION
ON A MIXED BACTERIAL POPULATION
Initial BA Sorption (S) Amount in
concentration or supernatant
(ng ml"1) desorption (D) (%)
12.8 S
D
7.5 S
D
21 + 1
22 + 2
23 + 3
21 + 3
Amount in
cells (%)a
78 + 2
59 + 1
72 + 3
50 + 1
Sorption
coefficient
(x lO"1*)
3.9 + 0.2
2.9 + 0.3
3.3 + 0.4
2.5 + 0.2
"The dry weight equivalent to 96 yg ml 1.
Sorption coefficients of viable cells and heat-killed cells were also
compared. The sorption coefficients with initial BA concentration of
11.4 ng ml"1 are shown in Table 5.8. Heat-killed cells had a higher sorption
coefficient.
TABLE 5.8. BENZ[a]ANTHRACENE SORPTION BY VIABLE
AND HEAT-KILLED BACTERIA3
Condition
of cells
Viable cells
Heat-killed cells
Cells dry
weight
(mg-liter"1)
91
76
Amount in
supernatant
(%)
22 + 1
11 + 1
Amount in
cells
(%)
71 + 5
86 + 1
Sorption
coefficient
(x 10-1*)
3.6 + 0.3
9.9 + 0.7
f\ 1
The initial concentration of BA was 11.4 ng ml"1.
As expected, BA sorption coefficient data were lower than those for
BaP, but they are still very high. These high sorption coefficients with
bacterial cells would reduce the BA in the aqueous phase "available" for
development of biodegradable organisms. These data also suggest that sorption
could be important in biomagnification up the food chain ladder.
5.5 CHEMICAL TRANSFORMATION
5.5.1 Photolysis Rate
Photolysis of BA is rapid in the solar region with half-lives of several
hours. Data for the photolysis of BA with air in sunlight and at 313 and 366 nm
are given in Table 5.9. Acetonitrile was used as the cosolvent to enhance
57
-------
TABLE 5.9. RATE CONSTANTS FOR PHOTOLYSIS OF BENZ[a]ANTHRACENE
00
BA
Light source
313 nm
313 nm
366 nm
366 nm
366 nm
366 nm
366 nm
366 nm
366 nm
366 nm
366 nm
Midday sunlight,
early March
al.OO ng ml"1 BA in
AN = acetonitrile.
concentration3
(ng ml-1)
19.6
18.0
17.4
9.8
19.6
9.8
22.4
9.3
19.6
222
17.3
19.6
water = 4.38 x
Solutionb
0.1% AN in pure water
1.0% AN in pure water
0.1% AN in pure water
1.0% AN in pure water
0.1% AN in Lake Tahoe water
1.0% AN in Lake Tahoe water
0.1% AN in Coyote Creek water
1.0% AN in Coyote Creek water
0.1% AN in Searsville Pond water
1.0% AN in Searsville Pond water
Humic acid (8 ug ml"1) in 1% AN
in pure water
0.1% AN in pure water
10~9 M.
%
Reaction
63
38
83
67
88
62
84
58
85
56
57
66
Rate constant
(K x 105 sec'1)
P
2.59 + 0.09c,d
1.98 + 0.08e
14.5 + 0.4^
12.3 + 0.3e
8.10 + 0.65
10.8 + 0.2
8.80 + 0.48
9.55 + 0.15
7.47 + 0.30
6.55 + 0.09
3.9f
5.91 + 0.078
Standard deviation.
d o
Quantum yield of 3.2 x 10"3.
eQuantum yield of 3.4 x 10~3.
Rate constant for initial photolysis; data do not follow first-order kinetic law.
8Half-life of 3.3 hours.
-------
the solubility of BA in water from the sources indicated because of the low
solubility of BA in pure water. (See Section 5.4.1.) Except for the photoly-
sis reaction in the solution containing 8 Mg ml"1 humic acid, good first-order
kinetic behavior was found for photolyses carried out to beyond two half-lives
of BA under laboratory conditions.
In experiments using 9.3 to 222 ng ml"1 BA, the photolysis rate
constant was independent of BA concentration, again confirming first-order
kinetic behavior. The quantum yields for direct photolysis at 313 and 366
nm were (3.3 ± 0.1) x 10~3. The data in Table 5.9 also indicate that
the photolysis experiments in the same water are not reproducible to within
two standard deviations (^ 95% confidence limits) of each experiment;
the average deviation between the two experiments in the same water ranged
from 4% to 14%. Beyond these limits, however, the photolysis rates in Lake
Tahoe or Coyote Creek water is about 70% of that in pure water, and in the
water from the pond near ISearsville Lake, the rate is half as fast as in pure
water. Since the absorb^nce of the natural waters at 366 nm is less than 0.02
(in a 1-cm cell), the lower photolysis rates observed in natural waters are not
due to light screening effects.
It was also found that photolysis at 366 nm of a 9.3 ng ml"1 BA solution
(1% acetonitrile in water) containing 8 ug ml"1 humic acid did not follow a
first-order kinetic law, since the rate decreased as the photolysis proceeded;
we have no explanation for this behavior. While not a first-order reaction,
a first-order rate constant calculated from the initial two data points is
4 x 10~5 sec"1, which is at least three times slower than the reaction in pure
water. Since the absorbance of the humic acid solution is 0.43 (1-cm cell),
the light screening effect of the humic acid should only reduce the reaction
rate by 37% (e.g., 0.63 times the rate in pure water). Some process other
than screening appears to be retarding the photolysis rates in the three natural
waters. Discussion of some possible explanations is given in the section on
BaP (Section 6.5.1).
The half-life for direct photolysis of BA in sunlight as a function of
the time of day was calculated by the procedure of Zepp and Cline (1977) using
a quantum yield of 3.3 x 10~3 and the measured UV spectrum of BA (Section 5.4.2).
The data for summer and winter seasons are plotted in Figure 5.7. A calculated
half-life of 1 to 2 hours for midday photolysis is in good agreement with the
3-hour half-life measured for midday on a cloudy day in early tlarch.
S 5.2 Oxidation Rate
The susceptibility of 19.6 ng ml"1 (8.59 x 10~8 M) BA to free radical
oxidation was examined using an AA-initiated oxidation reaction, as described
in Part I and Appendix B. The experiment was performed at 50.0°C in 1% aceto-
nitrile in water containing 9.6 x 10~5 M AA.
Analysis of controls (BA solutions without AA maintained under identical
reaction conditions) at reaction times of 3.0 and 5.4 hours showed a 36% loss
of BA compared with the original solutions; this was apparently due to sorption
59
-------
Q _
3
o
•= 5
AM
PM
6
6
7
5
9
3
10
2
11
1
12 NOON
TIME OF DAY
FIGURE 5,7 SEASONAL AND DAILY VARIATION OF PHOTOLYSIS HALF-LIFE
OF BENZ [a] ANTHRACENE
-------
of BA on glassware. Correcting for this sorption, we obtained a pseudo-first-
order rate constant (kox[RO^D of (1.53 + 0.18) x ID"4 sec"1 for oxidation of
BA under the reaction conditions. This rate constant corresponds to a half-life
of 1.3 hours. Under these conditions (Part I, Section 6.3), this corresponds
to a second-order rate constant kox of 4.5 x 101* M"1 sec-1 at 50.0°C. At 25°C,
k0x would then be 5.0 x 103 M"1 sec"1.
Assuming that [RC^*] = 10~9 M in the aquatic environment, the half-life
of BA toward oxidation is estimated to be about 38 hours. The free-radical
oxidation of BA in the environment is rapid and could be competitive with photol-
ysis or sorption processes.
5.5.3 Hydrolysis Rate
BA contains no groups that are hydrolyzable; therefore, no hydrolysis
studies were conducted.
5.5.4 Products from Chemical Transformation
One product from the photolysis of BA at 366 nm in air-saturated,
acetonitrile-water solvent was identified as benz[a]anthracene-7,12-quinone
(7,12-BAQ) by comparing the HPLC retention time and UV spectrum with that of
an authentic quinone sample.
At times when BA was 15%, 26%, and 46% reacted, 7,12-BAQ was present in about
26% yield based on the amount of BA reacted. The quinone was stable to photol-
ysis at 366 nm at these reaction times.
At low conversions of BA, we found that two other products peaks were also
present in photolysis reaction mixtures from experiments under a variety of
conditions. At higher conversions of BA, at least four additional, but smaller,
peaks were found in the HPLC traces of the reaction mixtures. These products
eluted earlier than BA or 7,12-BAQ on reverse-phase HPLC. Since these products
could not be separated and identified, their yields could not be determined.
These more polar, unknown products could account for the balance (74%) of the
reacted BA.
In an attempt to identify the photochemical processes responsible for
the complex product mixtures, we performed several additional experiments.
The photolysis of BA at 366 nm in 1:1 acetonitrileiwater that had been purged
of oxygen with nitrogen gave a photolysis rate that was about one-third of
61
-------
that found when oxygen was present. The same three major product peaks were
present in this reaction mixture.
It was also found that the photolysis of BA in oxygen-saturated, pure
acetonitrile solvent was slower than in tiie aqueous solutions. In this ex-
periment, less than 2% 7,12-BAQ was formed and one of the other two major
products significantly increased in yield. In nitrogen-purged, pure aceto-
nitrile solvent the photolysis of BA did not occur.
In another experiment it was also found that photooxygenation in methanol,
using rose bengal as the singlet oxygen sensitizer, produced the same mixture
of products that was found in the air-saturated aqueous solutions.
The effects of water and oxygen on the BA photolysis rates and the
finding that 7,12-BAQ is formed in water in the absence of oxygen is surprising
and indicates that the photolyses are more complicated than the simple photo-
oxygenation process expected on the basis of literature information. In photol-
ysis studies on BA, McGinnes and Snoeyink (1974) found a product that they
tentatively identified as 7,12-BAQ, in agreement with our work. They also
found a major product that they described as a complex organic acid with the
empirical formula C18H1303, which was not further characterized.
This polar, acidic compound could correspond to any of the products
that eluted early in our reverse-phase HPLC analysis. However, it is difficult
to rationalize how any product containing three oxygens could be a primary BA
transformation product. Since 7,12-BAQ is stable to photolysis in water, we
expect that such acid product(s) would be derived from other primary product(s),
which are rapidly photolyzed (or possibly hydrolyzed or oxidized). Evidence
for such unstable primary products, which could be other quinones or endoper-
oxides, could not be found in any of the HPLC traces of the reaction mixtures.
These findings indicate that the photoreactions of BA and possibly other
polynuclear aromatics are complex and require more detailed studies, especially
in aqueous sytems. Such investigations will be necessary in order to understand
the chemistry of these processes and the identity of the primary products.
The kinetic data obtained for the air-saturated, aqueous solutions presented
in Table 5.9 are, however, still reliable for environmental assessment purposes.
5.6 BIODEGRADATION
5.6.1 Development of Enrichment Cultures
Screening for enrichment cultures for BA degrading systems was conducted
with water samples in 9-liter aerated fermentors with 10 ug ml"1 of BA. The
samples were initiated with and without 10 pg ml"1 of phenanthracene added as
an analog inducer in conjunction with the BA. The water samples tested were
from Searsville Pond in Woodside, CA; Coyote Creek water; aeration effluents
from sewage plants in Palo Alto and South San Francisco, and from the Shell
Oil Refinery in Martinez, CA; and a composite sample of waters from ponds that
are part of the storm and emergency drainage system of the Shell Oil Refinery.
62
-------
Transfers were made from these 9-liter fermentors into shaker flasks
containing BA and basal salts with and without phenanthracene. Samples from
South San Francisco and Shell ponds were also transferred into shaker flasks
containing basal salts medium, BA, and naphthalene. Neither the 9-liter
fermentors nor subsequent transfers into flasks indicated the development of
BA degrading cultures within the specified six-week period.
Additionally, five single bacterial isolates obtained in other studies
from biphenyl and phenanthracene degrading systems were tested for BA degrada-
tion with 10 yg ml-1 of BA in the presence of the respective 10 pg ml"1 of
biphenyl or phenanthracene substrate with and without 100 ppm of Tripticase
Soy broth. No BA degradation was observed.
To test the possibility that polyaromatic carcinogens can affect metab-
olism or viability of bacteria, we examined the inhibition activity of 10 pg
ml"1 of BA on growth of pj-cresol and on quinoline degrading systems. No signi-
ficant effects were observed.
5.6.2 Biodegradation Kinetics
Since enrichment cultures that degraded BA were not obtained, biodegrada-
tion kinetic studies were not performed.
5.6.3 Discussion
The discussion about the biodegradability of benzo[ajpyrene (Section 6)
can be applied to BA. Enrichment procedures tend to favor certain types of
organisms. With the high sorption coefficients and the low solubility of BA,
it is possible that conventional enrichment procedures do not provide enough
"available" BA for organisms to utilize this substrate for growth in competi-
tion with other organisms that have a more favorable predator-prey relationship
under the conditions of the primary enrichment process.
The bacteria that can utilize BA may be too slow to develop within the
six>-week limit because the growth rates of bacteria in aromatic hydrocarbon
media are regulated by amount of dissolved substrates (Wodizinski et al., 1968,
1972, 1974). The fact that enrichment systems did not develop BA biodegrading
culture, mixtures within the time limit does not unequivocally classify BA as
a recalcitrant compound. BA biodegradation may also take place slowly on
sediments.
63
-------
6. LABORATORY INVESTIGATION OF BENZO[a]PYRENE
6.1 SYNOPSIS
The results of the laboratory investigations suggest that benzo [a] pyrene
(BaP) will accumulate in the sediment and biota portions of an aquatic envi-
ronment. A small amount of BaP will be dissolved and will photolyse rapidly.
The estimated half-lives of BaP dissolved in different types of water bodies
and the percent sorbed on suspended sediments are:
A Suspended
Half-life" Solution solids Sediments
(hr) (ng ml"1) (ng g"1) (ng g"1)
River 0.48 1.5 x 10~3 7.4 9.9
Pond 7.3 1.5 x 1(T6 7.0 x 10~2 7.0 x 10
~2
Eutrophic _
lake 7.4 1.1 x 10~5 5.4 x ID"1 1.4 x 10"1
Oligotrophic
lake 1.5 1.9 x 10~6 9.3 x 10~2 2.3 x 10~2
Several problems were encountered in these studies. Experimental diffi-
culties were encountered because BaP sorbed on glassware, and its low solubil-
ity in water made it difficult to prepare and work with aqueous solutions. We
suspect that sorbed BaP may be transformed by photochemical and chemical oxi-
dation as well as by biodegradation processes, but the extent of these processes
could not be evaluated by our laboratory procedures. Finally, there is litera-
ture evidence that biodegradation of BaP may take place under different condi-
tions and with different organisms than those used in these studies. There-
fore, the general conclusion is that BaP will accumulate in sediments and
microorganisms and may degrade slowly in sediments with half-lives of a few
years .
6.2 BACKGROUND
Benzo [a] pyrene is a polynuclear aromatic hydrocarbon (PAH) that is known
to be carcinogenic to animals and probably also to man. The time from ex-
posure to BaP to observation of cancer (referred to as the latent period) is
on the order of years, so the persistence and pathways for BaP in the aquatic
*
Predicted by one-compartment model.
64
-------
environment are subtle but important issues. Excellent reviews on the occur-
rence of BaP and other PAH in various environments have recently appeared in
literature (Radding et al., 1976; Harrison et al., 1975; National Academy of
Science, 1972; Andelman and Snodgrass, 1974; Andelman and Suess, 1970). Most
studies of PAH in natural waters do not discriminate between material that is
sorbed on particulate matter and that in "true" solution. As several reviews
have pointed out, however, the low solubility of BaP (and all PAH) in water
does not necessarily mean that all BaP is sorbed, since many natural waters
and industrial effluents also contain other more soluble organic material that
may increase the solubility of BaP.
BaP is widespread in the environment, having been found in soils, airborne
articulates, and plant materials as well as in natural, waste, and drinking
waters. In a 1970 review, Andelman and Suess reported that BaP concentrations
of over 300 ng ml"1 are present in some effluents from the coke and shale-oil
industries. Concentrations to 0.15 ng ml'1 BaP have been found in surface
waters. A review prepared by the International Agency for Research on Cancer
(1973) reported that concentrations of BaP in drinking waters ranged from
0.0001 to 0.023 ng ml"1.' A concentration of 56 yg kg"1 total PAH has been
found in river sediments; values of several thousand yg kg"1 BaP have been
found in analyses of marine sediments (Andelman and Suess, 1970).
The general physical properties of BaP that were found in literature are
given in Table 6.1.
TABLE 6.1. PHYSICAL PROPERTIES OF BENZO[a]PYRENE
Structure
Molecular weight
Melting point (°C)
Boiling point at 10 torr (°C)
Vapor pressure at 25°C (torr)
Solubility in water (ng ml"1) 1.2 ± 0.1
1.00 ng ml"1 (1 ppb) BaP = 3.96 x 10~9 M
Pleasured in this study (Section 6.4.1).
65
-------
The reported concentrations of BaP in sediments suggest that sorption
onto sediments and particulates is an important environmental fate. Although
reliable data were not available for the solubility of BaP, it is known that
the solubility is extremely low in pure water.
The oxidation and photochemistry of polynuclear aromatic compounds were
reviewed as part of a recent task report for EPA-OTS (Radding et al., 1976).
Autoxidation of BaP by peroxy radicals appears to be slow. Photooxygenation
(reaction with singlet oxygen) to produce quinones is expected to be an impor-
tant degradative process in the environment. BaP appears to serve as both the
sensitizer and the reactant hydrocarbon.
The extreme insolubility of BaP in water has made it necessary to conduct
the laboratory work in organic or mixed organic-aqueous solvent systems. The
photodegradation of BaP has been carried out in solution and sorbed on calcium
carbonate (Andelman and Suess, 1971) and on kaolinite (McGinnes and Snoeyink,
1974). When corrected to environmental solar flux values, these data indicate
that BaP may have a half-life of less than a day both in solution and in the
sorbed state. BaP has been reported to show an increased half-life beyond
about 60% photodegradation when sorbed on kaolinite) the authors ascribe this
reduction of the rate of photodegradation to competitive light sorption by the
quinone products. Since only small yields of these quinones are isolated, this
explanation is plausible. The solution photochemistry shows no such retarda-
tion of rate.
Gibson (1975, 1976) and Gibson et al. (1975) reported on the microbial
conversion of BaP to cis-9,10-dihydroxy-9,10-dihydrobenzo[a]pyrene and more
extensive degradation by Beijerenckia bacteria isolated from a polluted stream.
The wild strain was obtained by enrichment procedures using biphenyl as the
sole source of carbon for growth, and a mutant, B-836, was prepared by treating
the parent with N-methyl-N-nitrosguanidine (Gibson et al., 1973). The wild
strain oxidized BaP to acid products (Gibson, 1975, 1976; Gibson et al., 1975).
Gibson et al., 1975, reported a 1.7% yield of a fraction from B-856 conversion
of BaP that was primarily the cis-9,10-dihydroxy-9,10-dihydrobenzo[a]pyrene
and possibly a small amount of the 7,8-dihydrodiol isomer.
Shabad et al. (1971), in a continuation of many studies on the analysis,
presence, and degradation of BaP in the environment, demonstrated some of the
intricacies of research with this carcinogen. They found that some soils from
oil refinery plants contained up to 200,000 yg BaP kg"1, and from some BaP-
contaminated soils they isolated strains of bacteria that could decompose
polyaromatic. Some strains that could not decompose BaP in their liquid medium
decomposed 40% to 86% of the BaP when added to sterilized BaP-contaminated
soils.
Other studies further indicate the many characteristics of BaP degrada-
tion research (Fedoseeva et al., 1968; Khesina et al., 1969; Lohrbacher et al.,
1971; Malaney et al., 1967; Niaussat and Ottenwalder, 1969; Poglazova and
Meisel, 1971; Poglazova et al., 1966, 1967a,b, 1968, 1971, 1972; Shabad, 1968;
Sisler and Zobell, 1947).
66
-------
To a large degree the source of the BaP metabolizing cultures has been
material with long exposure to relatively high concentrations of BaP. Single
cell isolates have been used, the level of BaP available to the microorganisms
has been on the order of pg ml"1, other nutrients were present, and many
species of organisms have been implicated. Included in the identified de-
grading organisms are the yeasts Endomyces magnusii and Candida lipolytica and
bacteria such as strains of Bacillus megaterium, Mycobacterium flavum, M.
rubrum, M. smegmatis, Bacillus sphaericus, Pseudomonas aeruginosa,
Azobacter chrococcus, and Escherichia coli.
While we anticipated that sorption onto sediments and photolysis were
probably the most important pathways for BaP, the literature data were in-
sufficient to allow us to quantitatively estimate the importance of these
processes in comparison to biodegradation, volatilization, and oxidation.
Therefore, we decided that screening studies should be conducted for each of
these processes.
6.3 ENVIRONMENTAL ASSESSMENT
6.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigations of the
transformation and transport processes for BaP are summarized in Table 6.2.
TABLE 6.2. SUMMARY OF BENZfa]PYRENE LABORATORY DATA
Process
Sorption
Volatilization
Photolysis
Oxidation
Hydrolysis
Biodegradation
o
Sorption equilibrium Partition coefficient
S = K S K = (7.6 ± 2.4) x 10A
s p w p
Rate expression Rate constant at 25°C
kv[BaP]b k = k° x (0.0036 ± 0.0006)
k [BaP] = $ (EZxex)[BaP] k = (2.8 ± 0.2) x 10~" sec"1
k [RO.-lfBaP] k = 1.86 x 103 M"1 sec"1
OXL 2JL J ox
NA k = 0
u
, max 1 _d
^I O \T T7 *^1_ O
oZ Y K bZ
s
On Coyote Creek sediment.
See discussion in Part I, Section 5.3, and Appendix B.
"Noonday sunlight, in mid-December.
Biodegradation was not observed in the enrichment procedures,
67
-------
6.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved BaP calculated for individual transformation
or removal processes are listed in Table 6.3. Although these half-lives gener-
ally change from one water body to another, the half-life for photolysis is an
order of magnitude smaller than the half-lives of the other transformation
pathways. Sorption half-lives have not been measured, but they are probably
at least 100 times smaller than the half-lives for photolysis (Section 6.5.1).
Moreover, since the sorption partition coefficient is at least 50,000
(Section 6.4.4), it is probable that most of the BaP entering natural waters
is trapped in sediments. Subsequent loss from the sediments cannot be assessed
with our data, although loss by biodegradation is plausible (Section 6.6).
TABLE 6.3. TRANSFORMATION AND TRANSPORT OF BENZO[a]PYRENE
PREDICTED BY THE ONE-COMPARTMENT MODEL
Eutrophic Eutrophic Oligotrophic
Process River pond lake lake
Photolysis, half-life (hr)a 3.0 7.5 7.5 1.5
Oxidation, half-life (hr) >340 >340 >340 >340
Volatilization,
half-life (hr) 140 350 700 700
Hydrolysis, half-life (hr) NA NA NA NA
Biodegradation,
half-life (hr)b >10* >10* >10A >10A
Half-life for all processes,
except dilution (hr) 2.9 7.3 7.4 1.5
Half-life for all processes,
including dilution (hr) 0.48 7.3 7.4 1.5
Amount BaP sorbedc(mg m3) 5 15 2.5 2.5
Percentage BaP sorbed 83% 93% 71% 71%
*a
Based on the photolysis rates estimated for summer sunlight.
Biodegradation of BaP was not observed in our enrichment procedure.
However, the literature suggests that biodegradation of BaP may occur
in the environment (see Sections 6.3.5 and 6.6).
f* w 1
1 ng ml BaP is assumed in the solution phase.
68
-------
6.3.3 Persistence
The half-life, and hence the residence time, of BaP in the water column
attributable to transformation and volatilization is expected to be independent
of the dilution rate in lakes and ponds (Table 6.3). However, dilution may
account for as much as a tenfold reduction time in short river segments, per-
haps 1-20 km in length, since the major portion of the sorbed BaP should be in
the slowly settling, small particles. Effects of dilution should be less for
longer river reaches due to the sedimentation of BaP-contaminated particles.
6.3.4 Mass and Concentration Distributions Calculated Using
Computer Models
Table 6.4 summarizes the distribution of mass and concentration of BaP
expected at steady stateiwithin each of four types of water bodies, as calcu-
lated with the aid of the computer model. The pseudo-first-order rate con-
stants used in these simulations are presented in Appendix A.
The sediments and suspended solids clearly dominate the distribution
of BaP in the pond and lake simulations, with at least 98% of the BaP in the
bottom sediments and at least 70% of the remainder in the suspended fraction.
The quantities of BaP predicted to be in solution are on the order of
10 5 to 10~6 ng ml"1 in the pond and lake simulations. Only the river simu-
lation predicts concentrations approximating those encountered in natural
waters (1.5 x 10~3 ng ml"1), and even these concentrations are much lower
than the concentrations typical of surface waters (Suess, 1976; Andelman and
Snodgrass, 1974). Three explanations are possible: either the field data
include the BaP on the very fine suspended particulates as well as dissolved
BaP, or there are processes operative under natural conditions that were not
included in our study and that enhance the solubility of BaP or increase its
stability (such as complexation with humic acids), or the BaP concentrations
that we assumed for the inflows in the simulations are much too low. We
cannot discriminate among these three sources of disagreement with the data
in hand, but the first one is the simplest and most plausible.
However, these differences do not alter the conclusions that follow
from the simulations. Even if the BaP concentration were raised by two or
three orders of magnitude, the predicted distribution and transformation
half-lives would be the same, because the rate expressions would still be
pseudo-first order. That is, adjusting the concentration in the simulation's
inflows will proportionately change the concentration in each compartment.
The effects of incomplete mixing are also illustrated quantitatively
by Figures 6.1 through 6.4, which show the patterns of buildup and decline
of BaP in the pond, river, and lake simulations. Figures 6.1 and 6.2 show
the trends in the concentration of BaP on suspended solids in the pond and
river, but these paralleled the trends of dissolved BaP and are omitted from
Figures 6.3 and 6.4 for simplicity. In all four simulations the concentrations
of dissolved BaP rapidly approach steady states both before and after stoppage
69
-------
TABLE 6.4. DISTRIBUTION OF BENZO[a]PYRENE IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations of 1 ng ml"1 benzo[a]pyrene)
Pond
Compartment 1
(surface water)
Solution
Suspended solids
Mass
(g)
2.9 x 10-5
4.2 x 10~4
Cone.
(ng ml"1)
1.5 x 10~6
7.0 x 10~2
River
Mass
(g)
4.5 x ID"2
2.2 x 10"1
Cone.
(ng ml"1)
1.5 x 10"3
7.4
Eutrophic Lake
Mass
(g)
2.7 x 10"3
6.7 x 10"3
Cone.
(ng ml"1)
1.1 x 10~5
5.4 x 10"1
Oligotrophic Lake
Mass
(g)
4.7 x 10"4
1.2 x 10"3
Cone.
(ng ml-1)
1.9 x 10~6
9.3 x 10"2
Compartment 2
(surface water)
Solution
Suspended solids
Compartment 3
(surface water)
Solution
Suspended solids
Compartment 5
(bottom water)
Solution
Suspended solids
Compartments 7-9
(sediment)3
Solution
Solids
3.6 x 10-7
4.7 x 10~2
1.4 x ID"6
7.0 x 10~2
4.2 x 10~2 1.4 x 10-3
2.0 x 10"1 6.7
3.7 x lO"3 1.5 x 10~6 2.0 x 10~4 7.9 x 10~8
9.4 x 10~3 7.5 x ID'2 4.9 x 10~4 3.9 x 10~3
3.7 x 10~2 1.2 x 10"3 1.8 x 10~4 7.4 x 10~6 5.2 x 10~6 2.1 x 10~8
1.8 x 10-1 6.1 4.6 x 10~4 3.7 x 10~2 1.3 x 10~5 1.0 x 10~3
5.5 x 10~5 2.2 x 10~8 5.2 x 10~7 2.1 x 10~10
1.4 x 10~4 1.1 x 10~3 1.3 x 10~6 1.0 x lO-5
1.0 x 10"3 1.35 x 10"4 2.2 x 10~5 2.6 x 10"6 2.3 x 10~6 2.7 x 10~7
1.4 x 10~2 6.7 1.9 1.4 x KT1 3.1 x ID"1 2.3 x 10~2
Total mass
4.7 x 10-2
140
2.0
3.2 x ID"1
The amounts given for solid and solution phases in the sediment compartments are estimated from the sorption partition
coefficient for suspended solids and may be overestimated because it was assumed that biodegradation of sorbed
material does not occur. t
-------
7 x 10'2
10
10
100 150
TIME - hours
200
250
FIGURE 6.1 PERSISTENCE OF BENZO[a]PYRENE IN A TWO-COMPARTMENT POND SYSTEM
71
-------
8 x 10
5 6
TIME - hours
10
FIGURE 6.2 PERSISTENCE OF BENZO[a]PYRENE IN A PARTIALLY MIXED RIVER SYSTEM
72
-------
3 x 10'2
10
.-2
r
10"
z
LU
cc
a.
O
N
Z
UJ
CD
O
Z
UJ
U
O
U
10-6
10'7
ID"1
10'
,-9
COMPARTMENT
SEDIMENTS
SOLUTION
DISCHARGE
STOPPED
100
200 300 400
TIME - hours
500
600
700 720
FIGURE 6,3 PERSISTENCE OF BENZO[a] PYRENE IN A EUTROPHIC LAKE
73
-------
6 x 1(T3
10-
10'
10-'
-10
COMPARTMENT
=//
SEDIMENTS
SOLUTION
DISCHARGE
STOPPED
100
200
300 400
TIME - hours
500
600
700 720
FIGURE 6.4 PERSISTENCE OF BENZO[a]PYRENE IN AN OLIGOTROPHIC LAKE
74
-------
of discharge. Sediment concentrations in each simulation rise relatively
slowly and decline exceedingly slowly after discharge stops. Stoppage of
discharge causes a consistent 100-to 1000-fold decrease in the concentration
of dissolved BaP, but in all cases desorption from the sediments is sufficient
to maintain a very low concentration of BaP in solution, roughly equal to the
concentrations in unpolluted groundwaters (Suess, 1976; Andelman and Snodgrass,
1974).
6.3.5 Discussion
The greatest uncertainty in the foregoing analyses, the transformational
pathway of sorbed BaP, is less easily resolved. The stability of BaP is
clearly affected by the presence of other organic compounds (Suess, 1972), but
the net effect of other organic pollutants, and hence the role of chemical
transformation, cannot b4 assessed with confidence. It is quite possible that
the presence of other organics in polluted waters will increase the solubility
of BaP, but waters that are polluted enough for this to occur will probably
be too turbid for significant photolytic degradation (Andelman and Snodgrass,
1974).
Shabad et al. (1971) obtained some field data from experimentally con-
taminated agricultural fields and from industrially contaminated urban and
suburban sites that suggest that biodegradation of BaP may occur in soils.
Granting that these data appear to include minor losses due to leaching as
well as degradation and were obtained from aerobic environments, it is reason-
able to presume that they are within an order of magnitude of the rates to be
expected in aquatic sediments. If this assumption is indeed true, it is
probable that the half-life of BaP in aquatic sediments is one to ten years,
probably at least three to four years.
Thus it appears probable that virtually all the BaP entering all but
the cleanest natural waters is already sorbed on, or will rapidly sorb on
suspended sediments and will accumulate in the bottom sediments. It will be
lost from the sediments by biodegradation with a half-life of five to ten
years.
6.4 PHYSICAL PROPERTIES
6.4.1 Solubility in Water
The solubility of BaP in water at room temperature (about 22 ± 2°C)
was measured using the methods described by Haque and Schmedding (1975). The
average solubility was 1.2 ' 0.1 ng ml"1 (4.8 x 10~ M), based on replicate
determinations.
6.4.2 Absorption Spectrum
BaP absorbs light strongly in the solar region to 420 nm.
75
-------
The absorption spectrum of BaP in 20% acetonitrile/80% water was
measured from 295 to 650 nm. The pH of the solution was 6 and was not ad-
justed. Care was taken to assure that the BaP was in solution and not a
suspension. No particulate could be observed by light scattered by a colli-
mated beam of green light (the green filter absorbed light below about 450 nm
and eliminated the strong fluorescence observed in a beam of white light that
obscured the scattered light).
The absorption coefficients at wavelength intervals from 297.5 to
420 nm and at wavelengths of 313 and 366 nm are reported in Table 6.5.
6.4.3 Volatilization Rate
The volatilization rate of BaP was measured using the method of Hill
et al. (1976), which is described in detail in Part I, Section 5.3. The
volatilization half-life of BaP in aqueous solution was about 22 hours at a
fast stirring rate. The rate constants were calculated as follows:
* 0 _i
Oxygen reaeration rate: k = 8.41 ± 1.51 hr
R'l p _
BaP volatilization rate: k = 0.030 J- 0.002 hr *
v
n
Ratio: k /k = 0.0036 ± 0.0006
v v
While this is an appreciable volatilization rate, the estimates re-
ported in Table 6.3 show that volatilization of BaP is slow compared with
photooxidation. This is largely because most of the BaP is sorbed on sedi-
ment, and volatilization of sorbed material is presumed to be very slow.
6.4.4 Sorption on Clay and Sediments
The sorption partition coefficients of BaP were measured on four sedi-
ments: Ca-montmorillonite, Des Moines River sediment, Coyote Creek sediment,
and Searsville Pond sediment. Care was taken to assure that the samples were
protected from light at all times to minimize photooxidation of the BaP. The
initial concentration of BaP in each sample was always less than the satura-
tion concentration of 1.2 ng ml"1 (4.76 x 10~9 M) . Each experiment consisted
of samples at one BaP concentration and two of three sediment loadings, plus
blanks that contained only BaP. Duplicate flasks were analyzed for all sedi-
ment except clay. Periodic checks were made for sorption of BaP onto the
glassware; since none of the glassware extracts contained BaP, we presumed
that no sorption onto the glassware took pJace. The data, summarized in
Table 6.6, were fitted to a Freundlich isotherm.
Graphs of the data are shown in Figure 6.5 with the partition coeffi-
cients (K ) . The calculation of the partition coefficient was based on the
&
This value represents the average and standard deviation of five reaeration
measurements taken over the period of the experiment.
76
-------
TABLE 6.5. ABSORPTION SPECTRUM OF BENZO[a]PYRENE
IN 20% ACETONITRILE/80% WATER AT pH 6a
Center of Average
wavelength interval*3 absorption coefficient0
(nm) (M"1 cm-1)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330.0
340.0
350.0
360.0
370.0
380.0
390.0
400.0
410.0
420.0
46,600
27,700
13,900
6,670
4,840
3,970
3,890
3,650
3,730
3,570
3,650
5,400
8,330
12,300
18,100
19,680
21,910
15,160
2,100
1,100
0
aBenzo[a]pyrene concentration was 0.318 Ug ml
(1.26 x 10-6 M).
The wavelength intervals are given in Appendix B,
Table B.I.
r*
The absorption coefficients at 313.0 and 366.0 nm
are 3,730 and 23,650 M"1 cm""1, respectively.
77
-------
100
Q
LU
C/J
QC
O
O
EC
111
O.
O
111
CO
tr
O
0>
I I I
arsville Pond Sedime
A Coyote Creek Sediment, K
Searsville Pond Sediment, K = 150,000
• Des Moines Rwer Sediment, K = 35,000
0 Calcium Montmorillonite Clay, K
0.3 0.4 0.5 0.6 0.7 OB
CONCENTRATION OF BaP IN SUPERNATANT
AT EQUILIBRIUM - ng ml^lppb)
SA-4396 7?
FIGURE 6.5. SORPTION ISOTHERMS OF BENZO[a]PYRENE
78
-------
results ot the analysis of the supernatant at equilibrium. The amount of BaP
sorbed on the sediments was calculated by the difference between the amount
that was added and the amount that was found in the supernatant. To check for
material balance, sediment analyses were performed for these isotherms; how-
ever, the data are much more scattered than those from the supernatant. Re-
coveries of BaP from the supernatant plus the sediment ranged from approxi-
mately 70% to 110% (Table 6.6).
The relatively large experimental error in these data resulted from a
combination of three factors. First, the concentration of BaP in the super-
natant was less than 1 ng ml"1. Second, the extraction and concentration of
the samples required concentration of the extract of 100 ml of supernatant to
less than 1 ml of sample. Third, materials that interfered with the BaP
analysis were extracted from the water and from the sediment. There is, how-
ever, no doubt that sorption of BaP on sediments is very strong and that
sorption will be an important physical process for BaP in the environment.
BaP sorption onto sediments is strongly correlated with the organic
carbon levels in the sediments and is not related to the cation exchange
capacity. A regression of the partition coefficient versus the sediment
organic carbon levels was significant at the 0.5% level. The correlation
coefficient was 0.996. Since the Ca-montmorillonite clay, which has essen-
tially 0% carbon, did show sorption of BaP, it appears that BaP sorption in-
volves both a weak interaction with solid surfaces and a strong interaction
with organic material.
6.4.5 Biosorption
Biosorption studies were conducted with ng ml"1 BaP solutions and 0.05%
potassium phosphate buffer. Experiments using BaP solutions that did not con-
tain cells showed that significant amounts of BaP were sorbed onto the glass-
ware. These experiments are described in Appendix B. A correction was made
for sorption of BaP on glass walls by determining the recovery of BaP from
the centrifuge tubes when the cell suspensions were poured out of control
tubes.
Because of the low level of BaP, the optical densities of solutions
containing equal density concentrations of the four test bacteria (Azotobacter
beijerinkli ATCC 19366, Bacillus cereus ATCC 11778, Escherichia coli ATCC 9637,
and Serratia marcescens ATCC 13880) were reduced to 0.2. These solutions were
combined immediately before the experiment was carried out. The results from
the one-hour sorption and three-hour desorption studies are presented in
Table 6.7.
The high sorption coefficients were anticipated on the basis of our
earlier sorption studies with bacterial cell mixtures, the sorptive character-
istics on glass, and the technique that was necessary to extract BaP quantita-
tively from wet cells. This strong sorptive capacity of cells for BaP is also
discussed by Lohrbacher et al. (1974) and by Poglazova and Meisel (1971), and
accounts for the drop in interest in producing biomass for feed and food pur-
79
-------
TABLE 6.6. BENZO[a]PYRENE SORPTION ON COYOTE CREEK SEDIMENTS
Total
organic
carbon
Sediment (%)
Ca-montmorillonite 0.06
clay
Coyote Creek 1.4
CD
O
Des Moines River 0.6
Searsville Pond 3.8
Sediment
concentration
(Ug ml-1)
19
38
23
120
19
96
7.7
19
38
BaP
concentration
in supernatant3
(Ug ml-1 x 103)
0
0
0
0
0
0
0
0
0
.77
.65
.27
.69
.56
.18
.33
.19
.10
+
+
+
+
+
i
+
+
+
0.
0.
0.
0.
0.
0.
0.
0.
0.
15
10
06
013
11
15
09
04
03
BaP
concentration
on sediment
(ug g-1)
16
11
21
6.
22
7.
55
30
17
± 8
± 3
* 3
0 ± 0.1
± 6
4 ± 1.6
± 11
± 2
± 1
Partition coefficient0
K x 10""
P
LSS
a = 0 a ?* 0
0 O
1.7 ± 0.5 -2.7 ± 1.6
(a = 32 + 12)
o
7.6 ± 2.4 6.0 i 5.1
(a = 4 ± 11)
o
3.5 ± 2.7 2.5 ± 8.1
(a = 5 ± 35)
o
15 ± 2 9.7 ± 0.4
(a = 14 ± 10)
o
NLLS
S and S
w s
1.7 i 0.5
7.6 ± 4.0
4.2 ± 2.6
16.5±2.7
Concentration measured in supernatant with population standard deviation.
Concentration on sediment calculated from supernatant concentration with population standard deviation.
c
LLS = linear least squares) NLLS = nonlinear least squares; see Appendix B, Section B.I.4 for description of regressions.
Limits are 95% confidence limit.
-------
poses from liquid petroleum fractions. The elimination of BaP in activated
sludge of sewage plants is also attributed to this characteristic.
TABLE 6.7. BENZO[a]PYRENE SORPTION AND DESORPTION
ON A MIXED BACTERIAL POPULATION3
Initial BaP Sorption
concentration Sorption (S) or coefficient
(ng ml"1) desorption (D) (x 10~4)
1.01 S 33.
D 36.
0..
?5 S 44.
i D 49.
6
9
8
5
± 6
± 7
± 5
± 5
.3
.0
.4
.0
I)ry weight of bacterial mixture equivalent to
97 mg liter"1.
The adjustments in results in Table 6.7 for sorption of BaP on glassware
in sorption studies were 1.8% and 3.6%, and in desorption studies they were
3.6% and 4.0% for initial BaP concentrations of 1.01 and 0.55 ng ml"1,
respectively.
6.5 CHEMICAL TRANSFORMATION
6.5.1 Photolysis Rate
Photolysis of BaP in air-saturated water is rapid in the solar
region and gives a mixture of three quinones. Data for the photolysis of BaP
are given in Table 6.8. Acetonitrile was used as the cosolvent with water
from the sources indicated since the solubility of BaP in water alone is very
low (see Section 6.4.1). In sunlight the half-life of BaP was only 0.7 to
1.1 hour. Good first-order kinetic behavior was found for photolyses carried
to beyond two half-lives of BaP under laboratory conditions.
In the experiments using 260 and 13 ng ml"1, the photolysis rate was
independent of the BaP concentration. The quantum yields for direct photoly-
sis of BaP measured at 313 and 366 nm were 8.9 x lO"** and 5.4 x lO"*1, respec-
tively. These data indicate that there may be a slight dependence of the
quantum yield on wavelength, but this conclusion must be treated cautiously
in ,'iew oi the number of variables that might affect these measurements and
the limited number of experiments. Similar cautions apply also to the possible
small solvent effect shown in Table 6.8. For the photolysis of BaP at 366 nm
in the acetonitrile-pure water solvents, the standard deviation of the photol-
ysis rate constant among experiments is about 15% from the average of 3.77 x
10-" sec"1.
81
-------
TABLE 6.8. RATE CONSTANTS FOR PHOTOLYSIS OF BENZO[a]PYRENE
Irradiation
source
313 run
366 nm
366 nm
366 nm
366 nm
366 nm
366 nm
366 nm
Solution3
20% AN in PW
20% AN in PW
20% AN in PW
1% AN in PW
1% AN in PW
with hunic acid
1% AN in Lake
Tahoe water
1% AN in Sears-
ville Pond water
1% AN in Coyote
BaP con-
centration
UK ml-1)13
270
260
13
13
13
13
13
13
Extent of
reaction
(%)
71
84
77
87
43
54
76
46
Rate constant
k x 10* sec"1
P
0.105 + 0.009C'd
3.47 + 0.12e
3.36 + 0.22
4.37 + 0.30
2.9f
0.690 + 0.041
1.76 + 0.10
1.8f
3.98 + 0.13
1.70 + 0.07
Mid-morning
sunlight,
late January
Creek water
20% AN in PW
Noon sunlight 20% AN in PW
mid-December
270
270
86
91
1.2r
1.79 + 0.39
S,h
2.79 + 0.23&'-
AN = acetonitrile; PW = pure water.
D1.00 ng ml'1 BaP in water = 3.96 x 10~9 M.
i->
"Standard deviation.
Quantum yield for BaP disappearance at 313 nm is 8.9 x 10" A.
g
"Quantum yield for BaP disappearance at 366 nm is 5.4 x 10"*.
One data point for each photolysis experiment. Reactions were photolyzed
simultaneously in merry-go-round reactor for direct comparison of the
effect of the three waters indicated.
Rate constant not corrected for variation in sunlight intensity during
experiment.
^Half-life of 1.1 hour.
LHalf-life of 0.7 hour.
82
-------
While these differences are considered negligible in an environmental
assessment, the small variations in data between experiments may be subtle,
but real, phenomenological effects and not experimental error. Further studies
to elucidate the mechanism would clarify these effects, but such studies are
beyond the scope of this project.
The photolyses of BaP conducted in natural waters or in pure water
containing humic acid showed slower photolysis rates than photolyses conducted
in pure water alone. Photolysis of 13 ng ml"1 BaP in the presence of humic
acid in pure water (absorbance at 366 nm is 0.11) is five times slower than
the photolysis in pure water: the other natural waters showed intermediate but
definite retardation of the BaP photolysis. It is not known why the water from
Lake Tahoe shows greater inhibition of BaP photolysis than the water from the
Searsville Pond. It is unlikely that a light screening effect is responsible
for the slowed photolysis rates since the absorbance of the natural waters
at 366 nm is very small (< 0.02); the 0.11 absorbance of the humic acid solu-
tion at 366 nm should1 reduce the photolysis rate 13% if light screening is the
only effect operative.
The reason for the slower photolyses in natural waters can only be spec-
ulated upon at this time. It is not known how much or what types of organic
material are present in these natural waters. It is possible that substances
present in the natural waters (either organic or inorganic) may quench a BaP
excited state, singlet oxygen (if it is an intermediate oxidant), or any
excited state complex that occurs in the reaction.
Another possible process by which photolysis in natural waters may be
inhibited is by formation of a complex of BaP and natural substances (organic
or inorganic) in solution, which would alter the reactivity of ground state
BaP itself. Evidence for complex formation was found in an experiment with
approximately 30 ug ml~x humic acid and 270 ng ml"1 BaP in which about 50% of
the BaP could not be recovered from the aqueous solution by hexane extraction.
An example of possible inhibition of BaP photolysis on kaolinite clay
caused by complex formation during photolysis was reported by McGinnes and
Snoeyink (1974); products formed during photolysis are apparently responsible
for this effect. Such effects may be occurring by several mechanisms, in-
cluding competitive reactions of the associated organic material with the
oxidizing agent(s) formed during the BaP photolysis. If lipophilic/hydrophoic
compounds such as BaP do form such complexes, the photolysis rates of BaP in
lakes and streams may be much slower than in pure water solutions.
The half-life for direct photolysis of BaP in sunlight as a function of
the time of day was calculated by the procedure of Zepp and Cline (1977) using
a quantum yield of 8.9 x 10-£l and the measured UV spectrum of BaP (Section
6.4.2). The data for the summer and winter seasons are plotted in Figure 6.6.
The calculated half-life of 1.2 hour for midday photolysis in winter is in
excellent agreement with the measured half-lives of 1.1 and 0.7 hours measured
during the same season.
83
-------
I
HI
z
<
I
MEASURED -\
/ (WINTER)
I
AM
PM
6
6
8 9 10
432
TIME OF DAY
11
1
12 NOON
FIGURE 6.6 SEASONAL AND DAILY VARIATION OF PHOTOLYSIS
HALF-LIFE OF BENZO[aJPYRENE
6.5.2 Oxidation Rate
When the susceptibility of BaP to free radical oxidation was examined
using the AA-initiated oxidation reaction, BaP was found to be completely
consumed after the usual 100-hour reaction time. The oxidation experiment
was then repeated using a 270 ng ml"1 (1.0 x 10~6 M) BaP solution (20% ace-
tonitrile in water) containing 1.0 x 10"** M AA, and analyzed for BaP at reac-
The data from this experiment gave a first-
]) of (5.7 ± 1.1) x 10~5 sec"1. Under the
Appendix B, Section B.2), this corresponds to
tion times of 4.5 and 6.0 hours.
order rate constant term
(k [RO
reaction conditions at 50°C°X(se
a second-order rate constant
then be 1.86 x 103 M-1 sec"1
k of 1.68 x 10" M"1 sec"1. At 25°C, k would
, °x and assuming a [ROx] = 10"9 M, the haff-life
84
-------
for oxidation of BaP in aquatic environments would be 4.3 days. While this
is a reasonably fast process, the nonphotolytic oxidation of BaP is clearly
not competitive with photolysis or sorption of BaP under environmental
conditions.
6.5.3 Hydrolysis Rate
BaP contains no groups that are hydrolyzable.
studies were carried out.
Therefore, no hydrolysis
6.5.4 Products from Chemical Transformation
Products from the oxidation and photolysis of BaP have been reported
to be quinones of the BaP structure (Radding, 1976); however, no material
balances for the reactions have been reported. In our studies, the primary
products resulting from oxidation and photolyses of BaP were tentatively
identified as quinones by comparing their HPLC retention times with those
of authentic samples obtained from the National Cancer Institute. The quinones
found are shown below:
BaP-3,6-quinone
BaP-1,6-quinone
BaP-6,12-quinone
The yield of quinones and the material balances for the free radical oxidation
and the photolyses at 366 nm and in the sunlight are listed in Table 6.9. The
material balances for the oxidation and photolyses are good. This is contrary
to what would be expected from the work of McGinnes and Snoeyink (1974), who
reported that the quinone products are more photolabile than BaP itself. The
material balances in Table 6.9 show that the chemical transformation products
are accounted for by the three quinones.
85
-------
The data in Table 6.9 are interesting in several respects. The product
ratio of the 1,6-quinone and the 3,6-quinone is about 3 to 1 in both the sun-
light and the 366 nm photolyses; only traces of the 6,12-quinone were found.
This result suggests that the formation of the products of BaP photolysis is
independent of wavelength.
The mechanism for the photolytic transformation is not known. While
singlet oxygen may be involved, it is doubtful that there is an endoperoxide
intermediate involved; an alternative mechanism of electron-transfer/super-
peroxide radical anion oxidation is also a possibility. It is also inter-
TABLE 6.9 PRODUCT YIELDS AND MATERIAL BALANCE FOR
BENZO[a]PYRENE PHOTOLYSES AND FREE-RADICAL OXIDATION
Reaction
AA-initiated
oxidation of
2700 ng ml"1
BaP in 20% AN
Average
Sunlight
photolysis of
270 ng ml"1
BaP in 20% AN
Average
Photolysis
at 366 nm
of 270 ng ml-1
BaP in 20% AN
% BaP
reacted
50
74
100
50
55
74
25
56
% Yield3
1 , 6-quinone
36
35
40
37
54
78
81
71
80
77
% Yield3
3 , 6-quinone
22
18
24
21
16
24
26
22
20
20
% Yield3
6 , 12-quinone
32
27
40
33
Trace
Trace
Trace
Trace
Trace
Trace
Material
balance
(%)
90
80
104
70
102
107 -
100
96
Photolysis
at 366 nm
of 270 nm ml-1
BaP in 50% AN
8.5
19
71
66
24
18
Trace
Trace
95
84
Based on BaP reacted.
""Material balance = (BaP consumed/quinones formed) x 100.
"Trace, discernible peak in HPLC, but less than 5% yield.
86
-------
esting to note that the same quinone products in about the same product ratio
(3 to 1 ratio of 1,6-Q to 3,6-Q) have been reported for the ozonolysis of BaP
(Moriconi et al., 1961).
Table 6.9 shows that free radical oxidation of BaP also produces the
6,12-quinone in significant yield. The mechanism of the oxidation reaction
is unknown, probably involving RO • radicals and other oxidizing species,
and is likely different from that of the BaP photolysis. However, oxidation
at the 6-position on the BaP skeleton appears to be a common feature in both
types of reaction. It may be that a primary reaction occurs at the 6-position
of BaP, with the nature of the oxidant then determining the selectivity of
attack at other carbons on the ring (e.g., positions 1, 3, or 12).
6.5.5 Reaction with Singlet Oxygen
Several studies of the photolysis of polynuclear aromatic compounds
(PAHs) have implicated singlet oxygen (*02) as the oxidant in the formation
of quinone products (National Academy of Sciences, 1972). One suggested
mechanism is that the PAH serves as a singlet oxygen sensitizer and as a
reactant with X02 (Stevens and Algar, 1968).
PAH ^-> ^AH* (6.1)
XPAH* »- PAH (6.2)
1PAH* ~~> 3PAH* (6.3)
"PAH* + 302 —> PAH + X02 (6.4)
3PAH* > PAH (6.5)
3PAH + 302 —»• M02 (6.6)
102 + PAH -2+ M02 (6.7)
'Oa -^ 302 (6.8)
where the asterisk denotes an electronic excited state; 302 and X02 are the
triplet (the ground state) and singlet oxygen; PAH, 1PAH*, and 3PAH* are the
ground state, singlet, and triplet excited states of PAH; and M02 is the
oxidized PAH product or products. Reaction of 3PAH with 302 to form a complex
(exciplex) that gives products directly without formation of free 102 is also
a reasonable alternative mechanism.
3PAH* + 302 —>- (PAH, ^2)" (6.9)
(PAH, ^a)*-*- M02 (6.10)
87
-------
Since photolysis of BaP was found to be rapid in aqueous solution
giving quinone products (see Section 6.5.4), it was of interest to determine
the reactivity of ground-state BaP with 102. One measure of such reactivity
is the ratio of the rate constant k8 for the decay of X02 to ground state
( 02) to the rate constant k7 for the reaction of 102 with substrate (in this
system the BaP) . This ratio is referred to as the 3 value:
3 = (k8/k7)M~1 (6.11)
3 is also equal to the concentration of substrate required to trap half of
the singlet oxygen in the reaction system. Values of 3 are solvent dependent
because ke changes with solvent while k7 does not (Stevens et al., 1974).
The value for ka in different solvents has been measured by Merkel and
Reams (1972).
To evaluate k7 for BaP, (k p) , we used a competitive reaction tech-
nique in which BaP and a reference compound, for which k7 and B are known,
were exposed to singlet oxygen in the same reaction solution. Methanol was
used as the solvent for this study. Since the same concentration of singlet
oxygen is available to both reactants in solution, the ratio of the first-order
rate constants k^ and k' , corresponding to the disappearance of BaP and
reference compound, respectively, gives the relative reactivity of BaP to the
reference compound. With this reactivity ratio and the known absolute value
of k7 for the reference compound (k ) , the value of k in water can be
-i -< j "L Bar
calculated.
kRC=k
BaP
From the value for k and the known value for ke in water, the 3 value for
BaP in water can be calculated.
Singlet oxygen was generated in our experiments using the sensitizer
rose bengal. This sensitizer absorbs light above 500 nm, well beyond the
highest wavelength absorption of BaP or other PAHs.
Sens -^ "Sens" (6.13)
Sens •»'wv* Sens (6.14)
3Sens* + 302 y Sens + r02 (6.15)
Thus singlet oxygen is formed independently of the PAH, and the 3 values are
then related to the true bimolecular reaction (6.7) and without any contri-
bution to oxidation due to the possible PAH-102 complexes, reactions (6.9)
and (6.10). Photolyses were conducted using two 750-watt tungsten lamps and
a Corning 3-71 filter to cut off all light below 450 nm.
88
-------
Initial experiments to measure the 3 value for BaP found that it
reacted much slower with singlet oxygen than a reference compound, 1,3-
diphenylisobenzofuran (P = 0.06). Benza[a]anthracene (BA) was then chosen
as a reference compound because its 0 value was known (870 ± 250 at 25°C in
benzene solvent), because it reacted at a rate comparable to BaP, and because
HPLC analysis clearly resolved both PAHs (k7 for BA = 4.8 x 10**; Stevens
et al., 1974). The data for the rose bengal sensitized, singlet oxygen re-
action (also called photooxygenation) of BA and BaP in 10% acetonitrile in
water are given in Table 6.10. Rate constants for loss of BA and BaP (kpr
and k' p, respectively) were evaluated from first-order plots of In[PAH]
versus time.
TABLE 6.10. RATE CONSTANTS FOR SINGLET OXYGEN ADDITION
TO BENZO[a]PYRENE AND BENZ[a]ANTHRACENE3
Concentration
\ of PAH
Experiment
1
2
(M x
BaP
1.00
1.00
10s)
BA
1.00
1.20
k' x 105
BaP
0.97 ± 0.02
0.94 ± 0.09
sec 1
BA
2.04 ±0.05
2.28 ± 0.18
Ratio
kBA/kBaP
2.11
2.42
Average 2.26
Solvent was 10% acetonitrile in methanol with 2 x 10
rose bengal sensitizer.
-5
M
The averaged value of 2.3 for the ratio of the reactivity of BA to
BaP is in accord with the similar polycyclic structure of the two PAHs. It
is also interesting that the quantum yield for direct photolysis of BA at
366 nm was 3.3 x 10~~A (Section 5.5.1), slightly more than three times larger
than that of BaP (8.9 x 10"* at 366 nm). While perhaps fortuitous, the close
agreement between the relative reactivity of BA and BaP to singlet oxygen and
the ratio of direct photolysis quantum yields suggests that the direct photol-
ysis of these PAHs may involve reaction with X02 [reactions (6.4), (6.7), or
(6.9), (6.10)].
As discussed above, the value of k7 for reaction between a substrate
and 102 is independent of solvent. Using BA as the reference compound, k^
is then equal to 4.8 x 10** M"1 sec"1 at 25°C. From this rate constant and
the ratio kgap/kj = 0.44, kg p is calculated as 2.1 x 10" M"1 sec"1 at 25°C.
Based on a value for ke in water solvent = 5 x 10s sec"1 (Merkel and Kearns,
1972), the 3 value for BaP is calculated to be 24 M"1. For comparison, the
P value for BA in water is 10 M"1.
89
-------
6.6 BIODEGRADATION
6.6.1 Development of Enrichment Cultures
The three sets of experimental conditions used in attempting to develop
BaP degrading enrichment cultures failed to yield systems indicating any sig-
nificant breakdown of this carcinogen.
In one series, BaP was ground to a small particle size in a suspension
by the use of a low-clearance Potter-Elvehjem homogenizer. The BaP was added
at 3 yg ml"1 concentrations to the original 100 ml of water samples plus
buffer and ammonium salts in 500-ml Erlenmeyer shaker flasks, and to the basal
salts media used in subsequent transfers in 250-ml Erlenmeyer shaker flasks.
The first transfers from the flasks containing water samples from the
eutrophic pond near Searsville Lake in Woodside, California, and aeration
effluents from the wastewater treatment plants of Palo Alto, South San
Francisco, the Shell Oil Refinery in Martinez, California, and the Monsanto
Chemical Company, Anniston, Alabama, were arbitrarily made after 4 days in-
cubation at 25°C in a rotatory shaker. Subsequent transfers (2% vol) to fresh
media with suspended BaP appeared to develop 30% to 40% reduction of BaP, but
these results were obtained before fermentation broths were frozen and thawed
three times in the presence of some solvent before the final extractions were
made. At the time of the seventh transfer from each water sample, transfers
were also made to basal salts media with no added organic substrate. After
4 days of incubation, microbial counts in all flasks were all in the range of
(2.7 to 5.8) x 106 ml"1. The freeze-thaw (3 times) extraction procedure resulted
in complete recovery of the 3 yg ml"1 BaP, indicating no biodegradation of
BaP. Because there was an increase in numbers of organisms when new media
were inoculated, but there was no degradation of BaP, the organic material for
growth of the organisms had to originate from the distilled water used in the
preparation of the media and/or from the atmosphere.
The second series of enrichment studies was initiated with naphthalene
and pyrene as added organic substrates. The concept was that BaP may be so
highly sorbed by the organisms and particulate matter in the water samples that
there was insufficient free BaP for the growth of organisms that could or would
develop the capacity to degrade BaP. It was possible that cultures could
develop that would degrade naphthalene and/or pyrene, and then because of the
similarities of these compounds to parts of the structure of BaP, a BaP-bio-
degrading enrichment culture might develop if BaP were present in subsequent
transfer media.
To the 9-liter aerated bottles with eutrophic pond water and the aera-
tion effluents from the South San Francisco, Shell Oil Refinery (Martinez, CA),
and Monsanto Company Plant (Anniston, AL), naphthalene and pyrene were added
in DMSO equivalent to 50 and 1 yg ml"1, respectively, at days 1 and 2, and
100 and 1 pg ml"1, respectively, at days 3 and 4. On day 5, a transfer was
made into shaker flasks containing basal salts medium to which naphthalene
(300 yg ml"1) and pyrene (6 yg ml"1) had been added.
90
-------
Subsequent repeated transfers, made into shaker flasks every three days
into media containing naphthalene (600 pg ml-1), pyrene (12 pg ml"1), and BaP
(3 pg ml-1) all developed good growth as had been expected with naphthalene
providing organic carbon for growth. The fermented broths, after up to 11
days of fermentation, did not indicate any loss of BaP in the shaker flasks
when the broths were subjected to three freeze-thaw cycles in the presence of
ethyl acetate used for the first extraction.
The third series of enrichment studies was initiated with BaP (1 pgrnl"^)
and naphthalene (10 pg ml-1) and a weekly addition of naphthalene (10 pg ml-1)
to the contrary 9-liter first-stage enrichment fermentors with a water
sample from Coyote Creek and aeration effluent from the Palo Alto sewage
plant. During the six weeks of this study, degradation of BaP was not
observed.
6.6.2 Biodegradation' Kinetics
Since enrichment cultures that degraded BaP were not obtained, biodegra-
dation kinetic studies were not performed.
6.6.3 Discussion
The literature results summarized in Section 6.2suggest that, with long
exposure of microorganisms in the environment to BaP or other polyaromatic
hydrocarbons, there is a natural preselection and induction process that did
not occur within the time limitations of our enrichment studies. Our inability
to develop BaP degrading systems may also be due to the characteristics of the
water samples. Several of the reports from the laboratories of Poglazova and
Shabad used levels of BaP far above the level of solubility. With the high
sorption coefficientsof BaP on microorganisms and particulate matter and the
low water solubility of BaP, there is little BaP in solution for the growth of
specific microorganisms that may be able to metabolize or adapt to the metab-
olism of BaP when it is the only available carbon nutrient. These conditions
could put such microorganisms at a disadvantage with organisms having more
favorable predator-prey or microbial interaction relationships. To date, no
BaP biodegrading culture systems have been developed when BaP was used as a
sole carbon source.
The phenomena of microbial interactions or predator-prey relationships
depend on many factors including the species of organisms; rates of growth;
availabilities and characteristics of nutrients present as natural contaminants
or released by excretion or autolysis of other organisms; symbiosis or antag-
onisms by modifying pH, oxygen tensions, and/or production of new substrates;
growth of inhibitory substances; dilution rates of constitutive organisms;
and phagocytosis.
It is also possible that some of the degradations reported were conclu-
sions based on observations made as a consequence of artifacts of analytical
91
-------
procedures. However, with the many reports and analogies with microbial
attack on other aromatic compounds, it is reasonable to conclude that BaP can
be microbially metabolized. In aquatic systems this is probably most important
in sediments and may be the major pathway for degradation of BaP that is depos-
ited by sedimentation of microbes and other solids to which it is sorbed when
it enters a water body and escapes photolytic or oxidative degradation.
Additional discussion is given in Section 6.3.5.
92
-------
7. LABORATORY INVESTIGATION OF QUINOLINE
7.1 SYNOPSIS
The laboratory data suggest that the major degradation pathway for quino-
line in eutrophic systems is biodegradation. In oligotrophic systems, both
biodegradation and photolysis may be important. In rivers, volatilization may
be important during the winter; in other water bodies, volatilization is slow
compared with biodegradation and photolysis. Oxidation and sorption by suspen-
ded sediments are not significant compared with biodegradation and photolysis.
The nine-compartment environmental exposure model predicted the following
steady-state concentrations of quinoline in solution, suspended solids, and
sediments near point sources in the presence of a continuous discharge of
1 ug ml"1 (1 ppm) quinoline.
A Suspended
Half-life Solution solids Sediments.
(hr) (yg ml"1) (ug g"1) (yg g"1)
River 0.28 9.6 x 10"1 9.6 9.2
Pond 0.5 6.6 x 10~3 6.6 x 10~2 6.6 x 10~2
Eutrophic Q>5 1.8 x 10~2 1.8 x 10~2 1.4 x 10~* -'
lake
Oligotrophic 1-7 x 10-i 1>7 6_7 x 1Q-i
lake
7.2 BACKGROUND
Quinoline, isoquinoline, alkylated quinolines, and other derivatives of
quinoline are present in coal and coal products (Hayatsu et al., 1975; Vymetal,
1974a,b; Clemo, 1973), in diesel fuel or other petroleum products (Ben'kovskii
et al., 1974; Brodskii et al., 1975), in slightly decomposed sphagnum peats
(Naucke et al., 1972), and as components of a whole class of alkaloids. They
may be emitted to the atmosphere as a by-product of petroleum refining and
coal mining, coking, and the like. Since heterocyclic compounds containing
0, S, and N in the ring systems are particularly difficult to remove from the
effluents from coal gasification and liquefaction plants (Hayatsu et al.,
Predicted by one-compartment model.
Average concentration in sediments.
93
-------
1975), the environmental exposure to these compounds, including quinoline and
its derivatives, may increase.
Doehler and Young (1960) reported the sorption of quinoline on a Na- and
Ca-montmorillonite, illite, and kaolinite in water as a function of quinoline
concentration, pH, salinity, time, and temperature. The general trend found
for clays was that: reducing the pH increased the partition coefficient, tem-
perature had a very small effect on sorption, and the sorption equilibrium was
reached in less than 200 minutes and was not completely reversible. Although
Doehler and Young did not use the Freundlich isotherm, their data for Ca-
montmorillonite suggest that n = 1 at quinoline concentrations below about 50
yg ml-1. The value of the partition coefficients calculated from their data
for their Ca-montmorillonite at 25°C was 10.4 at pH 7.3 and 16.8 at pH 6.3.
No data on sorption of quinoline by natural sediments were available in the
literature.
Quinoline will not hydrolyze in environmental aquatic systems. No data
were available to evaluate the importance of the chemical oxidation of quino-
line in the environment. While it is known that quinoline absorbs strongly in
the solar spectrum to about 350 nm, no data on the rates or products of photol-
ysis were available for assessing its transformation in aquatic environments.
Several electron spin resonance studies on the photolysis of quinoline
and its derivatives, using the full spectrum of the Hg lamp, have been reported.
Castellano et al. (1975) detected the semiquinone radical derived from reaction
of the photoexcited quinoline with methanol solvent. Kataoka et al. (1966)
noted only that "Almost all the [quinoline] compounds were completely destroyed
by UV irradiation." No products were isolated and identified in these studies,
and it is not known how much of the photoreaction can be attributed to photol-
ysis at wavelengths greater than 290 nm. Photolysis of pyridine at 256 nm in
aqueous solution is reported to give an aminoaldehyde with a product quantum
yield of 0.07 (Joussot-Dubien and Houdard, 1967).
+ H20
It is not known whether quinoline reacts in a similar manner.
There is no information on the biodegradability of quinoline in aqueous
systems. Robbins (1916, 1917) and Funchess (1917) reported on the degradabil-
ity of quinoline in soils. More recently, Robbins (1971) reported using en-
richment procedures to isolate organisms that would degrade 2-hydroxyquinoline.
In this preliminary study, Robbins obtained an organism that would grow on
2-hydroxyquinoline, quinoline, glucose, or yeast extract as the only carbon
sources, but would not grow on coumarin, 2-methylquinoline, acridine, naphtha-
94
-------
lene, naphthols, phenol, salicylate, anthranilate, ^-aminophenol, or pyridine.
Respirometry studies, although reportedly poorly reproducible, indicated some
oxygen consumption with quinoline, 2-hydroxyquinoline, and coumarin.
The physical properties of quinoline are summarized in Table 7.1.
TABLE 7.1. PHYSICAL PROPERTIES OF QUINOLINE
Structure
Molecular \weight
(CRC Handbook, 1974)
Melting point (°C)
(CRC Handbook, 1974)
Boiling point (°C)
(CRC Handbook, 1974)
•a
Vapor pressure at 25°C (torr)
Solubility in water at 25°C
(mg ml"1)
(Albersmeyer, 1958)
(Perrin, 1965)
Ultraviolet spectrum in ethanol
[A (log e)]
max
(CRC Handbook, 1974)
9.1 x 10~3
6.11
4.8b
275 (3.51)
299 (3.46)
312 (3.52)
1 yg ml-1 is equivalent to 7.72 x 10 M
Extrapolated from data between 180° and 240°C (Maczynski and
Maczynska, 1965).
bQuinoline is 98%, 39%, and 0.6% protonated at pH 3, 5, and 7,
respectively.
95
-------
7.3 ENVIRONMENTAL ASSESSMENT
7.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory studies of the trans-
formation and transport processes for quinoline are summarized in Table
7.2.
TABLE 7.2. SUMMARY OF QUINOLINE LABORATORY DATA
Process
Sorption equilibrium
Partition coefficient
Sorption
Volatilization1'
Photolysis c
Oxidation
Hydrolysis
Biodegradation
S = K S
s p w
Rate expression
k [Q] = *(ZZAex)[Q]
[QJ
NA
*%ax
Y K
K = 11
P
Rate constant at 25 °C
kv = (2.7 ± 0.4) x 10~3 k°
kp = 7.8 x 10~7 sec"1
k = 2.8 M"1 sec-1
ox
kh-°
= 3.1 x 10~6 ml cell"1 hr"1
On Coyote Creek sediment having an organic content of 1.9%.
3See discussion in Part I, Section 5.3 of the final report.
T
"Assumes 12 hours of sunlight per day in late June.
7.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved quinoline calculated for individual trans-
formation or removal processes following a spill are listed in Table 7.3. The
half-lives in eutrophic and oligotrophic waters differ by roughly a factor of
1000, but among the three eutrophic systems half-lives differ by only a factor
of 2. Biodegradation is the dominant pathway in the eutrophic waters, and
the much slower photolytic pathway dominates in oligotrophic waters. Sorption
is relatively unimportant, and oxidation, volatilization, and hydrolysis are
unimportant in all water bodies.
96
-------
TABLE 7.3. TRANSFORMATION AND TRANSPORT OF QUINOLINE PREDICTED
BY THE ONE-COMPARTMENT MODEL
Process
Eutrophic Eutrophic Oligotrophic
Stream pond lake lake
Photolysis, half-life (hr)
Oxidation, half-life3 (hr)
Volatilization,
half-life (hr)
Hydrolysis,
half-life (hr) ;
i
Biodegradation, 1
half-life (hr) \
Half-life for all
processes, except
dilution (hr)
Half-life for all
processes, including
dilution (hr)
1,200 3,000
600
3,000
>10"
7.0 x 103 3.5 x 10" 2.8 x 10" 2.8 x 10'
NA
0.5
0.5
0.28
NA
0.5
0.5
0.5
NA
0.5
0.5
0.5
NA
10,000
600
600
Amount quinoline sorbed (mg m~3)
Percentage quinoline sorbed
1
0.1%
3
0.3%
0.5
0.05%
0.5
0.05%
Based on peroxy radical concentration of 10~9 M.
1 vg ml"1 is assumed in the solution phase.
7.3.3 Persistence
The persistence of quinoline following an acute discharge should be
virtually independent of dilution because the biotransformation rate of quino-
line is very fast. Even in the river simulation, which assumes a relatively
rapidly flowing stream, the apparent half-life at a given location is only
halved by the effects of dilution (Table 7.3). In the oligotrophic lake, where
biodegradation is expected to be relatively slow, the half-life is 25 days.
7.3.4 Mass Concentration Distributions Calculated Using Computer Models
The pseudo-first-order rate constants used in these simulations are
presented in Appendix A. The distributions of mass and concentration of
97
-------
quinoline expected at steady state during chronic discharge to each of four
types of water bodies are given in Table 7.4. The changes in concentration as
discharge begins and ends are shown in Figures 7.1 through 7.4.
The concentrations at steady state varied 100-fold between the water
bodies, as indicated by concentrations near the sources. Concentrations varied
10,000-fold among compartments within the eutrophic lake simulation but varied
only 10-fold or less within the simulation in the oligotrophic lake, where
transformation is slow, and in the river, where dilution is rapid (Table 7.4).
The approaches to steady-state concentrations were rapid in all water
bodies following start of the simulated discharge, as generally were the
transitions to new steady states after discharge stopped (Figures 7.1 through
7.4). The oligotrophic lake was an exception to the general pattern of rapid
decline, rising rapidly to concentrations approximating steady-state concen-
trations during discharge, but declining very slowly when discharge stopped.
Concentrations of quinoline on sediments declined slowly in all simulations
because of the long half-life for scouring of the sediments. Concentrations
on suspended solids paralleled concentrations in solution.
Despite the higher concentrations of quinoline on suspended solids
than in solution, virtually all the quinoline remained in solution because of
the much larger mass of water than suspended solids assumed in all the simu-
lations (see Part I of the final report). This pattern might be reversed in
actual water bodies if sediments are contaminated to depths greater than
those assumed in these simulations, but a substantial fraction of the quino-
line present in any water body can nonetheless be expected to be in the water
column and most of that fraction will be in solution.
These patterns were not changed in a sensitivity test by reducing the
concentration of quinoline in the inflow to 10 yg ml"1 and 10 yg ml"1. Fur-
thermore, since the steady-state concentrations in the pond simulation, which
are 7 x 10~3 yg ml"1 and 7 x 10~3 yg ml"1 for the 1 yg ml"1 and 10 yg ml"1
inflows, respectively, are much less than the value of Ks measured in the
laboratory (0.16 yg ml"1), the biodegradation kinetics can consistently be
expressed as:
R, _ ymax[S][X] K » [S] (7.1)
^ ~ Y K S
s
= \2 [S] [X]
7.3.5 Discussion
Quinoline appears to be toxic to aquatic organisms at concentrations
above 1 yg ml"1. For vertebrates LD50s range from 5 to 50 yg ml"1., and for
phyto- and zoo-plankton from 50 to 140 yg ml"1. Mammals appear to be more
resistant, with LD50s of 460 or more for oral and dermal exposures (Merck,
1960; Christensen and Lyginbyhl, 1975; McKee and Wolf, 1963). Consequently,
toxic concentrations are unlikely to be widespread even if concentrations in
discharges are as high as 1 yg ml"1.
98
-------
TABLE 7.4. DISTRIBUTION OF QUINOLINE IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations at 1 yg ml~1quinoline)
10
to
Compartment 1
(surface water)
Solution
Suspended solids
Compartment 2
(surface water)
Solution
Suspended' solids
Compartment 3
(surface water)
Solution
Suspended solids
Compartment 5
(surface water)
Solution
Suspended solids
Compartments 7-9
(Sediment)3
Solution
Solids
Pond
Mass Cone.
(kg) (vig g~l)
1.3 x 10~l 6.6 x 10~3 2.
3.9 x 10~l 6.6 x 10~a 2.
2 .
__ __ 7
— f, .
2.
2.
— — « __
~
1.7 x 10~3 6.6 x 10~3 6.
4.5 x 10~a 6.6 x 10~* 1.
River
Mass Cone.
(kg) (ug R~l)
9 x 10s
9 x 10"x
8 x 102
8 x 10"1
7 x 102
7 x ID"1
9 x 10"1
9 x 102
9.
9.
9.
9.
8.
8.
—
9.
9.
6 x ID'1
6
2 x 10~l
2
8 x ID'1
8
2 x 10'1
2
Eutrophic Lake
Mass Cone.
(ks) (UK a'1)
4.5
2.2
1.1
5.4
2.1
1.1
4.9
2.5
1.2
6.3
x 10" 3
x 10~"
xlO-2
x 10- '
x 10'3
x 10~*
x 10~s
x 10"1
•---
1.8 x 10~a
1.8 x 10"1
4.3 x 10~*
4.3 x 10~3
8.6 x 10~*
8.6 x 10-3
2.0 x 10"""
2.0 x ID"5
1.4 x 10~3
1.4 x 10~*
Oligotrophic Lake
Mass Cone.
(ks) (u« K-1)
4.2
2.1
2.4
1.2
2.0
1.0
2.7
1.4
5.8
9.0
x 10
x 10'2
x 102
x 10" l
x 10
x 10"a
x 10
x 10"2
x 10" l
1.7
1.7
9.7
9.7
8.0
8.0
1.1
1.1
6.7
6.7
x 10~l
x 10~2
x 10"1
x 10-2
x 10" *
x 10~2
x 10" l
x 10~2
x 10" *
The amounts given for solid and solution phases in the sediment compartments are estimated from the sorption
partition coefficient for suspended solids and may be overestimated because it was assumed that biodegradation
of sorbed material does not occur.
-------
2 x KT1
100 150
TIME - hours
200
250
FIGURE 7.1 PERSISTENCE OF QUINOLINE IN A TWO-COMPARTMENT POND SYSTEM
100
-------
3 x 10'
SEDIMENTS
10"°
456
TIME - hours
FIGURE 7.2 PERSISTENCE OF QUINOLINE IN A PARTIALLY MIXED RIVER SYSTEM
101
-------
3 x 10'2
10
t-2
2 x 10'
11
100 200
300 400
TIME - hours
500 600
700 720
FIGURE 7.3 PERSISTENCE OF QUINOLINE IN A PARTIALLY
MIXED EUTROPHIC LAKE
102
-------
3 x 10-
0>
a.
Ul
Z
_l
o
Z
o
u_
O
Z
o
Z
Ul
o
o
o
100
200
300
400
500
600
700 720
TIME - hours
FIGURE 7.4 PERSISTENCE OF QUINOLINE IN A PARTIALLY MIXED OLIGOTROPHIC LAKE
Degradation in all but the most oligotrophic waters will probably be
dominated by biodegradation; however, the effective bacteria populations
assumed in the oligotrophic lake simulations are very low, and merely doubling
or tripling bacterial population in the oligotrophic lake simulation would put
biodegradation and photolysis on an equal level. Moreover, since cometabolism
is omitted from all the simulations, our predictions of biodegradation rates
may be much too low. If so, biodegradation should be expected to be dominant
in all natural waters except in the coldest months.
During winter months in the Northeast, the Midwest, and the mountain
states, volatilization and photolysis rates may exceed biodegradation. Indeed,
103
-------
it is probably that volatilization will be dominant as long as the water sur-
faces are ice-free.
In summary, we can expect degradation to be rapid in all freshwaters,
with half-lives ranging from 5 hours or less in warm eutrophic waters to
about 25 days in oligotrophic waters. Moreover, unless cometabolism occurs
and is significantly affected by concentration at concentrations of 1 pg ml"1
or less, the patterns of degradation and distribution estimated by our simu-
lations should be generally applicable, as indicated by the similarity of
results of simulations using the concentrations of 1 ug ml"1 and 10 ng ml"1
for quinoline in the inflows. Degradation should be slowest in the winter,
particularly in ice-covered waters where the 25-day half-life may be greatly
exceeded due to suppression of photolysis and volatilization. At worst, under
thick ice cover, quinoline may not degrade. Nonetheless, as long as the con-
centrations of quinoline in the inflowing waters are less than 1 yg ml"1,
the hazard should be low.
7.4 PHYSICAL PROPERTIES
7.4.1 Solubility in Water
The solubility of quinoline in water at room temperature (25°C) is
reported in the literature as 6.11 mg ml"1 (Albersmeyer, 1958).
7.4.2 Absorption Spectrum
A small sample of quinoline was purified by preparative gas chroma-
tography. A portion of this purified material was used to prepare a
1.5 x 10~4 M solution of quinoline in distilled deionized water. The pH of
this solution was adjusted to pH 4.7 and 6.8 with dilute HC1 or NaOH, and
the UV absorption spectra were measured. The absorption coefficients at
wavelength intervals from 297.5 to 370 nm are reported in Table 7.5. The
pKa of quinoline is 4.8 (Perrin, 1965). Therefore, quinoline is 98%, 39%,
and 0.6% protonated at pH 3, 5, and 7, respectively. The inciease in the
absorption coefficient with the decrease in pH is undoubtedly a result of the
increase in the fraction of protonated quinoline.
-f-
OIO
7.4.3 Volatilization Rate
The volatilization rate of quinoline from water was measured using the
method of Hill et al. (1976), which is described in detail in Part I, Section
5.3.3 of this report.
104
-------
TABLE 7.5. ABSORPTION SPECTRA OF QUINOLINE IN WATER"
at pH 4.7 and pH 6.8
Center of
wavelength
interval^
(nra)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330.0
340.0
350.0
360.0
370.0
Average absorption
(M"1 cm
pH 4.7
4320
4660
4850
5070
4980
5220
5720
4880
3820
2920
2270
1370
529
143
9.3
0
coefficient0
-1)
pH 6.8
2910
3050
2740
2480
2050
2440
2920
1680
622
269
119
26
9
0
0
0
Quinoline concentration 1.61 x 10~A mole liter *.
The wavelength intervals are given in Appendix B,
Table B.I.
CThe absorption coefficients at 313.0 and 366.0 nm are 5690
and 0 at pH 4.7.
105
-------
The volatilization half-life of quinoline in aqueous solution was
about 69 hours at a moderate stirring rate. The rate constants were calcu-
lated as follows:
Oxygen reaeration rate k = 4.01 ± 0.22 hr"1
constant:
Quinoline volatilization k^ = 0.011 ± 0.001 hr"1
rate constant:
Ratio: kQ/k° = 0.0027 ± 0.0004
v v
Additional measurements were not made because biodegradation was found, in
separate experiments, to be so rapid (see Section 7.6.2).
7.4.4 Sorption on Clay and Sediments
Sorption isotherms were measured on Ca-montmorillonite clay and
Coyote Creek sediment. Each experiment consisted of duplicate samples at two
quinoline concentrations and two sediment loadings, in addition to blanks
containing quinoline only and sediment only. All samples were adjusted to
pH 6.0 ± 0.2 with dilute HC1. The data, summarized in Table 7.6, were fitted
to a Freundlich isotherm where n = 1 and in which either the regression line
was forced through the origin (a = 0) or a non-zero intercept was calculated.
Calculation of the partition coefficients was based on analysis of the
supernatant at equilibrium, and the initial amount present, as determined by
analysis of samples containing no sediment. The estimated values of Kp
reported in Table 7.6 indicate that sediment sorption is not an important
pathway for the disappearance of quinoline; thus, no further sorption studies
were conducted.
7.4.5 Biosorption
Since biodegradation of quinoline was so rapid and sediment sorption
so weak, no biosorption studies were conducted.
7.5 CHEMICAL TRANSFORMATION
7.5.1 Photolysis Rate
Data for the photolysis of quinoline in pure water and in water con-
taining 9.5 yg ml"1 humic acid are given in Table 7.7. The photolyses at
313 nm were carried out to 50% to 70% conversion of quinoline with good first-
order kinetic behavior found. The rate constant for direct photolysis of
quinoline at pH 6.9 was unaffected by the presence of oxygen. From the rate
constant kp measured for the direct photolysis, the quantum yield for quino-
line conversion was calculated as 3.3 x 10"''.
106
-------
TABLE 7.6. QUINOLINE SORPTION ON SEDIMENTS
O
-J
Total
organic
carbon
Sediment (percent)
Ca-
montmorillonite 0.06
clay
Coyote
Creek 1.4
Quinoline
Sediment concentra-
concentra- tion in
tion
(mg ml""1)
39
80
39
78
38
77
38
77
Supernatant
i (UR ml'1)3
4.3
3.7
9.0
7.0
4.0
2.9
8.5
6.2
± 0.20
± 0.10
± 0.40
± 0.60
± 0.2
± 0.1
± 0.2
± 0.1
Quinoline „ ... . ,.,.. . ,,
concentra- Partition coefficients, Kp
tion on LLS
Sediment „ , .
( U 2 £~ 1 O O
37
26
62
57
44
36
87
74
±5 7.3 ± 0.5 5.8 ± 15
±1 (a = 11 ± 10)
± 10
± 7
± 5 10.9 ± 0.4 9.7 ± 1.1
±2 (a = 7.7 ± 6.1)
± 5
± 2
Q
Concentration measured in supernatant with population standard deviation.
Concentration on sediment calculated from supernatant concentration WJth population standard deviation.
Calculated by linear least squares statistical treatment, as described in Appendix B, Section B.I.4.
Limits are 95% confidence limit.
-------
Since quinoline is a weak base (pKj, = 9.2) and will be mostly in the
conjugate acid form at low pH, an experiment on the direct photolysis at pH
4.4 was carried out. At this pH, quinoline is 71% protonated; less than 1%
is protonated at pH 6.9.
Figure 7.5 plots the log of quinoline concentration at pH 6.9 and pH 4.4 as
a function of time. Although the pH 6.9 data show good first-order kinetic
behavior, the pH 4.4 data indicate that the reaction is progressively inhibited
as the photolysis proceeds. The pH 4.4 data are not correlated by any simple
integral or half integral kinetic rate law for quinoline conversion. The
reason for this behavior is not known. The effect of pH may be to produce
two different mechanisms for photolysis. It is unlikely that the products
from the reaction affect the photolysis at pH 4.4, since their concentrations
are less than 1 yg ml"1. The total absorbance of the solution is so low that
it does not affect the total light intensity passing through the solution.
Any product that acts as a triplet quencher could not compete with the oxygen
(about 10~3 M) in the air-saturated solution. There is insufficient data for
determining the processes that are taking place.
The data in Table 7.7 show that in the photolyses at 313 run, the pre-
sence of humic acid reduced the photolysis rate of quinoline by 56%. The
absorbance of the humic acid solution was 0.56 (compared with 0.01 calculated
for 0.7 ug ml"1 quinoline) so that the humic material absorbed about half of
the incident light. The effect of humic acid in slowing the photolysis of
quinoline can then be mostly accounted for by a screening effect. As discussed
previously, the slower rate may also be due partially to the low pH of the
solution (5.25); at this pH quinoline would be 20% in the protonated form. The
photolyses in humic acid solutions, however, showed no deviation from first-
order kinetics for disappearance for quinoline.
Evidence for the absence of any photoinitiated radical reaction or photo-
sensitized processes involving humic acid was provided by the photolysis at
366 nm. At this wavelength, quinoline has no absorption, whereas humic acid
absorbs appreciably. Correcting for the difference in light intensity and the
humic acid absorbance at 313 nm versus 366 nm, we calculated from the experi-
ments at 366 nm that any sensitized photolysis with 9.5 )Jg ml"1 humic acid
must have a rate constant less than 10 times that of the direct photolysis.
This calculation assumed that the quantum efficiency for any humic acid sensi-
tized photolysis was independent of wavelength.
The half-life for direct photolysis of quinoline in sunlight was calcu-
lated according to the procedure of Zepp and Cline (1977) using the quantum
yield of 3.3 x lO"4* and the absorption spectrum (Section 7.4.2). As shown in
Figure 7.6, the half-life for direct photolysis of quinoline will vary from
160 days in winter to about 23 days in summer. The latter value is in excel-
lent agreement with the measured half-life of 21 days during June (Table 7.7).
108
-------
o
VO
tO
o
X
LU 3
I]
O
a
pH 4.4 SOLUTION
pH 6.9 SOLUTION
I
I
10 20 30 40 50 60 70 80
TIME — hours
90
100 110
120
SA-4396-48
FIGURE 7.5 DIRECT PHOTOLYSIS OF QUINOLINE AT pH 6.9 AND pH 4.4
-------
TABLE 7.7. RATE CONSTANTS FOR PHOTOLYSIS OF QUINOLINE
Irradiation
source
313 nm
313 nm
313 nm
Sunlight ,
late June
Sunlight,
late June
Quinoline
concentration Rate constant
Solution (ug ml'1) (k x 106 sec"1)
Pure water, air
saturated, pH 6.9
Pure water, nitrogen
sparged, pH 6.9
9.5 pg ml"1 humic acid
in pure water,0 pH 5.4
Pure water, air
saturated, pH 6.9
9.5 pg ml"1 humic acid
in pure water, pH 5.4
0.82
0.69
0.69
0.67
0.68
0.82
0.67
6.07 ±
5.85 ±
5.71 ±
2.60 ±
2.52 ±
0.770 ±
3.04 ±
o.oob
0.14
0.36
0.18
0.21
0.060d'e
o.iod>f
1.00 ug ml"1 quinoline in water = 7.74 x 10~6 M.
Standard deviation.
"Absorbance of humic acid solution at 313 nm is 0.56 (1-cm cells).
Calculated assuming 12 hours of sunlight per day in late June; to obtain
average rate constant for a fall calendar day (24 hours), divide rate
constant by two.
3
"Half-life of 21 calendar days.
Half-life of 5.3 calendar days.
The data of Table 7.7 for the solar photolysis of quinoline in the 9.5
yg ml"1 humic acid solution indicate a half-life of about 5.3 days, about four
times faster than for the direct photolysis of quinoline at the same time of
year. This result is contrary to what was expected based on the laboratory
photolyses at 313 and 366 nm, which indicated that humic acid acted only as a
light screen and gave no evidence of a photoinitiated or photosensitized
reaction. We have no explanation for the difference in results for the labora-
tory and the field experiments; the problem deserves further investigation.
7.5.2 Oxidation Rate
The susceptibility of 0.83 pg ml"1 (6.4 x 10~6 M) auinoline to free
radical oxidation was examined using the AA-initiated oxidation reaction. See
110
-------
170
160
140
120
| 100
I
1U
LL
T" 80
LL
I
60
40
20
I \ I
1 I
I I I I I
JAN FEB MAR
APR MAY JUN JUL AUG SEP OCT NOV
MONTH OF YEAR
DEC
FIGURE 7.6 ANNUAL VARIATION OF PHOTOLYSIS HALF-LIFE OF QUINOLINE
-------
Appendix B, Section B.2 for discussion and procedure. This experiment was
carried out at 50°C in pure water containing 1.0 x ICT^ M AA. At the end of
100 hours reaction time, only 3% of the quinoline had been consumed. A control
solution without AA under identical reaction conditions showed no change
(± 1% in the quinoline concentration.
A pseudo-first-order oxidation rate constant (kox [R02*]) of 8.5 x 10~8
sec"1 was obtained for oxidation of quinoline under the reaction conditions.
For the reaction at 50°C, this corresponds to a second-order rate constant
k0x of 25 M"1 sec"1 (Part I, Section 5.4.2). At 25°C, kox would then be
2.8 M~x sec"1, and for a RC^' radical concentration in the aqueous environment
of 10~9 M. the half-life ot quinoliae toward oxidation is about 8 years. The
free radical oxidation of quinoline by R02* is clearly not competitive with
biodegradation or even photolysis i..ider aquatic environmental conditions.
7.5.3 Hydrolysis Rate
Since quinoline contains no groups that are hydrolyzable, no hydrolysis
studies were carried out.
7.5.4. Products from Chemical Transformation
Since screening studies determined that photolysis and oxidation were
not competitive with biodegradation, detailed kinetic studies and product
studies were not performed for quinoline.
7.6 BIODEGRADATION
7.6.1 Development of Biodegrading Cultures
Screening for quinoline-biodegrading cultures developed by an enrichment
process was initiated with quinoline levels of 10 yg ml"1. The 9-liter aerated
fermentors were used for water samples from Searsville Pond near Woodside,
California; water from Coyote Creek in San Jose, California; aeration effluent
from the Palo Alto, California, sewage plant; and water from Lake Tahoe.
Fernbach flasks with 1.5 liters of 4:5 diluted aeration effluents were used
with samples of aeration effluents from the South San Francisco and Shell Oil
Refinery (Martinez, California) wastewater treatment plants. The Fernbach
flasks were placed in a Psycrotherm Shaker with rotary action.
All incubations were at 25°C except the Lake Tahoe samples, which x^ere
incubated at 15°C. Progress in the development of quinoline biodegradation
was monitored by GC analyses and by UV absorptions of ethyl acetate extracts
at 308, 312, and 320 nm.
Quinoline biodegrading systems were obtained from all waters. The gen-
eral pattern of quinoline disappearance was similar to the"adaptation period
indicated in Figure 7.7, which also includes cascade series breakdowns of
quinoline. Degradation, once initiated, was rapid. Greater than 95.^ break-
112
-------
9-liter BOTTLE
SERIES 1 SERIES 2 SERIES 3 SERIES 4
101
H
U)
FIGURE 7.7 CASCADE BATCH FERMENTATIONS WITH A FRESHLY DEVELOPED DEGRADING
SYSTEM TRANSFERRED IN FRESH POND WATER MEDIA
-------
down of quinoline occurred with samples from Coyote Creek, Palo Alto Sewage
Plant aeration effluent, Searsville Pond and Lake Tahoe in 48 hours, 40 hours,
60 hours, and 11 days, respectively. It is of interest that with the adapta-
tion in Lake Tahoe water, the time for greater than 95% breakdown of 10 j.:g ml"1
quinoline after start of degradation was 8 hours. The samples of aeration
effluent waters from the South San Francisco and Shell Oil Refinery treatment
plants contained substances that interfered with UV monitoring for quinoline
and, arbitrarily, at 48 hours, a 2% transfer was made to flasks containing
basal salts/10 yg ml"1 quinoline medium. Degradation of quinoline in these
flasks was observed in 24 hours.
In subtransferring from the original water sample mixtures in 9-liter
bottles to 50 ml of basal salts media in 250-ml Erlenmeyer flasks, no prob-
lems were encountered in using quinoline as the only carbon source or in
stepping up quinoline to 30 and 100 yg ml"1 levels. A pink color developed
in some of the degrading systems.
7.6.2 Biodegradation Kinetics
Two types of kinetic studies were conducted: batch fermentations with
low-level inocula and cascade batch fermentations with a freshly developed
degrading system serially transferred in fresh pond water media.
Batch Fermentations with Low-Level Inocula—Figures 7.8 and 7.9 present
data from batch fermentations with low-level inocula of quinoline decomposing
organisms developed from a eutrophic pond. Figure 7.8 plots viable cell
counts in flasks witii initial quinoline levels of 0, 1, 3, 5, and 10 yg ml"1
respectively. The increase in counts in the control flask with no quinoline
is attributable to the growth of some organisms or secretion products of
other viable cells and/or autolysis products of dead cells. It is also
possible that trace quantities of metabolizable organic compounds were intro-
duced into the media by the distilled water or air during shaker flask agita-
tion (Kayser, 1975). Because quinoline is so readily metabolized, increases
in cell counts in flasks with quinoline are most likely due to good survival
and growth of quinoline utilizers (primary and secondary).
The specific growth rate (u) in each medium was determined by measuring
the slope of the curves in the logarithmic phase of growth. For the media
initially containing 1, 3, 5, and 10 yg ml"1 quinoline, y values were 0.65,
0.69, 0.73, and 0.73 hr~l, respectively. Comparing counts and residual
quinoline data in Figures 7.8 and 7.9, it is apparent that significant in-
creases in cell counts occurred after quinoline was essentially all degraded.
Figure 7.10 is a Lineweaver-Burk plot developed by the least squares
method. Ks and ym were calculated as 0.16 ppm and 0.74 hour"1, respectively.
Table 7.8 presents the data and calculated values for Y and kt obtained
in one low-inoculum batch fermentation study with one biodegrading system.
The average yield (Y) was 1.9(±1.0) x 10s cells yg quinoline"1 and the first-
order rate constant in X is 4.9(±2.5) x 10~7 yg quinoline cell"1 hr"1. The
second-order rate constant k^2> first order in X and S, is k^/Kg = 3.1 x 10~6
ml cell"1 hr"1.
114
-------
8 12
TIME — hour
16
20
SA-4396-52
FIGURE 7.8
VIABLE COUNTS IN QUINOLINE DEGRADATION BY BATCH
FERMENTATIONS WITH LOW-LEVEL INOCULA
Counts for initial concentrations: 0 (® ); 1 /ig ml"1 (V);
10x counts for 3 pig ml'1 ( D); 102x counts for 5 pig ml'1 (A);
103X counts for 10 pig ml'1 quinoline (O).
115
-------
100
CD
a.
UJ
a
o
LT
8
TIME — hour
12
16
SA-4396-53
FIGURE 7.9 QUINOLINE BIODEGRADATION BY BATCH FERMENTATIONS WITH LOW-LEVEL
INOCULA
Initial quinoline concentrations = 1 (®), 3 (D ), 5 (A ) and 10 ( O ) /ng ml"1
-------
15
10
5
So
10
SA-4396-54
FIGURE 7.10 LINEWEAVER-BURK PLOT FOR DATA FROM QUINOLINE
BIODEGRADATION IN BATCH FERMENTATIONS WITH
LOW-LEVEL INOCULA
117
-------
TABLE 7.8. QUINOLINE BIODEGRADATIONS IN BATCH FERMENTATIONS WITH LOW-LEVEL INOCULA
oo
Initial
quinoline
cone.
(UK ml'1)
1
3
5
10
Time
(hr)
12
14
14
14
Change in
bacterial counts (AX)
(cells ml"1 x 106)
1.5
5.5
6.7
6.1
Quinoline
degraded (AS)
-------
The decreasing value of Y with increasing initial concentrations of
quinoline may reflect an increasing accumulation of a metabolite that has an
inhibitory effect.
Cascade Batch Fermentations with a Freshly Developed Degrading System
Transferred in Fresh Pond Water Media — The data obtained for quinoline levels
and viable cell counts in a cascade study with water from a eutrophic pond are
presented in Figures 7.7 and 7.11 and in Table 7.9. Because the samplings
were limited, only the data from the 10 yg ml"1 quinoline series were analyzed
for determination of the yield factor and the first-order rate constant. The
initial viable counts (X ) after 0.2% inoculation from the previous degrada-
tion culture mixture ranged from 0.6 x 10s to 2.2 x 105 cells ml"1. These
counts were at least two orders magnitude smaller than the cell counts used
in calculations of Y, and they were neglected in the change in population
used to calculate this factor. The values of ^ were determined from the
slopes of the slots of In ;AS versus t in Figure 7.11 (see Section 7.3, Part
I of this report for the details of this kinetic analysis). Since Y = AS /Ax,
k^ could be calculated (Table 7.9). Data from series 2 in the cascade studies
were not used in these calculations because of insufficient data points in
the critical time periods. When the Kg value of 0.16 yg ml"1 from batch fer-
mentations with low-level inocula is used, the second-order rate constant
, is (2.4 ± 1.5) x 10~6 ml cell"1 hr~l.
7.6.3 Identification of Metabolites
Quinoline metabolites were observed during the kinetic studies. Two
intermediates were found after they were converted to the corresponding tri-
methylsilyl derivatives and examined by gas chromatography/mass spectrometry.
The major metabolite was identified as 2-hydroxyquinoline (I) based on compara-
tive spectra of an authentic standard (supplied by Pf altz and Bauer company) .
The mass spectrum of the second metabolite corresponded to the di-
trimethylsilylether of a dihydroxylated quinoline [M/e = 305 (M+)]. This
metabolite is expected to be 2,3-dihydroxyquinoline (D). However, this
structure has not been verified because an authentic standard is not avail-
able.
119
-------
20
10
I 8
6
O)
a.
Q 4
UJ
Q
01
Q
O
1
0.8
0.6
0
SERIES 4
I
I
SERIES 1
I
8 12
TIME — hour
16
20
SA-4396-55
FIGURE 7.11 QUINOUNE UTILIZATION IN CASCADE BATCH FERMENTATIONS
120
-------
TABLE 7.9. QUINOLINE DEGRADATIONS IN CASCADE BATCH FERMENTATIONS'
Flask
aeries
1
2
4
Time (t)
(hr)
16
14
14
Cell counts (X)
(cells ml"1 x 106)
4.6
4.3
14.0
Quinoline
degraded (AS)
(lag ml""1)
4.0
9.9
10.0
Cell yield (Y) •-,- ym
(cells ug"1 S x 106) (hr-1)
1.15 0.301
0.43 0.289
1.40 0.289
Average 1.0 ± 0.5
First-order
rate constant (kjj]_)
(yg cell"1 hr"1 x 10" 7)
2.6
6.7
2.1
Average 3.8 ± 2.5b
Freshly developed quinoline biodegrading system sequentially transferred (0.2% inoculum) into fresh
pond water sample media.
Standard deviations.
-------
8. LABORATORY INVESTIGATION OF BENZO[f]QUINOLINE
8.1 SYNOPSIS
The results of the laboratory investigations suggest that photolysis is the
major transformation pathway for benzo[fJquinoline (BQ) in all types of water
bodies. Biodegradation may occur slowly in eutrophic waters. Sorption by sedi-
ments may be significant in turbid waters.
The nine-compartment model predicted the following steady-state concentra-
tions of BQ in solution, suspended solids, and sediments near point sources in
the presence of a continuous discharge of 10 ng ml"-'- (10 ppb) BQ:
Suspended
Half-life Solution solids Sediments
(hr) (yg ml"1) (pg g"1) (yg g"1)
River 0.5 8.1 x 10"1 1.1 x 103 1.0 x 103
Pond 6.9 7.8 x 10~3 1.1 x 101 1.1 x 101
Eutrophic lake 7.0 2.4 x 10~2 3.3 x 101 8.3
Oligotrophic lake 1.4 4.3 x 10~3 6.0 1.5
8.2 BACKGROUND
Benzoquiuolines have been identified in various crude oils (Brodeskii et al.,
1975; Snyder, 1969; Ball, 1962) and are produced in the low-temperature carbon-
ization of coal (Karr and Ta-Chung, 1958).
Brodeskii et al. (1975) reported that the acridines and benzoquinolines
constituted 12% to 20% of the basic components in Sakhalim petroleums; the con-
centrations of the quinolines were 50% to 60% of these same fractions. Snyder
(1969) observed that in a California crude oil the total benzoquinoline content
in two distillate fractions was approximately one-sixth that of the quinolines.
It is apparent that the quinolines and benzoquinolines are usually found to-
gether and should be present in both petroleum and coal processing wastes.
Benzoquinolines have been found in airborne particulates in urban atmospheres
(Sawicki et al., 1965; Brocco et al., 1973; Ray and Frei, 1972). Benzo[f]-
quinoline (BQ) has been selected for this study; its physical properties are
listed in Table 8.1.
A
Predicted by the one-compartment model.
122
-------
TABLE 8.1. PHYSICAL PROPERTIES OF BENZOfflOUINOLINE
Structure
Molecular weight
Melting point (°C)
Boiling point (°C)
(Lange's Handbook of
Chemistry, 1973)
Vapor pressure at 20°C (torr)
(McEacherti et al., 1975)
I1 1 a
Solubility in water (yg ml x)
UV spectrum (ethanol)
X nm (log e)
Handbook)
350 (721 torr)
202-5 (8 torr)
,-5
2.8 x 10
76.1 ± 2.2
316 (3.34)
323 (3.26)
330 (3.62)
338 (3.28)
346 (3.71)
5.15
(Ray and Frei, 1972)
1 yg ml"1 BQ = 5.58 x 10~6 M
Measured in this study (Section 8.4.1).
""pIC = 8.85; pIC for quinoline is 9.5.
No information was found concerning the water solubility, volatilization,
or sorption properties of BQ on soils or sediments. Doehler and Young (1960)
studied the sorption of quinoline on Na and Ca-montmorillonite. They reported
that the partition coefficient is pH dependent; sorption increases as the pH
decreases.
BQ should not hydrolyze under the conditions of environmental aquatic
systems.
Data were not available 'for evaluating the importance of free radical
oxidation of BQ under environmental conditions. In the study of the free radi-
cal oxidation of quinoline, an environmental half-life of at least 100 years
was estimated.
As seen from the UV spectral data (see Table 8.1), BQ absorbs strongly in
the solar region out to 350 nm. Although no information on the photolysis of
123
-------
BQ was available, our studies on quinoline indicated a half-life of about 21
days for direct photolysis; in the presence of humic acid, the photolysis half-
life of quinoline was about 6 days (See Section 7.5.1).
No reports were found regarding the biodegradation of the benzoquinolines.
In Section 7, we reported that quinoline biodegrading systems could be readily
obtained by enrichment processes with oligotrophic and eutrophic waters and with
aeration effluents from sewage plants. BQ also has structural similarities with
phenanthrene, which can be biodegraded.
In summary, the literature information and our data from quinoline suggested
that photolysis and biodegradation might be important processes in the re-
moval of BQ from natural waters.
8.3 ENVIRONMENTAL ASSESSMENT
8.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigations of the
transformation and transport processes for BQ are summarized in Table 8.2.
TABLE 8.2. SUMMARY OF BENZO[f]QUINOLINE LABORATORY DATA
Process
Sorption equilibrium
Partition coefficient
Sorption
Photolysis
Oxidation
Hydrolysis
Biodegradation
S = K S
s p w
Rate expression
K = 1300
P
Rate constant at 25°C
Volatilization k [BQ]
k [BQ] « $ (ZZ.e.)[BQ]
p A A
k [RO -][BQ]
OX *•
NA
kv = (2.2 ± 0.9) x 10~4 k°
k = (1.4 ± 0.7) x 10~4 sec"1
P
k <2.8 M"1 sec"1
ox
NA
k. 0 = (3.6 ± 0.01) x 10 8 ml
*2 cell-1 hr-
Measured on Coyote Creek sediment.
3See Part I, Section 5.3.
•>
"Assumes 12 hours of sunlight in mid-June.
8.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved BQ calculated for individual transformation
or removal processes following a single, brief input are listed in Table 8.3.
124
-------
Photolysis is the dominant pathway in all standing waters, but is roughly five
times slower than dilution under the assumed conditions of rapid river flow.
Sorption is important only in the most turbid waters, but nven in these waters
it is unlikely to account for more than 40% of the BQ lost from solution. In
all waters, biodegradation is slow, and volatilization and oxidation are ex-
ceedingly slow. The half-lives in eutrophic and oligotrophic waters differ by
roughly a factor of 4.
TABLE 8.3. TRANSFORMATION AND TRANSPORT OF
BENZO[f]QUINOLINE PREDICTED BY THE ONE-COMPARTMENT MODEL
o
Photolysis, half-life (hr)
Oxidation, half-life (hr)
Volatilization, half-life (hr)
Hydrolysis, half-life (hr)
Biodegradation, half-life (hr)
River
2.8
>io5
>io4
—
190
Eutrophic
pond
7.0
>io5
>io5
—
190
Eutrophic
lake
7.0
>io5
>io5
—
190
Oligotrophic
lake
1.4
>io5
>io5
—
>io6
Half-life for all processes
except dilution (hr) 2.8 7.0 7.0 1.4
Half-life for all processes
including dilution (hr) 0.48 6.9 7.0 1.4
Amount BQ sorbed (rag m~3)b 140 720 70 70
Percentage BQ sorbed 12% 42% 6.5% 6.5%
r*
Estimates are the average photolysis rates on a summer day at 40° latitude.
Photolysis rates in midwinter are at least three times slower.
Assumes 1 yg ml"1 of BQ in aqueous phase and partition coefficient of 200.
8.3.3 Persistence
Benzoffjquinoline should not be "persistent" in any shallow or thoroughly
mixed natural waters. However, in lakes such as Lake Superior and Lake Tahoe,
which are so deep as to be unproductive and poorly lit throughout much of their
depth, benzo[f]quinoline may persist at relatively high concentrations for
decades or centuries. The persistence .of concentrations that approximate the
initial concentrations following an acute discharge should be virtually inde-
pendent of dilution except in all but the largest streams, such as the hypo-
thetical river assumed in the river simulation of Table 8.3.
125
-------
8.3.4 Mass and Concentration Distributions Calculated Using Computer Models
The distributions of mass and concentration of BQ expected at steady
state during chronic discharge to each of four types of water bodies are shown
in Table 8.4. The changes in concentration as discharge begins and ends are
shown in Figures 8.1 through 8.4. The pseudo-first-order rate constants used in
the simulations are presented in Appendix A. Under our assumptions regarding
the geometries of the water bodies and the flow rates within them, roughly 95%
of the BQ is in the sediments at steady state and roughly 90% of the remainder
is in solution.
The concentrations of dissolved BQ at steady state varied 100-fold among
the four simulated water bodies, the highest concentrations being in the eutro-
phic waters where photolysis is slower than in the oligotrophic lake. Con-
centrations within individual waters varied 10,000-fold among aqueous compart-
ments of the oligotrophic lake, 1,000-fold within the eutrophic lake, and about
15% in the river, where dilution is very rapid (Table 8.4). Average concentra-
tions in the sediments differ by roughly a factor of 10 or less from the
concentrations in the aqueous compartment that receives the inflowing waters
(Compartment 1 in each simulation). Concentrations of BQ on solids in the water
columns and in the sediments tend to be 1,000 times higher than concentrations
in solution.
The approaches to steady-state concentrations were rapid in all water
bodies following start of the simulated discharge, and the transitions to new
steady states after discharge stopped were very rapid in the surface waters of
the lake simulations (Figures 8.3 and 8.4). Concentrations of BQ in the sedi-
ments declined slowly in each simulation, reflecting the large mass of BQ in the
sediments and the relatively slow rates of exchange between the sediments and
the water column.
8.4 PHYSICAL PROPERTIES
8.4.1 Solubility in Water
The solubility of BQ in water at 25°C, measured using the method of
Campbell (1930), was 76.1 ± 2.2 yg ml"1.
8.4.2 Absorption ^Spectra
The absorption spectra of BQ in water were measured at pH 7.1 (adjusted
and 6.3 (nonadjusted). The absorption coefficients at wavelength intervals
from 297 to 400 nm and at 313 and 366 nm are given in Table 8.5. The absorp-
tion coefficients at pH 6.3 were slightly higher than at pH 7.1, probably due
to the higher concentration of protonated BQ. (BQ is 1.1% protonated at pH 7.1
and 7.1% at pH 6.3.)
126
-------
TABLE 8.4. DISTRIBUTION OF BENZO[f]QUINOLINE IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations of 1 wg ml"1 benzo[f]quinollne)
Pond River
Compartment 1
(surface water)
Solution
Suspended solids
Compartment 2
(surface water)
Solution
Suspended solids
Compartment 3
(surface water)
Solution
Suspended solids
Compartment 5
(bottom water)
Solution
Suspended solids
Compartments 7-9
(sediment)3
Solution
Solids
Mass Cone. Mass
(kp,) (HE ml"1) (kgl
1.55 x ID'1 7.75 x 10"3 2.43 x 10a
6.50 x 10"a 1.08 x 10 3.40 x 10
2.25 x 10a
3.15 x 10
2.08 x 10*
2.92 x 10
—
—
1.94 x 10~3 7.75 x 10""3 5.63
7.29 1.08 x 10 2.13 x 10*
Cone,
(ug ml~11
8.10 x
1.13 x
7.50 x
1.05 x
6.95 x
9.73 x
__
—
7.51 x
1.05 x
10" l
10s
10- l
103
10- l
10a
10" '
10s
Eutrophic Lake
Mass
5.93
4.15
2.28
1.59
6.56
4.59
i
5.40
3.78
5.20
1.13
,
x 10" l
x 10" l
x 10- l
x 10- a
x 10- a
x 10" 3
x 10" '
x 10'
Cone.
CUg ml"1)
__..
2.37
3.32
9.13
1.27
2.62
3.67
2.16
3.02
5.94
8.32
x 10" s
x 10
x 10"*
x 10-*
x 10" '
x 10" 3
x 10" a
x 10" 3
Oligotrophic Lake
Mass
(kg)
1.07
7.53
6.54
4.58
3.76
2.63
4.15
2.92
8.87
2.03
x 10- a
x 10" a
x 10- 3
x ID'*
x 10" 3
x 10"*
x 10" s
x 10"*
x 10
Cone
(ug ml'
4.30 x
6.02
2.61 x
3.66 x
1.50 x
2.10 x
1.66 x
2.32 x
1.06 x
1.49
ll)
io-3
10"'
10" a
10" *
10" 3
10- 7
10"*
10"'
aThe amounts given for solid and solution phases in the sediment compartments are estimated from the sorption partition coefficient for
suspended solids and may be overestimated because it was assumed that biodegradation of sorbed material does not occur.
-------
2 x 10
10
E
o>
UJ
Z
o
z
a
****'
'o'
N
01
m
Z
2 10
o
o
o
r- /
I /
-2
10-
4 x 10'1
SUSPENDED SOLIDS
50
100 150
TIME - hours
200
250
FIGURE 8.1. PERSISTENCE OF BENZO[f]QUINOLINE IN A TWO-COMPARTMENT POND SYSTEM
128
-------
10
I
z
o
z
o
O
Z
10-
I I I I [ l_ I l_
COMPARTMENT ^^***^^'"~
^ ^ ** l
4
SEDIMENTS
SOLUTION
I I
I I
_L L
TIME - hours
9 10
FIGURE a 2 PERSISTENCE OF BENZOffjQUINOLINE IN A PARTIALLY. MIXED RIVER SYSTEM
129
-------
SEDIMENTS
SOLUTION
10
100
200
300 400
TIME - hours
500
600
700 720
FIGURE 8.3. PERSISTENCE OF BENZO[f]QUINOLINE IN A EUTROPHIC LAKE
130
-------
4 x i(
n
10
-i
OJ
I
I'0"
O
ui
co
LL.
O
ui
O
|io-7
10"
10
-9
3 x 10
,-10
COMPARTMENT
3
T
r
_ 2
DISCHARGE !
STOPPED
SEDIMENTS
SOLUTION
100
200
300 400
TIME - hours
2 =
5
500
600 700 720
FIGURE 8.4. PERSISTENCE OF BENZO[f]QUINOLINE IN AN OLIGOTROPHIC LAKE
131
-------
TABLE 8.5. ABSORPTION SPECTRA OF B_ENZO[f JQUINOLINE IN WATER
Center of
wavelength
intervalb
(nm)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330
340
350
360
370
380
390
400
Absorption coefficients0
(M-! cm'1)
pH 7.1
3980
3910
2070
1400
1140
1070
1330
1600
1410
1340
1460
3010
1600
1380
89
44
30
0
0
pH 6.3
3960
3910
2140
1500
1240
1180
1430
1670
1480
1380
1490
3020
1680
1530
250
185
96
37
0
Benzoff]quinoline concentration was 1.35 x 10~4 M in a 1-cm cell.
The wavelength intervals are given in Appendix B, Table B.I.
CThe absorption coefficients at 313 and 366 nm are 1361 and 37 at
pH 7.1 and 1479 and 207 at pH 6.3.
8.4.3 Volatilization Rate
The rate of volatilization of BQ from water was measured using the method
of Hill et al. (1976). At an ayerage oxygen reaeration rate k^ = 4.00 ± 1.51 hr"1,
the ratio k$°>/k$ was (2.2 ± 0.9) x 10~4. Since this ratio is very small, vola-
tilization was not an important pathway for BQ compared with photolysis;
therefore, no further volatilization experiments were conducted.
8.4.4 Sorption on a Sediment
The sorption partition coefficient of BQ on Coyote Creek sediment was
measured. Replicate samples were run at two BQ concentrations and two sediment
loadings as well as blanks. All solutions were buffered to maintain pH 7.
Table 8.6 summarizes the results.
132
-------
CO
CO
TABLE 8.6. SORPTION OF BENZOff]QUINOLINE
ON COYOTE CREEK SEDIMENTa
Sediment
concentration
(mg ml"1)
4.8
14.5
4.8
14.5
BQ concentration
in supernatant3''3
(yg ml"1)
0.99 ± 0.19
0.61 ± 0.02
1.35 ± 0.17
0.84 ± 0.01
BQ concentration Partition coefficient, K
on sediment0 LLSd P
(Mg g'1 x 10-^) a0 = 0 a0 * 0
12.04 ± 0.39e
4.27 ± 0.01
22.95 ± 0.34f 1313 ± 170 2150 ± 440
8.01 ± 0.01 (a0 = -856 ± 433)
dTotal organic content = 1.4%
Concentration measured in supernatant with population standard deviation.
c
Concentration on sediment calculated from supernatant concentration with
population standard deviation.
LLS = linear least squares regression; see Part I, Section 5.4.2.3.
e —1
Measured concentration on sediment for one of the flasks was 8.80 ± 0.10 ug g
Measured concentration on sediment for one of the flasks was 17.63 ± 0.10 yg g .
-------
8.4.5 Biosorption
Benzo[f]quinoline was dissolved in phosphate buffer and added to a
suspension of the four types of bacteria prepared as described in Part I of
this report. The sorptions were conducted with suspensions of viable (optical
density of 2) and heat-killed cells at 25°C for 1 hr in 10-ml volumes in 15-ml
Corex tubes agitated in a Rotor drum. All sorptions were conducted in tripli-
cate. The centrifuged cell pellets and clear supernatants were extracted with
ethyl acetate and analyzed for BQ by HPLC.
The sorption coefficients developed with two levels of BQ had a greater
difference between those for heat-killed and viable cells (Table 8.7) than were
observed for any other compound examined under this contract. The higher sorp-
tion coefficient for the heat-killed cells suggests that the sorption is not a
receptor or enzymatic phenomenon.
TABLE 8.7. BIOSORPTION OF BENZoLfJQUINOLINE
Initial
BQ concentration Sorption
Condition of cells (ng ml) coefficient
Viable 115 158 ± 8
Viable 390 139 ± 7
Heat-killed 115 560 ± 67
Heat-killed 390 520 ± 19
o
Dry weights of bacterial cell mixture in viable-cell sorptions
= 1.05 g liter"1, and in heat-killed cells = 0.847 g liter"1.
—&——— / ug BQ ml"1 supernatant after 1 hr equilibration.
g cells
8.5 CHEMICAL TRANSFORMATION
8.5.1 Photolysis Rate
In air-saturated water, the direct photolysis half-life of BQ in sun-
light is less than a day. Photolysis experiments were conducted with BQ in
pure water and in natural waters from several sources. These solutions also
contained 1% methanol to increase the solubility -of BQ. Data for photolysis of
BQ in sunlight and at 313 nm are summarized in Table 8.8. The quantum yield
measured at 313 nm for photolysis of BQ in air-saturated pure water containing
1% methanol is 1.4 x 10 .
Photolysis of BQ in water that had been purged of air by nitrogen showed
a rate acceleration of 70%, indicating that oxygen may be quenching a photo-
excited state of BQ and that the reaction observed is a photohydrolysis or
rearrangement. The data in Table 8.8 also show that, while the photolysis rate
134
-------
TABLE 8.8. RATE CONSTANTS FOR PHOTOLYSIS OF BENZO[fJQUINOLINE*
u>
tn
Irradiation
source
, 313 nm
313
313
313
313
313
313
313
Sunlight, midday
mid- June
Sunlight , midday
mid-June
BQ concentration
(UR ml'1)
7.3
7.3
9.1
9.1
9.1
12.1
9.1
9.1
10.2
9.1
b
Solution
Pure water, air saturated
Pure water, N2 purged
Lake Tahoe water
Searsville Pond water
Coyote Creek water
8.0 ug ml"1 humic acid in
pure water
Methanol
Acetonitrile
Pure water
Pure water
Extent of reaction
(%)
39
55
— C 1
JJ-
•29
49
14
21
17
87
83
Rate constant
(k x 10s sec'1)
P
1.73 ± 0.15C'd
2.90 ± 0.27
1.92 ± 0.10
0.748± 0.071
1.34 ± 0.12
0.275 ± 0.055
0.475 ± 0.031
0.372 ± 0.040
P
14.0 ± 0.66
f
18r
a1.00 yg ml~l benzo[f]quinoline in water = 5.59 x 10"6 M.
Solutions contained 1% methanol as cosolvent.
Standard deviation.
Quantum yield for disappearance of BQ was 1.4 x 10"2.
6Half-life of 1.4 hours.
Half-life of 1.1 hours; based on one datum point at 83% reaction.
-------
of BQ in Lake Tahoe water is the same as that in the pure water solvent, the
rates in the waters from Coyote Creek and Searsville Pond are 28% and 57% slower,
respectively. Since the absorbances of the creek and pond waters (both about
0.07 at 313 nm) would account for less than a 10% reduction in rate due to light
screening effects, other processes must also be contributing to the rate decrease
in these waters.
A similar conclusion applies to the photoreaction in the water containing
8.0 ug ml~ humic acid where the rate constant is only 16% of that in pure water.
Light screening by humic acid because of the absorbance at 313 nm of 0.81 would
account for a rate constant 45% of that in pure water. The apparent sensitivity
of BQ triplet state (probably) to quenching by oxygen makes it probable that BQ
triplet is also quenched efficiently by other materials including those present
in natural waters.
Photolyses were also carried out using methanol and acetonitrile as the
solvents. The data in Table 8.8 show that the rates in these solvents are less
than 25% of the rate in water. Based on the limited number of experiments, it
is impossible to determine whether the rate difference was due to a change in
lifetime of the excited state owing to solvent interaction, to quenching by the
probable higher concentration of oxygen in organic solvents, or to reaction of
excited state BQ with solvent (that is, water versus methanol rates). The
solvent effects and the apparent quenching effects of oxygen and natural substances
pose many questions regarding the chemistry of these reactions, and further
studies on the phototransformation of BQ in water seem indicated.
The half-lives for direct photolysis of BQ in sunlight as a function of
the time of day and season were calculated by the procedure of Zepp and Cline
(1977), using a quantum yield of 1.4 x 10~2 measured at 313 nm and the measured
UV spectrum of BQ. These data are plotted in Figure 8.5. The half-life of less
than an hour calculated for midday in summer is in good agreement with measured
half-lives of 1.1 and 1.4 hours measured in midday in mid-June. As discussed
above, the presence of naturally occurring substances in water may slow the
photolyses significantly, so application of the direct photolysis data to
environmental assessments should be used with some caution.
8.5.2 Oxidation Rate
The susceptibility of 9.1 yg ml'1 (5.1 x 10~5 M) BQ to free radical
oxidation was examined using the AA-initiated oxidation reaction. (See Part I,
Section 6.3, of this report and Appendix B for discussion and procedure for this
experiment.) The BQ oxidation experiment was conducted at 50.0°C in 1% methanol
in water containing 1.0 x 10"^ M AA. At the end of 96 hours reaction, less than
1% of the BQ had been consumed. Similar lack of reactivity toward oxidation
was found for quinoline (Section 7.5.2), where only 3% loss of quinoline was
found for a 100-hour reaction time. If a 3% experimental error in the analysis
of BQ is assumed, the second-order rate constant k at 25°C is the same as
that for quinoline, or less than 2.8 M"1 sec"1. Assuming [R02.]= 10~9 M in the
aquatic environment, the half-life of BQ toward oxidation is more than 8 years.
The free radical oxidation of BQ by R02' radicals is then not competitive with
photolysis under environmental conditions.
136
-------
MEASURED -r
(SUMMER) f
12 NOON
4 3
TIME OF DAY
FIGURE 8.5. SEASONAL AND DAILY VARIATION OF PHOTOLYSIS HALF-LIFE OF
BENZO[f]QUINOLINE
137
-------
8.5.3 Hydrolysis Rate
BQ contains no groups that are hydrolyzable; therefore, no hydrolysis
studies were performed.
8.5.4 Products from Chemical Transformations
Two early eluting primary product peaks were found in the HPLC traces of
photolyzed BQ solutions. The relative size of these peaks was not affected by
the presence of oxygen during photolyses, although the photolysis did proceed
more slowly (by a factor of over 2) when oxygen was present (see Section 8.5.1).
The result suggests that oxygen serves to quench a BQ excited state (probably
triplet) during the photoreaction.
Different products were observed when the photolyses were carried out in
acetonitrile or methanol rather than in water. The rate in deoxygenated aceto-
nitrile was only one-eighth the rate in deoxygenated water (Table 8.8) and there
was no significant effect of oxygen on rate or products.
These observations indicate that the photoreaction of BQ in water is some
kind of photohydrolysis process. Attempts to identify products by concentration
and characterization by silylation and GC-MS were unsuccessful. Assuming that
the HPLC response factors are about equal to those of the BQ, these peaks
accounted for approximately 50% of the reacted BQ. In view of the importance
of photolysis of BQ in aquatic systems, more information is needed on the
structure of the photolysis products.
8.6 BIODEGRADATION
8.6.1. Development of Encrichment Cultures
Because BQ is frequently found in the presence of quinoline and quinoline
is readily biodegraded (see Section 7.6), enrichment procedures for BQ biode-
grading cultures were conducted with BQ as the sole carbon compound added to the
water samples and with BQ plus quinoline as carbon source, each at 10 ug ml"^.
All enrichments were initiated in 9-liter fermentors, and when BQ was being
degraded, transfers were made to shaker flasks containing basal salts medium
containing BQ alone or BQ plus quinoline. As with other compounds, when inducers
were used, attempts were made to grow the various mixed cultures, derived from
BQ plus quinoline enrichments, on basal salts/BQ medium.
The development of BQ biodegrading cultures was monitored by determining
the differences in UV absorptions at 345 and 355 nm of ethyl acetate extracts
prepared from fermentation broths with 2 volumes of ethyl acetate at pH 7.
Although BQ has a UV absorption peak at 345 nm and a larger peak at 266 nm, too
much interference was observed at this lower wavelength to make this practical
in our rapid assays. Quinoline has UV absorption maximum at 312-320 nm. Critical
analysis were confirmed by HPLC with a fluorescence detector.
138
-------
Table 8.9 indicates the times (days) at which biodegradations of BQ were
observed in the 9-liter fermentors when BQ alone, or BQ plus quinoline were the
organic substrates added to the diluted (4:5) water samples. Table 8.9 also
shows our success in developing culture systems that would biodegrade BQ in
basal salts medium with BQ, or BQ plus quinoline, as the only carbon substrates.
With every water sample, BQ-degrading systems were readily developed when
quinoline was present, but with only three of these six water samples was a
culture system developed that could utilize BQ as the sole carbon substrate in
the basal salts medium. All biodegrading cultures derived with the BQ plus
quinoline mixture rapidly adapted to growth on media with 50 ug ml~l BQ and
20 ug quinoline ml" 3-, and transfers were made daily.
The enrichment studies initiated with Coyote Creek water to which only
BQ was added developed a BQ-degrading system in 15 days. With the eutrophic
pond water, BQ degradation was observed at 43 days. All other 9-liter ferment-
ors with only BQ did no: have any detectible BQ degradation during this period.
i
The culture system derived from Coyote Creek water (with BQ as the only
added carbon substrate) degraded BQ in basal salts/BQ (10 yg ml"-'-) and Trypti-
case soy broth (100 ug ml~^)/BQ (10 ug ml"-*-) media. The BQ-degrading mixed
culture derived from aeration effluent from the Shell Oil Refinery treatment
plant could degrade BQ only if quinoline were present in the basal salts medium.
It also failed to degrade BQ when transferred to Trypticase soy broth (100 ug
ml~-1-)/BQ (10 ug ml ) medium or basal salts medium to which glucose (20 ug ml"-*-)
Difco yeast extract (2 yg ml"-"-), and BQ (10 ug ml"-*-) were added; but it grew
well on these two media. This indicated that with this culture system, BQ
degradation was dependent on cometabolism with quinoline.
8.6.2 Biodegradation Kinetics
Pseudo-first-order kinetic studies were conducted with a BQ-degrading
culture developed by original enrichment of Coyote Creek water with BQ. The
inoculum was first grown with shaking for 2 days in basal salts/BQ (10 ug ml" )
medium, 600 ml/2-liter Erlenmeyer flask, diluted to 2.8 liters and distributed
in four sterile 2-liter Erlenmeyer flasks and incubated with shaking for 18
hours. The combined broths were centrifuged, washed twice with basal salts
medium, suspended in 100 ml basal salts medium and shaken for 4 hours, centri-
fuged again and suspended to appropriate volumes for kinetic studies with BQ at
1000 ng ml"-1- in basal salts medium. The data in Table 8.10 indicate the con-
stancy of cell counts throughout the kinetic studies.
Figure 8.6 plots In BQ concentrations versus time. The pseudo-first-
order rate constants (k£) calculated from the slopes of the plots are 1.04 and
0.56 hr"1 for series 1 and 2, respectively. Consequently, the respective second-
order rate constants are 3.5 x 10~8 and 3.7 x 10~8 ml cell"1 hr"1. These data
correspond to a half-life (In 2/k£) of 193 hours assuming 105 degrading organisms
in a total population of 10^ organisms per milliliter. Since all enrichment
cultures grown on BQ plus quinoline degraded BQ more rapidly than when BQ was
the sole carbon source, these rate constants should be considered lower limits.
The degradation rates under cometabolic conditions in the environment should be
significantly higher.
139
-------
TABLE 8.9.. DEVELOPMENT OF BENZOff 1QUINOLINE BIOPEGRADING ENRICHMENT CULTURES
Q_
Source of water sample
Aeration effluent from
Palo Alto Sewage Plant
Eutrophic pond,
Woodside, CA
Coyote Creek, San Jose, CA
Aeration effluent from
South San Francisco plant
Aeration effluent from
Shell Oil Refinery, Martinez, CA
Lake Tahoe, CA
Compounds added
to water sample
BQ + Q
BQ
BQ + Q
BQ
BQ + Q
BQ
BQ + Q
BQ
BQ + Q
BQ
BQ + Q
BQ
Days for first
biodegradation
3
c
3
43
2
15
2
c
3
c
8
c
Development of
culture on salts
medium with
h b
BQ BQ + Q
No Yes
No
Yes Yes
Yes Yes
Yes
Yes Yes
No
No Yes
No
No Yes
No
samples diluted 4:5 with buffer and (NH.
^BQ = benzoff]quinoline, Q = quinoline.
"No degradation observed in 42 days.
solution.
-------
2000
1000
800
600
400
z
o
til
O
O
O
Ul
z
_l
o
a
N
200
8°
60
40
20
A
A —
I
40
80
120
TIME — minutes
160
200
240
FIGURE a6. BENZOmaillNOLINE BIODEGRADATION IN BATCH FERMENTATIONS
WITH HIGH CELL COUNTS
Initial cell concentration: O 3.0 x tO7 celts ml"1
A 1.5 x 107 cells ml"1
141
-------
TABLE 8.10. VIABLE CELL COUNTS DURING PSEUDO-FIRST-ORDER
BIODEGRADATION STUDIES WITH BENZO[fJQUINOLINE
Time
(hr)
0
60
120
180
210
240
Counts
Series 1
3.0
2.9
2.8
2.9
3.0
^•^
x 107 ml-1
Series 2
1.5
1.42
1.51
1.46
1.51
1.41
8.6.3 Metabolites
HPLC analyses of an extract prepared during biodegradation kinetic studies
revealed that one metabolite was formed. This product was not identified.
8.6.4 Discussion
The results of the enrichment studies illustrate the strong influence
that closely related products can have on the development of biodegrading culture
systems. The effluents of coal processing or petroleum plants would contain
series of structurally related heteroaromatics such as the quinolines. These
conditions would favor biodegradation processes similar to those observed with
BQ plus quinoline. BQ would not be expected to persist for prolonged periods,
but the susceptibility of derivatives of BQ cannot be estimated at this time.
The development of a cometabolic system from the aeration effluent of the Shell
Oil Refinery clearly indicated that the mixed culture would grow in media con-
taining conventional nutrients plus BQ but would not degrade BQ unless quino-
line were present. In nature, it is possible that other products could function
in cometabolism of BQ.
The half-life indicated in the above pseudo-first-order kinetic studies
was measured for a culture developed when only BQ was added to the diluted
buffered Coyote Creek water. For these studies, transfers on basal salts/BQ
were made at three- to four-day intervals. However, when cometabolic cultures
were developed with BQ plus quinoline, transfers could be made daily and the
half-life for BQ under these conditions would probably be reduced.
142
-------
9. LABORATORY INVESTIGATION OF 9H-CARBAZOLE
9.1 SYNOPSIS
The results of the laboratory investigation suggest that 9H-carbazole
should not be a significant environmental hazard. Both photolysis and biodeg-
radation are rapid. Carbazole" is not sorbed strongly, and it will not accum-
ulate in either sediments or biota. The volatilization and oxidation rates
are slow. The photolysis transformation products appear to be coupling prod-
ucts containing one or two oxygen atoms, but their structures were not posi-
tively identified. No;biodegradation products were observed, and it is likely
that complete degradation occurs.
The nine-compartment environmental exposure model predicted that the
steady-state concentrations of carbazole in solution, suspended solids, and
sediments near point sources in the presence of a continuous discharge of
1 pg ml"1 (1 ppm) carbazole are expected to be:
Suspended
Half-life"*" Solution solids Sediments*
(hr) (ME ml"1) (PR g"1) (MR g"1)
River 0.5 9.3 x 10"1 190 180 "
Pond 10 1.3 x 10~2 2.5 2.5
Eutrophic
lake 10 3.1 x 10~2 6.2 0.88
Oligotrophic
lake 3 4.0 x 10~2 8.0 1.2
9.2 BACKGROUND
"-The pyrrole nucleus of carbazole is found in natural products and in
products derived from natural substances. Carbazoles are present in coal,
petroleum, peat, and the smoke from combustion of these materials (Wu and
Storch, 1967; Utkina, 1965; Poulsen et al., 1974; Snyder, 1969; Braids et al.,
"Carbazole and 9H-carbazole are used interchangeably in this section.
^Predicted by the one-compartment model.
Average concentration in sediments.
143
-------
1967; Hartung and Jewell, 1962; Filatova et al., 1973). Carbazoles have also
been found in tobacco smoke (Hoffman et al. 1968, 1969). Carbazoles and indoles
(including alkylated derivatives) have been found in the effluents from the
coal processing industry (Schmidt et al., 1974; Karr et al., 1967).
9H-Carbazole was chosen for study because it is the simplest carbazole-
type nitrogen heterocycle found in energy-related effluents. The results are
also interesting for comparison with the results from the 7H-dibenzo [c,g]~
carbazole study (Section 10) to assess the relative importance of increasing
the number of aromatic rings of a basic molecular nucleus in terms of effects
on the transport and transformation processes. No information on the phys-
ical transport or chemical transformation of carbazole in the environment
was available when our laboratory studies began. One report was found re-
garding the biodegradation of carbazole. Medvedev et al. (1973) reported
that carbazole was less readily degraded by soil organisms than cresols,
aniline, fluorene, acenaphthene, or acridine but more readily degraded than
g-naphthylamine, (3-naphthol, pyrene, or chrysene.
The general physical properties of carbazole are given in Table 9.1.
TABLE 9.1. PHYSICAL PROPERTIES OF 9H-CARBAZOLE
Structure
Molecular weight
(CRC handbook, 1976-1977)
Melting point (°C)
(CRC handbook, 1976-1977)
Vapor pressure at 20°C (torr)
Solubility in water 20°C (ug ml"1)
pKa (see text)
1.00 ug ml"1 (1 ppm) carbazole in water = 5.98 x 10~6 M
167.21
247-248
7 x 10-4a
1.03 + 0.05b
a, -3
Estimated using vapor pressure data from CRC Handbook,
1976-77, and Clausius-Clapeyron equation.
Measured in this study, see Section 9.3.1
Carbazole is reported to have a pKa intermediate between those of indole
and pyrrole (Zherebtsov and Lopatinskii, 1970). The pKa values of the compounds
144
-------
are -2.4 and -3.8, respectively (Perrin, 1965). This indicates that carbazole
would be less than 1% in the protonated form at pH 1 and that the environmental
chemistry of carbazole will involve only the free base.
The literature information on carbazole was insufficient for making
preliminary assessment of the relative importance of the various transport and
transformation processes in the aquatic environment. Therefore, screening
studies were conducted for each of these processes.
9.3 ENVIRONMENTAL ASSESSMENT
9.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigation of the
transformation and transport processes for 9H-carbazole are summarized in
Table 9.2.
TABLE 9.2. SUMMARY OF 9H-CARBAZOLE LABORATORY DATA
Process
Q
Sorption
Volatilization
Q
Photolysis
Oxidation
Hydrolysis
Biodegradation
Sorption equilibrium
S = K S
s p w
Rate expression
k [C]
VL J
kp[C] = $ (ZZxeA)[C]
k [RO •] [C]
ox1 2 J L J
NA
Pmax
^>2 Y K
s
Partition coefficient
K = 175 ± 21
P
Rate constant at 25°C
k < k° x 10~4
V V
k = 6.6 x 10~5 sec"1
P
k =29 (M"1 sec"1)
ox
\ = o
k = 5 x 10-7 (mi cell"1^^1)
*a
On Coyote Creek sediment.
See discussion in Part I, Section 5.3 of the final report.
°Midday sunlight in late January.
9.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved 9H-carbazole calculated for individual
transformation of removal processes following a spill are listed in Table 9.3.
Although these half-lives vary among water bodies, photolysis consistently
dominates the transformation and removal processes. Biodegradation is at
least 2 to 5 times slower even in the presence of large acclimated microbial
145
-------
populations. Oxidation half-lives are expected to be an order of magnitude
greater than biodegradation half-lives except in oligotrophic lakes, where
oxidation is roughly 5 times faster than biodegradation. Sorption half-lives
have not been measured, but the times needed to reach equilibrium are generally
less than half an hour for those compounds for which the rates have been measured.
Nonetheless, since the partition coefficients were generally low, the amount of
carbazole removed from the water by sorption is expected to be low, perhaps
1% to 6% of the amount entering the water body.
Thus, excluding the effects of dilution, the half-life for removal of
carbazole from natural waters may be as short as 3 hours in oligotrophic lakes
during the day in the summer, and a long as 4 days in eutrophic lakes during
the winter.
TABLE 9.3. TRANSFORMATION AND TRANSPORT OF
9H-CARBAZOLE PREDICTED BY THE ONE-COMPARTMENT MODEL
Photolysis
Oxidation,
, half -life (hr)a
half-life (hr)
Eutrophic
River pond
6 15
>2£0 >240
Eutrophic
lake
15
>240
Oligotrophic
lake
3
>240
Volatilization, half-life (hr) > 10 > 10
Hydrolysis, half-life (hr)
Biodegradation, half-life (hr) 14 14
Half-life for all processes
except dilution (hr) 5 10
Half-life for all processes
including dilution (hr) 0.5 10
Amount carbazole sorbed
(mg m~3)b 20 60
Percentage carbazole sorbed 2.0% 5.6%
> 10"
14
10
10
10
1%
> 1Q
>1000
10
1%
Estimates are the average photolysis rates on a summer day at 40° latitude.
Photolysis rates in midwinter are at least three times slower.
1 ug ml"1 of 9H-carbazole in aqueous phase and partition coefficient of
200 are assumed.
146
-------
9.3.3 Persistence
The persistence of concentrations that approximate the initial concen-
trations following an acute discharge should largely be a function of the
dilution rate in fast flowing streams. Persistence of relatively high concen-
trations in lakes and ponds is expected to be determined by rates of trans-
formation within the aqueous phase. Sorption is not expected to significantly
affect the persistence of concentrations approximating those prevailing immed-
iately after a spill, but it will be a major determinant of the relatively low
concentrations of carbazole to be expected in ensuing weeks as desorption from
contaminated sediments becomes the major source of carbazole in solution.
Carbazole may be more persistent in eutrophic water than in oligotropic
waters if the major transformation pathway, photolysis, is less effective in
such waters, as assumed in Table 9.3. This tendency is unlikely to be present
except in frequently contaminated waters. Even in these waters there may be a
lag of several hours before biodegradation begins. Moreover, since the sediment
loading is typically higher in eutrophic waters, the amount of carbazole sorbed
to suspended solids should be greater than in oligotrophic waters, creating a
larger reservoir of carbazole that is subject to later desorption and therefore
the presence of low concentrations of carbazole for long periods.
9.3.4 Mass and Concentration Distributions Calculated Using Computer Models
Table 9.4 summarizes the distribution of mass and concentration of
carbazole expected at steady state during chronic discharge to each of four
types of water bodies, as calculated with the aid of the computer model. The
pseudo-first-order rate constants used in these simulations are presented in
Appendix A.
The steady-state concentrations of carbazole in solution in the simu-
lated stream segments imply a removal from solution of approximately 3% per
kilometer. However, in smaller, and hence more slowly flowing rivers than we
have assumed in the simulations, 10% to 20% of carbazole may be removed
within the same river reach.
The steady-state concentrations of carbazole in solution are roughly
10~2 to 10~^ pg ml in pond and lake simulations. This is rou.ehly 100 to
10,000 times less than the input concentrations (Table 9-4). In the pond
simulation,the steady-state concentration of carbazole is estimated to be 1%
of the concentration in inflowing waters. Within the lake simulations, the
concentrations of dissolved carbazole in the surface waters decrease 100-fold
with increasing distance from the source, and concentrations in the hypolimnion
are predicted to be 1,000 times lower than the concentrations in inflowing
waters.
The model predicts that virtually all the carbazole is in the bottom
sediments at steady state, because it is assumed that carbazole concentration
in sediments equals the concentrations on suspended solids at steady state and
that carbazole does not undergo transformation following sorption. We suspect,
however, that biodegradation of carbazole may take place in the sediment layers.
Hence, the estimated values of carbazole in sediment are probably upper limits
147
-------
TABLE 9.4. DISTRIBUTION OF 9H-CARBAZOLE IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations of 1 yg ml"l 9H-carbazole)
•P-
oo
Compartment 1
(surface water)
Solution
Suspended solids
Compartment 2
(surface water)
Solution
Suspended solids
Compartment 3
(surface water)
Solution
Suspended solids
Compartment 5
(bottom water)
Solution
Suspended solids
Compartments 7-9
(sediment)3
Solution
solids
Total mass
Pond
Mass Cone.
(kg) (yg g-1)
2.6 x 10"1 1.3 x 10"2 2
1.5 x lO-2 2.5 5
2
5
__ __ 2
— — — — 5
«— — .« —
—
3.3 x 10-3 1.3 x IO-2 6
1.7 2.5 3
2.0 4
River
Mass
(kg)
.8 x IO2
.6
.7 x IO2
.4
.6 x 102
.2
_
-
.8 x 10
.7 x IO3
.4 x IO3
Cone.
(vs g~i)
9.3 x
1.9 x
9.0 x
1.8 x
8.6 x
1.7 x
__
—
9.0 x
1.8 x
10-1
102
10-1
102
10-1
102
10-1
102
Eutrophic
Mass
(kg)
7.8
7.8 x 10~2
2.3
2.3 x 10-2
8.2 x 10"2
8.2 x 10-4
2.9 x 10~2
2.9 x 10-4
3.84 x 10"2
12
22
3.
6.
9.
1.
3.
6.
1.
2,
4.
8.
Lake
Cone.
(yg g"1)
1 x lO"2
24
2 x ID"4
8 x 10"!
3 x 10"4
6 x 10~2
1 x 10"5
3 x lO"3
4 x 10~3
8 x 10-1
Oligotrophic Lake
Mass
(kg)
1.0
1.0
3.8
3.8
1.7
1.7
1.4
1.4
5.12
16
21
x 10
x IO-1
x ID'2
x IO"1
x 10'3
x 10-1
x 10~3
x IO-2
Cone.
(yg g"1!
4 x
8
1.5
3.0
6.8
1.4
5.6
1.1
5.8
1.2
ID"2
x 10" 3
x 10"!
x 10-4
x 10"1
x 10~5
x 16"2
x 10~3
The amounts given for solid and solution phases in the sediment compartments (7, 8, and 9) are based on the assumption
that the carbazole concentration in the solution phases of these compartments at steady state is the same as the
concentrations in the solution phases of the adjacent water compartments (2, 5, and 3, respectively). These concentrations
may be overestimated because it was assumed sorbed material is not biodegraded.
-------
for the concentrations that may be expected in natural environments. Values
shown in Table 9.4 for sediments are estimated from the above assumption. Note
that the interchanges between water compartments and sediment compartments are
slow processes. Sediment compartments may require very long times before
steady-state conditions are established.
If an acclimated microbial culture is developed in the receiving water
bodies, biodegradation may contribute significantly to carbazole removal.
Figure 9.1 illustrates the effects of varied assumptions regarding metabolic
activity in a partially mixed two-compartment pond system. Steady-state con-
centrations of carbazole in the pond simulations vary 1.5 to 3.0-fold in response
to a 30-fold variation in biotransformation rates. Variations in the steady-
state concentrations following termination of discharge are as much as 20-fold,
roughly the range by which the biotransformation rates varied.
Figure 9.1 also shows the buildup and decline of carbazole in the pond
simulation during and following discharge. Steady states are reached within
two days for the solution and suspended solid phases, but concentrations in
the sediments are still increasing slowly when discharge is stopped on the
seventh day. Predicted carbazole concentrations in the suspended solid and
solution fractions decline fairly rapidly after discharge stops and approach a
new steady state in about four days. Desorption then provides a lower limit
to the decreases in the concentrations of carbazole in the dissolved and sus-
pended solids fractions, giving a relatively long-lived loading in the water
column of roughly 10"" 7 yg ml~l. The half-life of carbazole following a spill,
as estimated from the rate of decline in the pond simulation, is approximately
the same as that predicted by the one-compartment model.
Similar patterns prevail in the river simulation (Figure 9.2) except
that steady-state conditions are obtained far more rapidly (within hours in-
stead of days). Because of the high dilution rates in the stream and the
presence of multiple compartments, the concentration of carbazole in solution
changes abruptly in the first compartment after discharge is stopped, since
from then on it receives only uncontaminated water. Successively less abrupt
changes occur in the downstream compartments because they receive water that
is increasingly contaminated by carbazole desorbed from the sediments of the
previous compartment. This downstream movement of the area of highest concen-
tration of dissolved carbazole conceals the role of the initial source, which
is now 2 km upstream from the area of maximum concentration. Similar shifts
occur in the concentrations of carbazole on suspended solids, but are omitted
from the figure for the sake of clarity.
Buildup and declines in the lake simulations (Figures 9.3 and 9.4)
most nearly resembled the pond simulations. During the period of accumulation
of carbazole, large amounts of carbazole appear in the compartment receiving
the discharge, little in the distant surface compartments, and even less in
the bottom waters because of the slow transports between hypolimnion and
epilimnion. However, during the fall and spring overturns, rapid vertical
mixing of the lake waters should result in significant transport of carbazole
to the lower portions of the lake, causing relatively high., concentrations in the
hypolimnion for one to three weeks. After discharge stops, the concentration
of carbazole in the hypolimnion is predicted to be higher than that in the
surface compartment furtherest from the source.
149
-------
4 x 10
Ui
O
JTION X ^
I fe \
y i "
DISCHARGE -+J
'
THE PSEUDO FIRST ORDER
BIODEGRADATION RATES CONSTANTS
.231 x 10'2hr'1
231 x 10"1hr-1
.693 x 10"1 hr'1
10
100 150
TIME - hours
250
FIGURE 9.1 PERSISTENCE OF 9H-CARBAZOLE IN A PARTIALLY MIXED TWO-COMPARTMENT
POND SYSTEM
-------
10'1
SL
I
ui
O
N
00
(T
O
u.
O
HI
o
8
10'2
10
10-
I
COMPARTMENT
DISCHARGE
STOPPED
SOLUTION
SEDIMENTS
10
TIME-hours
FIGURE 9.2 PERSISTENCE OF 9H-CARBAZOLE IN A PARTIALLY MIXED
RIVER SYSTEM
151
-------
3 X 10'1
/
• -• r
NJ
y
**
S
/
/
f
^
3 _
^^-^"s""
COMPARTMENT
' l~^
I
I
I
I
I
I
I
I
DISCHARGED
STOPPED
SOLUTION
SEDIMENTS
100
200
300 400
TIME - hour*
PERSlSTENCE
500
600
700 720
-------
5 x 10
SEDIMENTS
100
200 300 400 500
TIME - hours
600
700 720
FIGURE 9.4 PERSISTENCE OF 9H-CARBAZOLE IN AN OLIGOTROPHIC LAKE
153
-------
9.3.5 Discussion
Despite the uncertainties inherent in extrapolating laboratory data
to the field, it is evident that the concentrations of carbazole should rapidly
decline by 5 to 6 orders of magnitude after discharges to the water are stopped.
The concentrations of carbazole in lakes should reach a new steady state within
one to two weeks after discharges are stopped and within hours in rivers.
The most significant uncertainty lies in the estimation of the ultimate
concentration of the carbazole, since this is determined largely by the proper-
ties of the sediments, and in the potential biological effects. Since the
partition coefficient of carbazole is sensitive to the composition of the sedi-
ments, the ultimate concentration of carbazole in water bodies with a long his-
tory of pollution with carbazole may vary by one or two orders of magnitude
even if discharges are stopped. However, the significance of the variations
cannot be appraised with the available tbxicological data. We can only say
that the concentrations will be quite low shortly after discharges stop.
9.4 PHYSICAL PROPERTIES
9.4.1 Solubility in Water
The solubility of 9H-carbazole in water at room temperature (20.0 + 0.1°C)
was measured using the method described by Campbell (1930). The average solu-
bility was 1.03 + 0.05 ug ml [6.16(+ 0.30) x 10~6 M], based on replicate
determinations.
9.4.2 Absorption Spectrum
9H-Carbazole absorbs light strongly in the solar region to 400 nm.
The absorption spectrum of 9H-carbazole in 26% acetonitrile/74% water
was measured from 210 to 500 nm. The pH of the solution was 7 and was not
adjusted. The absorption coefficients at wavelength intervals from 297.5 to
400 nm and at wavelengths of 313 and 366 nm are reported in Table 9.5.
The pH-dependence of the absorption spectrum was not measured because
our estimate of the pKa of 9H-carbazole ('v- -3, Section 9.2), indicated that less
than 1% would be in the protonated form even at pH 1. The environmental chem-
istry of carbazole will involve only the free base form.
9.4.3 Volatilization Rate
The volatilization rate of 9H-carbazole was measured using the method
of Hill et al. (1976), which is described in detail in Appendix B
of this report. The concentration of carbazole did not change measurably
over the four-day period of the experiment. The average oxygen reaeration
rate was 3.76 + 1.82 hr~*. An upper limit for the volatilization rate
* This value represents the average and standard deviation of four reaeration
measurements taken over the period of the experiment.
154
-------
TABLE 9.5. ABSORPTION SPECTRUM OF 9H-CARBAZOLE IN 26%
ACETONITRILE/74% WATER (V/V) AT pH 7a
Center of , Average
wavelength interval absorption coefficient
(M"1 cm"1)
297.5
300.0 .
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330
340
350
360
370
380
390
400
5540
3100
2440
2270
2390
2530
2600
2700
2920
3190
3170
2900
1520
166
23
13
12
2
0
a9H-Carbazole concentration was 40.8 pg ml"1 (2.44 x 10"^ M)
b,,
The wavelength intervals are given in Appendix B, Table B.I.
The absorption coefficients at
and 12 M~l cm"1, respectively.
°The absorption coefficients at 313.0 and 366.0 nm are 2600
155
-------
constant was therefore estimated based on the measurement error. The 95%
confidence limit (two times the average standard deviation; of the measurements
of the carbazole concentration was 0.654 + 0.24 ug ml-1. Therefore, the low-
est possible final concentration of carbazole at the end of the experiment was
about 0.63 ug ml"-'-. Therefore
kv < [-ln(0.63/0.654)]/94 = 4 x 10"4 hr"1 (9-1)
and
k° -4
_Y_ < 1 x 10 (9.2)
k°
v
9.4.4 Sorption on Clay and Sediments
The sorption partition coefficients of 9H-carbazole were measured on
Ca-montmorillonite clay and Coyote Creek sediment. Care was taken to assure
that the samples were protected from light at all times to minimize photolysis
of the carbazole (Section 9.5.1). The initial concentration of carbazole
in each sample was always less than the saturation concentration of 1.0 ug ml"-'-.
Each experiment consisted of replicate samples at one carbazole concentration
and two sediment loadings, plus blanks that contained only carbazole. The
data are summarized in Table 9.6.
The calculation of the partition coefficient was based on the results
of the analysis of the supernatant at equilibrium. The amount of carbazole
sorbed on the sediments was calculated by finding the difference between the
initial amount and the amount that was found in the supernatant after equili-
bration with the sediment. The estimates reported in Table 9.6 show that
sediment sorption is not an important environmental pathway for 9H-carbazole
compared with photo-oxidation and biodegradation. Therefore, no further studies
of the sorption of carbazole by either sediments or microorganisms were carried
out.
9.4.5 Biosorption
Because of the rapid development of enrichment cultures, rapid photol-
ysis , and low sorption coefficients on clays and sediments, biosorption studies
did not appear to be warranted and were not conducted.
9.5 CHEMICAL TRANSFORMATIONS
9.5.1 Photolysis Rate
The photolysis of 9H-carbazole is rapid and unaffected by oxygen; in
sunlight the half-life for direct photolysis is less than a day. Photolysis
experiments were conducted with 1.0 pg ml~l concentrations of 9H-carbazole in
pure water and in natural waters from several sources. These waters also con-
156
-------
TABLE 9.6. SORPTION OF 9H-CARBAZOLE ON SEDIMENTS
Ul
-j
Total
organic
carbon
Sediment (percent)
Ca-montmor-
illonite 0.06
clay
Coyote Creek 1.4
Carbazole Carbazole
Sediment
concentration
(yg ml-1)
13
25
6.8
14
concentration concentratioi
in supernatant on sediment
(UR ml'1)* (MR g'1)13
0.52
0.50
0.24
0.14
± 0.01 1.9 ± 1.1
± 0.01 1.7 ± 0.5
± 0.03 41 ± 4
± 0.01 29 ± 1
Partition Coefficient, k
LLSC P
a = 0 a ^ 0
0 O
3.2 ± 1.1 -3.k ±24
(a = 18 ± 12)
175 ±21 87 ± 52
(a = 18 ± 10)
'Concentration measured in supernatant with population standard deviation.
Concentration on sediment calculated from supernatant concentration with population standard deviation.
"LLS = linear least squares; see Appendix B, Section B.I.4 for description of regressions.
Limits are 95% confidence limit.
-------
tained 1% acetonitrile to increase the solubility of carbazole. Data for photol-
ysis of 9H-carbazole in sunlight and at 313 nm are summarized in Table 9.7.
Good first-order kinetic behavior was found for photolyses at 313 nm carried
out to beyond two half-lives of carbazole. The quantum yield for 9H-carbazole
phototransformation in pure water (containing 1% acetonitrile) is 7.6 (+ 0.3)
x 10~3.
Photolysis experiments carried out at 313 nm on 1.0 yg ml"1 9H-carbazole
solutions that had been purged with nitrogen showed that oxygen had no effect
on the photolysis rate. The photolysis of 9H-carbazole in 100% acetonitrile
with 313 nm light was slow. The rate constant was less than 2 x 10~6 sec"1,
which is only 3% of the rate constant found for the direct photolysis of 9H-
carbazole in 1% acetonitrile in water as solvent (see Table 9.7). The re-
action must then be some kind of photohydration process.
The data in Table 9.7 show that the photolysis rates of carbazole in
waters from Lake Tahoe or Coyote Creek are the same within experimental error
as the rate in pure water. The experiments with Searsville Pond water and in
pure water containing 8.0 yg ml"-*- humic acid do show retardation of the photol-
ysis rate, but these slower rates are accounted for by screening by the
natural substances in the waters. The absorbances of 0.08 for Searsville
Pond water and 0.43 for humic acid in water account for the 25% and 66% de-
crease in rate due to reduced light intensity.
TABLE 9.7. RATE CONSTANTS FOR PHOTOLYSIS OF 1.0 yg ml"1
9H-CARBAZOLE3
Irradiation
source
313 nm
313 nm
313 nm
313 nm
313 nm
Solution53
Pure water
Lake Tahoe water
Coyote Creek water
Extent of
reaction (/
85
87
73
Searsville Pond water 65
Humic acid (8 yg ml
in pure water
-1)
35
Rate constant
(k x 105 sec-1)
°)
7.50 + 0.18C'd
7.69 + 0.38
7.66 + 0.39
5.74 + 0.23
2,46 + 0.11
Sunlight, midday
late January Pure water
57
6.59 + 1.21
a1.00 yg ml-1 9H-carbazole in water = 5.98 x 10~6 M.
Solutions contained 1% acetonitrile as cosolvent.
£
Standard deviation
Quantum yield for disappearance of 9H-carbazole at 313 nm was 7.6 x
eHalf-life of 2.9 hours.
158
-------
The half-life for direct photolysis of carbazole in sunlight as a
function of the time of day was calculated by the procedure of Zepp and Cline
(1977) using a quantum yield of 7.6 x 10" 3 measured at 313 nm and the measured
uv spectrum of carbazole. The data are plotted in Figure 9.5. The instant-
aneous half-life of 2.7 hours calculated for midday in winter is in excellent
agreement with the measured half-life of 2.9 hours.
9.5.2 Oxidation Rate
The susceptibility of 1.0 yg ml"1 (6.0 x 10~6 M) 9H-carbazole to free
radical oxidation was examined using the AA-initiated oxidation reaction.
See Part I, Section 6.3 of this report and Appendix B for discussion and
procedure. This experiment was carried out at 50.0°C in 1% acetonitrile in
water containing 1.0 x 10~4 M M. At the end of 136 hours, 32% of the car-
bazole had been reacted. A rate constant (k [RO™-]) of 8.08 (± 0.36) x 10~7
.sec"1 was obtained for the oxidation of 9H-carbazole under the reaction con-
ditions; this rate constant corresponds to a half-life of 9.9 days. Under
these conditions (Part I, Section 6.3), this corresponds to a second-order
rate constant kQx of 2.6 x 102 M~l sec"l At 25°C, k would be 29 M"1 sec"1
and with the assumption that [RO •] = 10 M in the environment, the half-life
of carbazole toward oxidation is 280 days. The free radical oxidation of
carbazole by RO • is clearly not competitive with photolysis or biodegradation
of carbazole under aquatic environmental conditions.
9.5.3 Hydrolysis Rate
Carbazole contains no groups that are hydrolyzable; therefore, no
hydrolysis studies were carried out.
9.5.4 Products from Chemical Transformation
Four compounds were isolated by HPLC from the direct photolysis of
carbazole and subjected to mass spectrometry. However, only two products
could be identified as being derived from 9H-carbazole, giving apparent molec-
ular ions of m/e 364 and 368 (9H-carbazole has m/e = 167). These values
indicate that both products are coupling products of carbazole (molecular
weight of 332), with the m/e 364 having two added oxygen atoms. Since the
photolysis rate was unaffected by the presence of oxygen and photolysis did
not occur except in the presence of water, we assume that the source of the
oxygen is water. The m/e 368 product corresponds to addition of two water
molecules to a 9H-carbazole coupling product (36 + 332). No further infor^
nation was provided by the mass spectra. Some additional structure information
seems necessary to identify these products.
9.6 BIODEGRADATION
9.6.1 Development of Enrichment Cultures
Enrichment studies were initiated with 5-liter liquid volumes in 9-
liter, aerated bottle-fermentors. 9H-Carbazole (50 mg ml"1 in DMSO) was added
159
-------
H
01
O
12 NOON
TIME OF DAY
FIGURE 9.5 SEASONAL AND DAILY VARIATION OF PHOTOLYSIS HALF-LIFE OF
9H-CARBAZOLE
-------
to these fermentors to give initial concentrations of 10 yg ml"-*-. This level
is about 10 times the solubility of 9H-carbazole (Section 9.4.1) and some
precipitation was observed. Water samples used in these enrichment studies
and the characteristics of the first observations of 9H-carbazole biodegrada-
tions are presented in Table 9.8.
TABLE 9.8. DEVELOPMENT OF 9H-CARBAZOLE BIODEGRADATION
ENRICHMENT CULTURES
First observations of
biodegradations
Sources of samples
Days of
incubation
Percent
degraded3
Coyote Creek, San Jose, CA
Pond near Searsville Eake, Woodside, CA
Palo Alto Sewage Plant, Palo Alto, CA
South San Francisco Sewage Plant,
South San Francisco, CA
Shell Oil Refinery Industrial Waste
Plant, Martinez, CA
Lake Tahoe, CA
6
6
5
5
11
42
70%
90
99
70
60
Determined by UV assay.
In all cases, subtransfers were readily adapted to growth on 9H-
carbazole/inorganic salts media in shaker flasks incubated at 25°C. The
9H-carbazole concentrations in the media were raised progressively from 10,
to 30, and finally to 100 yg ml"-'-. Before the individual culture systems
were preserved in the vapor space in a Dewar flask containing liquid nitrogen,
> 90% biodegradation of 100 yg ml 9H-carbazole media was generally observed
in 24 to 48 hours of incubation with 2% (v/v) inocula.
9 • 6 .-"2 Biodegradation Kinetics
Because of the low solubility of carbazole and the lower limits of
detection by HPLC (0.1 yg ml ), pseudo-first-order kinetic studies (see Part
I, Section 7.2 of this report) were conducted with a biodegrading culture
system developed from the Palo Alto sewage plant aeration effluent. These
studies involved two series of batch fermentations (experiments 1 and 2,
Table 9.9) with initial 9H-carbazole concentrations of 0.7 and 0.8 yg ml"1,
respectively. Each series of kinetic studies was initiated with two levels of
stationary phase cells (a and b, Table 9.9).
The carbazole concentration versus time data are plotted in Figures
9.6 and 9.7. The complete metabolism of 9H-carbazole did not result in any
161
-------
TABLE 9.9. CELL COUNTS3 DURING KINETIC STUDIES
OF THE BIODEGRADATION OF 9H-CARBAZOLE
Time
(hr)
0
0.5
1.0
1.5
2.0
2.5
Experiment
la Ib
4.7 1.41
4.6 1.57
4.7 1.51
1.65
2a
1.69
1.44
1.63
1.62
2b
0.76
0.69
0.70
0.78
0.91
0.71
aCell x 10~7 ml-1.
significant change in microbial mass, as shown by the data in Table 9.9. In
the lower cell-count experiment Ib, there was a lag period of approximately
6 minutes before rapid degradation took place. No lag period was observed wit1"
the higher cell population.
The pseudo-first-order rate constants, Ic = kT~X were calculated
from the slopes of the plots of In [carbazole] versus time (Figures 9.6 and
9.7). From these data, the second-order rate constants were calculated by
using the following relationship:
k£2 (ml cells"1 hr"1) = ^"l (9.3)
cells ml
Values obtained for k;*, and kf~ are listed in Table 9.10.
After experiments la and Ib, the mixed culture was refrigerated for
two days, and three sequential 2% (v/v) transfers were made daily in 100 ug
ml~l carbazole/basal salts medium, followed by two daily 10% (v/v) transfers
into 50 yg ml~l carbazole/basal salts medium. In contrast with previous ob-
servations, the first fermentation with the 10% inoculum resulted in slower
degradation of carbazole. However, with the second 10% inoculum, carbazole
was completely degraded in 18 hours.
These cells were used for the second kinetic study (experiments 2a
and 2b) conducted under conditions similar to those described above; results
are presented in Figure 9.7 and Table 9.10. In this study, with both levels
of cultures, there were lag periods in which carbazole was slowly degraded.
After the lag period, the degradation rates were greater then those in the
first kinetic study (experiments la and Ib, Figure 9.6). For experiments 2a
and 2b the rate constants were calculated from data obtained during the period
of rapid degradation. Considering the fact that the mixed culture behaved
quite differently in the two experiments during preparation of cells for kinetic
162
-------
40
60
TIME — minutes
80
100
120
FIGURE 9.6 9H-CARBAZOLE BIODEGRADATION IN BATCH FERMENTATION
WITH HIGH CELL COUNTS (Experiment 1)
Initial cell concentrations: 4.7 X lO'mr1 (O).1.41 X lO'mf'jA)
163
-------
40
60
TIME — minutes
80
100
120
»™HHCEU
WITH HIGH CELL COUNTS (Experiment 2)
inltfal cel, concentrations: ,.«, X 107 ml" (O). 0.76 X lo'm." (A)
164
-------
TABLE 9.10. RATE CONSTANTS FOR BIODEGRADATION OF 9H-CARBAZOLE
Pseudo-first-
Initial order rate Second-order
cell level constants rate constant
Experiment3 (cell ml-1) x 107 (hr~l) (ml cell-1 hr"1) x 107
la
Ib
2a
2b
1.4
4.7
0.76
1.7
2.7
9.2
7.3
9.7
Average
1.9
2.0
9.6
5.7
4.8 ± 3.7
i
a <'
Data from experiments 1 and 2 are plotted in Figures 9.6 and
9.7, respectively.
studies, this is a surprisingly close agreement. The average value of kTo ,
(4.8 ± 3.7) x 10" 7 ml cell"1 hr"1, obtained is probably not significantly
different in environmental situations.
9.6.3 Major Metabolites in Carbazole Biodegradation
In the many HPLC analyses (fluorescence and UV detectors) performed on
concentrated extracts of acidified fermentation samples, no significant evidence
for metabolites was observed. These observations suggest that after the initial
attack on the carbazole molecule, degradation took place rapidly to products
that did not fluoresce or absorb in the ultraviolet. Gas chromatography
analyses also gave no evidence of metabolites.
9.6.4 Discussion
The ease of development of carbazole biodegrading enrichment cultures
was comparable to our experience with methyl parathion. The data suggest that
biodegradation of carbazole would probably take place under a variety of en-
vironmental conditions and, therefore, carbazole should not present any major
pollution problem. n-Alkylated carbazoles would also possibly degrade rapidly.
During the biotransformations of 100 yg ml /basal salts medium with
mixed cultures the biodegrading bacteria were in fact exposed to levels of
carbazole equal to the saturation concentration (approximately 1 ug ml"1) and
not to the total initial concentration in the medium, which included solid
carbazole. In the presence of other solutes, it is possible that the saturation
concentration of carbazole could be elevated, but there is no available infor-
mation on the antimicrobial activity of this compound.
165
-------
The differences in behaviorial patterns of the mixed cultures used in
the two sets of kinetic studies illustrated in Figures 9.6 and 9.7 are inter-
esting and undoubtedly reflect the complexities of mixed culture fermentations.
No study was made of the component cultures, but the results may be due to
relative changes of some strains and/or to the development of mutants during
the five days between the two sets of experiments. In ecological studies of
microorganisms in aquatic or soil habitats, much more significant behavior
changes have been observed over a one-week period (Silvey et al., 1975).
166
-------
10. LABORATORY INVESTIGATION
OF 7H-DIBENZO[c,g]CARBAZOLE
10,1 SYNOPSIS
The results of the laboratory investigations suggest that 99% of the
7H-dibenzo[c,g]carbazole (DBG) entering natural waters will be sorbed onto
the sediment and that the remaining 1% will be rapidly transformed photochem-
ically to yield two products, one of which has added oxygen from reaction
with water and is probably phenolic. Photolysis of DBG is expected to be the
dominant process in clear waters. In turbid waters, sorption by suspended
sediments will be more important. No information was obtained concerning
possible phototransformation of DBG sorbed on suspended sediments. Enrichment
cultures that could degrade DBG were not obtained. Volatilization of DBG will
be extremely slow.
Steady-state concentrations of DBG in solution, suspended solids, and
sediments near point sources in the presence of a continuous discharge of
10 ng ml"1 (10 ppb) are expected to be:
&
Half-life Solution .Suspended solids Sediments
(hr) (ng ml-1) (ng g"1) (ng g"1)
River 0.36 2.8 55,000 49,000
Pond 1.5 0.01 180 200
Eutrophic
lake 1.5 2.9 x 10~2 580 81
Oligotrophic
lake 0.5 1.5 x 10~3 30 4.4
These concentrations may be hazardous even though they are approximately 1000
times lower than those tested to date under relevant conditions (Christensen
et al., 1976). The few toxicity data available suggest that even low concentra-
tions and brief exposures are carcinogenic, and caution certainly is warranted.
10.2 BACKGROUND
7H-dibenzo[c,g]carbazole (DBG) is a known carcinogen (Christensen et al.,
1976). It has been found in the combustion products of tobacco smoke and is
Predicted by one-compartment model.
167
-------
present in the particulate matter of polluted atmospheres. No reports on the
distribution of DBC in the aquatic environment were found in our literature
search.
DBC was chosen for study because of its known health hazard and because
it is representative of the indole- and carbazole-type nitrogen heterocycles
found in energy-related effluents (Schmidt et al., 1974; Karr et al., 1967).
The results of the DBC study are also useful for comparison with the results
from the carbazole study (Section 9) to assess the relative importance of the
transport and transformation processes, which may be affected by the aromatic
ring structure. No information on the physical transport or the chemical and
microbiological transformation of DBC was available when our laboratory
studies began.
The general physical properties of DBC are given in Table 10.1.
TABLEJ.0.1. PHYSICAL PROPERTIES OF 7H-DIBENZO[c>g]CARBAZOLE
Structure
Molecular weight 267.31
Melting point (°C) 155-159
Vapor pressure at 20°C (torr) 10~9
Solubility in water (ng ml~l) 63 + 3b
1.00 ng ml"1 (1 ppb) DBC in water = 3.74 x 10~9 M
a
Estimated by comparison with benzo[a]pyrene, Section 6.
Measured in this study, see Section 10.3.1.
The pK of DBC is not known, but by analogy to carbazole (see Section 9),
the pK of D§C is expected to be less than zero, indicating that DBC would be
less tnan 1% protonated even at pH 1. The environmental chemistry of DBC will
then involve only the free base (neutral) form.
While we anticipated that sorption onto sediments and photolysis were
probably the most important pathways for DBC, the literature data were insuffi-
cient to allow us to quantitatively estimate the importance of these processes
in comparison to biodegradation, volatilization, and oxidation. Therefore, we
decided that screening studies should be conducted for each of these processes.
168
-------
10.3 ENVIRONMENTAL ASSESSMENT
10.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigations of the
transformation and transport processes for DEC are summarized in Table 10.2.
TABLE 10.2. SUMMARY OF 7H-DIBENZO[c,g]CARBAZOLE LABORATORY DATA
Process
Sorption equilibrium
Partition coefficient
Sorption S = K S K = 20,500 ± 2,000a
v s p w P
Volatilization k
1 Rate expression Rate constant at 25°C
v[DBC]b kv * (9.4 x 10-3) k°
Photolysis k [DBC]° = $ (ZZ. e,. ) [DEC] k = 5.2 x I0~k sec~i
p A A p
Oxidation k [RO..][DBC] k = 830 M"1 sec"1
ox 2 ox
Hydrolysis NA k, = 0
Biodegradation k
- max k - 0
b2 ~ Y K b2
s
On Coyote Creek sediment.
See discussion in Part I, Section 5.3 and in text.
Assumes midday sunlight in mid-January.
^Biodegradation was not observed during our enrichment procedures.
10.3.2 Environmental Assessment Using the One-Compartment Modej
The half-lives of dissolved DEC calculated for individual transformation
or removal processes following spills or other acute discharges are listed in
Table 10.3. Although these half-lives generally change from one water body to
another, the half-lives of photolysis and adsorption appear generally shortest.
Sorption half-lives have not been measured, but they are probably at least
comparable to the half-lives for photolysis. The laboratory estimates of the
partition coefficient varied greatly (Section 10.4.4), due to variations in
organic content of the sediments. Consequently, since photolysis is rapid, the
pathway of DEC in natural waters is likely to be extremely sensitive to the
suspended sediment concentration; transformation should dominate in clear waters,
and sorption should dominate in turbid waters with an expected range of 50-82%
sorption onto suspended solids in most natural waters. Transformation of the
sorbed material is expected to be extremely slow, and accumulation large, unless
biodegradation in natural environments is significantly faster than was observed
in our experiments.
169
-------
TABLE 10.3 TRANSFORMATION AND TRANSPORT OF 7H-DIBENZO[c,g]CARBAZOLE
PREDICTED BY THE ONE-COMPARTMENT MODEL
Process
Photolysis, half-life (hr)a
Oxidation, half-life (hr)
River
1.0
>700
Eutrophic
pond
1.5
>700
Eutrophic
lake
1.5
>700
Oligotrophic
lake
0.5
>700
Volatilization, half-life
(hr)
Hydrolysis, half-life (hr)
Biodegradation, half-life
15,000
37,000
73,000
73,000
(hr)
Half-life for all processes
except dilution (hr)
Half-life for all processes
including dilution (hr)
Amount DBC sorbed (yg m~3)
Percentage DBC sorbed
Very long
1.0
0.36
20
66%
Very long
1.5
1.5
60
82%
Very long
1.5
1.5
10
50%
Very long
0.5
0.5
10
50%
Estimates assume winter insolation.
No acclimated cultures were obtained during the screening studies.
"Assumes 10 ng ml"1 (10 ppb) of DBC in aqueous phase and sorption partition
coefficient of 2 x 10\
10.3.3 Persistence
The persistence of DBC following acute discharges is expected to be
largely a function of the dilution rate in large, fast flowing streams. Per-
sistence in lakes and ponds should be dominated by rates of transformation within
the sediments. However, very low concentrations can be expected indefinitely
in all natural waters if biodegradation is indeed very slow in natural sediments,
as is implied by our failure to obtain microbial cultures capable of transform-
ing DBC. If so, desorption from the sediments will maintain low concentrations
of DBC in solution indefinitely, the significance of which will probably depend
on biological concentration processes, which cannot be appraised with our data.
10.3.4 Mass and Concentration Distributions Calculated Using Computer Models
The distributions of mass and concentration of DBC expected at steady
state within each of four types of water bodies, as calculated with the aid of
the computer model, are summarized in Table 10.4. The pseudo-first-order rate
constants used in these simulations are presented in Appendix A,
170
-------
TABLE 10.4. DISTRIBUTION OF 7H-DIBENZO[c,g]CARBAZOLE IN VARIOUS AQUATIC SYSTEMS
AT STEADY STATE (input concentrations of 10 pg ml"* 7H-dibenzo[c,g]carbazole)
Pond
Mass Cone. ,
(ke) (us ml"1)
Compartment 1
(surface water)
Solution 2.0 x 10~* 1 x 10-2
Suspended solids 1.1 x 10~3 18
Compartment 2
(surface water)
Solution — —
Suspended solids — —
Compartment 3
(surface water)
Solution — —
Suspended solids —
Compartment 5
(bottom water)
Solution —
Suspended solids — —
Compartments 7-9
(sediment)3
Solution 2.5 x 10-6 o.Ol
Solids 1.4 x 10-1 2 x Ifl2
Total mass 1.4 x 10
River
Mass
(ke)
8.4 x 10"1
1.7
7.4 x 10- *
1.5
6.4 x ID"1
1.3
__
—
1.8 x 10-1
1 x 103
1 x 103
Eutrophic lake
Cone. Mass
(UK ml'1) (ke)
2.8 7.2
5.5 x 104 7.2
2.5 1.5
4.9 x 10* 1.5
2.13 2.5
4.3 x 10* 2.5
1.2
1.2
2.5 3.7
4.9 x 10* 1.1
1.1
x 10~3
x 10~3
x 10-3
x 10-3
x 10-5
x 10~5
x 10-5
x 10~5
x 10-5
Cone. ,
(UK ml'1)
-
2.9 x 10~2
5.8 x 102
6.0 x 10~4
12
1 x 10-4
2
4.3 x ID"5
1 x 10"5
6.1 x ID"3
81
Oligo trophic lake
Mass
(ke)
3.9
3.8
5.1
5.1
5.6
5.6
1.9
1.9
1.9
6.0
6.0
x 10'4
x lO-^
x 10-5
x 10-5
x 10-7
x 10- 7
x 10-7
x 10~ 7
x 10-2
x 10-2
x 10~2
Cone. _i
-------
The sediments clearly dominate the distribution of DEC in each of
these simulations, with virtually all the DEC in the bottom sediments and
less than 1% in the suspended fraction. The amounts of DEC sorbed to the
suspended solids and in solution are equal in the aqueous compartments of
the lake simulations, but the suspended solids dominate in the aqueous com-
partments of the pond and river simulations. These predictions contrast
rather sharply with the results of the one-compartment, acute pollution
analyses, given in Section 10.3.2, which ignored interactions with the
bottom sediments and the continual replenishment of the suspended solids by
exchange with the sediments.
The expected steady-state concentrations of DEC in solution are roughly
10~3 to 10~6 ng ml~l, which is roughly one thousand to one million times less
than the input concentrations (Table 10.4). The highest steady-state concen-
trations of DEC in solution occurred in the river, with 100-fold lower concen-
trations in the eutrophic pond and lake simulations, and a 1000-fold lower
concentration in the oligotrophic lake simulation. Within the lake simulations,
concentrations of dissolved DEC decrease by 100- to 1000-fold with increasing
distance from the source. Concentrations in the river simulation decline by
25% in the first 3 kilometers. Concentrations in the sediments are roughly
equivalent to the concentration of DEC in the inflowing waters in the river
simulation, and roughly one-tenth the concentration of the inflowing rates in
the lake simulations.
The buildup and decline of DEC in ihe pnnd simulation, shown in Figure
10.1, shows the dynamics of DEC accumulation following initiation and termina-
tion of discharge. Steady-state conditions are reached within two days for
the solution and suspended solid phases, but the concentration of DEC sediments
in Figure 10.1 appears to rise until the ninth day (216 hours), when it declines
due to desorption. Desorption then provides a lower limit to the decreases in
the concentration of DEC in the dissolved and suspended solids fractions, which
began with the cessation of discharge after 150 hours, giving a relatively
long-lived loading of roughly 2.5 ng ml"1. The half-life of DEC following a
spill, estimated from the rate of decline, is much longer than that predicted
by the one-compartment model (7 versus 1.5 hours) since the DEC in solution
in the computer simulation is continually replenished from the sediments.
Half-lives of DEC during actual spills should lie between these two limiting
values.
Similar patterns prevail in the river simulation (Figure 10.2), ex-
cept that steady-state conditions are obtained far more rapidly (within hours
instead of days).
Buildup and declines in the lake simulations (Figures 10.3 and 10.4)
are rapid, as in the pond simulation, showing the effects of the high parti-
tion coefficient in one strong localization of DEC near the source.
10.3.5 Discussion
The principal uncertainties in the preceding analyses lie in the
estimates of partition coefficients and the extrapolations of the laboratory
172
-------
3 x 10
0>
UJ
O
CQ
cc
1
1
111
CO
I
cc
m
o
10'1 -
50
100 ISO
TIME — hours
200
250
FIGURE 10.1 PERSISTENCE OF 7H-DIBENZO[c.g] CARBAZOLE IN A
PARTIALLY MIXED TWO-COMPARTMENT POND SYSTEM
173
-------
SEDIMENTS
SOLUTION
10
456
TIME - hours
FIGURE 10.2 PERSISTENCE OF 7H-DIBEN20 [ c,g] CARBAZOLE IN A
PARTIALLY MIXED RIVER SYSTEM
174
-------
3 x 10
-2
10
-2
10"
1 _u_J I L ij
COMPARTMENT
_ /
/
|7
1
10'
i
01
_l
O
N
CO
c
<
O
O
N
Z
Ul
—
Q
X
z
O
Se
DC
ui
U
1
10-
10-
10
-7
10"9
SEDIMENTS
SOLUTION
DISCHARGE
STOPPED
10'
,-10
100
200
330 400
TIME - hours
500
600
700720
FIGURE 10.3 PERSISTENCE OF 7H-DIBENZO [c.g ] CARBAZOLE IN A
EUTROPHIC LAKE
175
-------
2 x 10"
10"
ID"
KT
I
ui
O
N
CD
CC
O
N
UI
CO
D
p»
U.
O
g
i-
(C
u>
u
o
o
E/
1
10-
JU, 10
'7
10-
10
•a
io-11
10
.-12
I I
COMPARTMENT
/
~l /
/
I
DISCHARGE
STOPPED
SEDIMENTS
SOLUTION
100
200
300 400
TIME - hours
500
600
700 720
FIGURE 10.4 PERSINSTENCE OF 7H-DIBENZO [c,g] CARBAZOLE
IN AN OLIGOTROPHIC LAKE
176
-------
microbial data to the field. Even though our assumption that biological
transformation is unimportant may be wrong, it is probable that most of the
DBG released to natural waters will be rapidly sorbed by suspended particulates
because of the size of the sorption partition coefficient and the rapidity
with which sorption typically occurs. Hence, further research to reduce the
uncertainty in our analysis seems unwarranted unless the transformation
products of DBG are themselves of interest, or the predicted concentrations
are found to be carcinogenic.
10.4 PHYSICAL PROPERTIES
10.4.1 Solubility in Water
The solubility of DBG in water at room temperature (about 22 ± 2°C)
was measured using the methods described by Haque and Schmedding (1975). The
average solubility was 63 ± 3 ng ml"1 [2.4 (± 0.1) x 10~7 M], based on replicate
determinations of solutions that were subjected to long centrifugation (30 min
at 15,000 rpm) until a constant value for the solubility was obtained.
10.4.2 Absorption Spectrum
DBG absorbs light strongly in the solar region to 380 nm, weakly to
390 nm. The pH of a solution of DBG in 5% acetonitrile/95% water was adjusted
with HC1 (Table 10.5). The absorption spectra of DBG in 50% acetonitrile/
50% water was also measured from 295 to 650 nm at pH 3.3, 5.2, and 7.2 (Table
10.6). Care was taken to assure that the DEC was in solution and not a sus-
pension. No particulate could be observed by light scattered by a collimated
beam of green light (the green filter absorbed light below about 450 nm and
eliminated the strong fluorescence observed in a beam of white light that ob-
scured the scattered light).
These data show that pH has no effect on the absorption spectrum in the
pH range of environmental interest. However, there is a solvent effect on the
absorption coefficient. At 360 nm, it is 12,400 cm"1 M"1 in 5% acetonitrile
and 17,400 cm"1 M"1 in 50% acetonitrile. This solvent effect was not examined
because it would have been necessary to reduce the concentration of DBG to be
sure that it remained in solution, and our spectrophotometer does not have the
required sensitivity to make the measurement with good precision. Also, the
data reported in Table 10.5 in 5% acetonitrile are adequate for estimating the
photolysis rate within the precision required for our predictions of the photol-
ysis half-lives. These results do point out the necessity of making the ab-
sorption spectra measurements with the minimum amount of cosolvent.
10.4.3 Volatilization Rate
The volatilization rate of DBG was measured using the method of Hill
et al. (1976), which is described in detail in Part I and Appendix B. Vola-
tilization of DBG in aqueous solution was undetectable, within experimental
error, at an average oxygen reaeration rate of 5.7 (± 1.6) hr"1 over a 4-day
177
-------
TABLE 10.5. ABSORPTION SPECTRUM OF 7H-DIBENZO[c,g]CARBAZOLE
IN 5% ACETONITRILE/95% WATER AT pH 4.5a
Center of Average
wavelength interval*3 absorption coefficient0
(nm) (M~J cm"1)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330.0
340.0
350.0
360.0
370.0
380.0
390.0
400.0
16,500
15,900
12,300
8,760
6,480
4,990
4,340
4,070
3,930
3,960
4,260
5,830
9,220
11,000
12,700
7,890
770
10
0
7H-Dibenzo[c,g]carbazole concentration was 985 ng ml"1
(3.69 x 10-6 M).
The wavelength intervals are given in Appendix B, Table B.1,
Q
The absorption coefficients at 313.0 and 366.0 nm are
4,210 and 13,000 M"1 cm"1, respectively.
178
-------
TABLE 10.6. ABSORPTION SPECTRUM OF 7H-DIBENZO[c,g]CARBAZOLE
IN 50% ACETONITRILE/50% WATERa
Center of
wavelength interval"
(nm)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330
340
350
360
370
380
390
Absorption coefficient0
(M"1 cm-1) i
pH 3.3
20,400
21,800
18,200
13,400
9,500
6,630
:
! 5,380
j 4,910
4,740
4,500
5,090
7,180
12,000
14,100
17,100
11,500
619
0
pH 5.2
21,000
21,800
18,000
12,800
8,860
6,510
5,410
4,850
4,650
4,500
5,060
6,970
12,000
14,200
17,100
11,300
590
0
pH 7.2
21,200
22,500
19,200
14,400
9,910
6,830
5,670
5,170
4,940
4,850
5,170
6,970
12,200
14,700
18,000
12,100
—
—
Average
ibsorption coefficient
(M"1 cm"1)
20,900
22,000
18,500
13,500
9,420
6,660
5,490
4,980
4,780
4,620
5,110
7,040
12,100
14,300
17,400
11,600
604
0
-.-...
a7H-Dibenzo[c,g]carbazole concentration was 9.20 yg ml (3.44 x 10"^ M).
The wavelength intervals are given in Appendix B, Table B.I.
°The absorption coefficients at 313.0 and 366.0 nm are 5,420 and
19,300 M"-1 cm"1, respectively.
period. The average DBG concentrations at 0, 21, and 96 hours were 15.1
(± 0.8), 15.5 (± 1.5), and 14.3 (± 0.5) ng ml"1. We assumed that a 5% loss
of DBC in 96 hours is the maximum loss that could have occurred without being
detected, within experimental error. The upper bound for the kS/kP ratio was
estimated as:
DBC volatilization rate:
Ratio:
= ln(1
5.3 x 10~* hr"
k^/kj < 9.4 x 10"=
179
-------
This is an extremely low value of the ratio. We concluded that volatilization
of DEC would not be a significant pathway in the environment. Therefore, no
further volatilization experiments with DBC were performed;
10.4.4 Sorption on Clay and Sediments
The sorption partition coefficients of DBC were measure on three
sediments: Des Moines River sediment, Coyote Creek sediment, and Searsville
Pond sediment. Care was taken to assure that the samples were protected from
light at all times to minimize photo-oxidation of the DBC. The initial
concentration of DBC in each sample was always less than the saturation
concentration of 63 ng ml"1 (2.4 x 10~7 M). Each experiment consisted of re-
plicate samples at two DBC concentrations and two sediment loadings, plus
blanks that contained only DBC or sediment. Triplicate analyses were made on
each flask. Periodic checks were made for sorption of DBC onto the glassware;
since none of the glassware extracts contained DBC, we presumed that less than
5% of the DBC was sorbed onto the glassware. The stock solutions of DBC in
water were centrifuged at 7500 rpm for 1 hour to remove any suspended DBC
before the sediment/DBC mixtures were prepared.
The data, summarized in Table 10.7, were fitted to a Freundlich isotherm
with n = 1 (equation B.6 ). Graphs of the isotherm data, calculated on the
basis of the supernatant concentrations of DBC, are shown in Figure 10.5 with
the partition coefficients (K ). The partition coefficients were calculated two
ways. First, we used only the measured values of the supernatant DBC concen-
trations in the blanks and the sediment/DBC mixtures. The calculation of the
partition coefficient was based on the results of the analysis of the super-
natant at equilibrium. The amount of DBC sorbed on the sediments was calculated
by the difference between the amount that was added and the amount that was
found in the supernatant. These data are also plotted in Figure 10.5, Our
opinion is that these values are the most valid (the estimated error is + 10%) .
The sediments were also extracted to determine the amount of sorbed
DBC. From these data, partition coefficients and material balances were cal-
culated. However, the data are much more scattered than those from the super-
natant, and a number of experimental difficulties were encountered. The sediment
in one flask at each level of sediment and DBC was extracted with ethyl acetate,
and the extract was concentrated and analyzed for DBC. These concentration
data are less reliable than the supernatant analyses due to analytical problems
encountered (the estimated error is + 25%).
We experienced several analytical difficulties with the measurement
of the DBC isotherms, but we did not attempt to repeat the measurements. The
analyses of the supernatant DBC, after centrifugation of the samples for 20
min at 12,000 rpm, were made by direct injection of 0.1 ml (Searsville and Des
Moines) or 1 ml (Coyote Creek) of the aqueous solution directly onto the LC.
After the isotherms on the Des Moines and Searsville sediments were complete,
we discovered that some of the DBC was held up on the sample loop while the
loop was being filled. Dilution of the aqueous samples with 25% by volume of
acetonitrile eliminated this problem. The Coyote Creek isotherm data were
analyzed by this procedure, and the standard deviations calculated from the
180
-------
TABLE 10.7. SORPTION OF 7H-DIBENZO[c,g]CARBAZOLE ON SEDIMENTS
Tot- jil
organic
carbon
Sediment (Z)
DCS Molnes River 0.6
sorption
Coyote Creek 1.4
)-> sorption
03
Searsville Pond 3.8
sorption
Sediment
concentration
(us ml"1)
84.2
168.0
84.2
168.0
35.8
71.7
35.8
71.7
27.5
82.5
27.5
82.5
DBC
Partition coefficient, K (x 10~3)d
concentration concentration
in supernatant3 on sedlmentb Recovery'
(ng ml-1) (ng nT1 x 10" ') (%)
6.7 ± 0.4
3.6 ± 1.4
2.7 ± 0,6
1.3 ± 0.3
15.2 ± 1.4
8.5 ± 0.5
7.9 ± 0,5
4.8 ± 0,5
13.0 ± 0.4
8.6 ± 1.0
6.6 ± 1.4
3.7 ± 0.1
95 ± 13
52 ± 2
38 ± 5
30 ± 4
198 ± 18
89 ± 6
108 ± 1
62 ± 4
159 ± 9
65 + 9
129 ± 30
39 + 10
65
52
47
50
94
63
104
83
77
62
69
47
LLS
: S only S amd S
w w
a-0 a t 0 a - 0
32.6 ± 5.9 19.5 ± 7.4 14 ± 2.6
a - 56 ± 27
18.5 ± 2.8 14.2 ± 4.5 15.0 ± 28
a • 46 ± 44
27.6 ± 5.1 17.9 ± 1.1 12 ± 7
a - 96 ± 93
NLLS
8
a i« 0 "w only "s anj "w
13 ± 6 -- 14.8 ± 1.4
a - 9 ± 23
-13 ± 44 20.5 ± 2.0 12.2 ± 1.4
a - 200 ± 28
11 ± 28 9.2 ± 1.6
a - 12 ± 240
'Concentration measured In supernatant with population standard deviation.
Concentration measured on sediment with population standard deviation.
°Based on blank flasks to determine total material added to each flask.
LLS - linear least squares) NLLS • nonlinear least squares; S • concentration in supernatant;
See Appendix B for description of regressions. Limits are 95% confidence limits.
S - concentration on sediment.
-------
00
ro
ui
5
5
w
V)
DC
ai
a.
Q
iu
m
a:
o
o
m
a
350
300 —
260 —
200 —
O
• OES MOINES RIVER SEDIMENT. Kp - 32,600
O COYOTE CREEK SEDIMENT, Kp - 18,500
A SEARSVILLE POND SEDIMENT, Kp - 27,600
5 10 15
CONCENTRATION OF DBC IN SUPERNATANT AT EQUILIBRIUM - ng ml"1 (ppb)
20
FIGURE 10.5. SORPTION ISOTHERMS OF 7H-DIBENZO[c,g] CARBAZOLE
-------
triplicate analyses are a good estimate of the error associated with these
measurements. The error introduced into the Des Moines and Searsville data was
about 30%. However, since all flasks for all three sediments were made from
the same stock solution of DEC in water, we could correct all the supernatant
data because we had analyzed blanks containing only DEC (along with each iso-
therm measurement). These corrections have been made on the data reported in
Table 10.7.
To measure recovery, the sediment from one flask at each level was
shaken with ethyl acetate to extract the sorbed DEC. The ethyl acetate extracts
from the Coyote Creek sediments were analyzed by injecting 0.1 ml of extract
onto the LC. The volume of ethyl acetate extracts injected during the analysis
of the Des Moines and Searsville sediment was 1 ml. This sample volume injected
was too large, and much of the DEC was eluted with the solvent front. However,
some DBG was retained and eluted at the usual elution volume. The overall
sensitivity of the method was reduced by about a factor of 10, with a corre-
sponding decrease in precision and accuracy. However, calibration curves could
be constructed from data obtained with standards that had been prepared in
water-saturated ethyl acetate, and this curve was used to calculate 1'ie concen-
tration of DEC in the sediment extracts. Hence, the amount of DEC sorbed on
the sediments could be estimated within about ± 20%. Recoveries ranging from
47% to 104% were obtained. Values of K were calculated from the measured
supernatant and sorbed DEC concentrations. These data are also reported in
Table 10.7. However, we have less confidence in the values of K and we prefer
to use the values calculated from the supernatant concentrations calculated on
the basis of these data only.
In spite of the analytical difficulties, a reasonable range of K
values for DEC is 15,000 to 40,000. The K for Coyote Creek, which is tRe most
reliable experiment, gave K = 20,500 for the nonlinear least squares fit to
the data for the supernatant data, in good agreement with the value of 20,000
used in the environmental assessment.
10.4.5 Biosorption
Biosorption studies were conducted with DEC solutions and 0.05% potas-
sium phosphate buffer, pH 7.0. Corrections were made for sorption of DEC on
glass walls by determining the recovery of DEC from the centrifuge tubes when
the cell suspensions were poured out of the control tubes.
Because of the low levels of DEC that were being used, each suspension
of the four test bacteria (Azotobacter beijerinkii ATCC 19366, Bacillus cereus
ATCC 11778, Escherichia coli ATCC 9637, and Serratia marcescens ATCC 13880)
was reduced to an optical density of 0.4. These suspensions were combined
immediately before the experiment was initiated. The results from the one-hour
sorption and three-hour sorption studies are presented in Table 10.8.
The adjustments for sorption of DEC on glassware in Table 10.8 were
4.6%, 3.7%, and 6.9% for sorption studies conducted in experiment 1 with
22.3 ng DEC ml"1 and in experiment 2 with 20 and 6.2 ng DEC ml"1, respectively.
The corresponding adjustment for glassware sorption values in the desorption
183
-------
TABLE 10.8. 7H-DIBENZO[c,g]CARBAZOLE SORPTION
AND DESORPTION WITH A MIXED BACTERIAL POPULATION
Experiment
1
2
2
initial DEC
concentration
(ng ml"1)
22.3
20.0
6.2
Sorption(S) or
desorption (D)
S
D
S
D
S
D
Sorption
coefficient
(x 10-")
9.0 ± 2.2
10.7 ± 2.3
9.2 ± 1.3
6.4 ± 1.1
8.0 ± 1.2
6.9 ± 1.0
aDry weight of bacterial mixture equivalent to 178 and 162 mg liter
liter~l in experiments 1 and 2, respectively.
studies were 7.7%, 6.4%, and 5.7%, respectively. The quantities of DBC sorbed
on the glassware are very close to the values for DBC in the supernatants and
present a problem in definitive interpretations of the sorption coefficients
beyond the fact that DBC is highly sorbed by biological and inorganic materials.
These corrections for sorption of DBC on glassware could be in error
because some bacterial cells may have remained on the walls of the centrifuge
tubes. In preliminary experiments with BaP, it was observed that the presence
of microorganisms in a glass vessel resulted in lower residual BaP values on
the glassware than in the controls with no microorganisms. However, these
results were complicated by the fact that bacteria also sorbed BaP, which low-
ered the available concentration of BaP for sorption on glassware. As was
observed with BaP, DBC was readily sorbed by clean glass surfaces.
This problem could best be resolved by a series of studies using
isotopically labeled DBC and bacteria. With the data we have obtained, the
sorption coefficient values would have been changed in the range of 5%—a
value less than our standard deviations. It would appear that this Afas a
futile exercise, but in view of the high sorbability of DBC on glassware and
the low levels of DBC that were to be studied, this adjustment procedure was
warranted.
The high sorption coefficient of bacterial cells for DBC would indicate
that, in activated sludge sewage treatment facilities, most of the DBC would
probably be removed on the sludge. The fate of DBC in flowing streams or
stationary waters is more difficult to predict. The high degree of sorption
on such a relatively inert surface as glass as well as natural sediments and
the high probability that DBC in bacterial mass is more dependent on sorption
phenomenon than on surface sorption, probably indicate that, in aquatic systems
with low microbial counts, sorption onto suspended sediments may be the prime
initial factor in decreasing DBC levels in waters. Subsequently, degradation
can take place by microbial and/or chemical/physical phenomena acting on free
or available DBC in equilibrium with the sorbed DBC.
184
-------
10.5 CHEMICAL TRANSFORMATION
10.5.1 Photolysis Rate
The photolysis of DBC is rapid and is unaffected by oxygen. In sun-
light the half-lives are less than a day. Since the solubility of DBC in
pure water is 63 ng ml"1, we used 0.1% acetonitrile as a cosolvent to enhance
the solubility of DBC in waters from the natural sources. Data for photolysis
of DBC in sunlight and at 366 nm are given in Table 10.9. Good first-order
kinetic behavior was found for photolyses at 366 nm carried out to beyond two
half-lives of DBC. The quantum yield for DBC phototransformation in pure
water is (2.8 ± 0.2) x 10"3.
Photolysis experiments carried out at 366 and 313 nm on 9.1 ng ml"1
DBC solutions that had been purged with nitrogen showed that oxygen had no
effect on the photolysis rate. One experiment at 366 nm found, however, that
DBC was stable to photolysis when pure acetonitrile (no water) was the reaction
solvent. Therefore water, but not oxygen, is involved in the rate-determining
step, and the reaction is probably some type of photohydration process like
that found for carbazole (see Sections 9 and 10.5.4).
Photolyses of 9 ng ml"1 DBC solution were also carried out at 313 nm but
these experiments provided an enigma. In 0.1% acetonitrile solution, a 15%
loss of DBC was found in the initial 15-minute photoperiod, but no further loss
of DBC was found after 45 additional minutes irradiation. A similar observa-
tion was made for 10 ng ml"1 DBC in water containing 0.1% methanol, where a
45% loss of DBC was found in the initial 20-minute photoperiod, but no further
loss after 60 additional minutes of irradiation. The DBC concentrations of the
unphotolyzed solutions were confirmed by the direct-injection HPLC analysis,
so the observation is considered a chemical phenomena and not an artifact of
the experiment procedure. The possible error introduced by holdup of DBC in
the sample loop (Section 10.4.4) is no greater than 5%. It is doubtful that
products from the DBC phototransformation could explain the observations either
through chemical or physical (quenching) processes, and at this time we can offer
no explanation for the unusual photolysis behavior at 313 nm.
The data in Table 10.9. show that the photolysis rate of DBC in water
from Lake Tahoe or Searsville Pond is not changed from that in pure water.
The water from Coyote Creek (absorbance at 366 is 0.06 in a 1-cm cell) shows
only a slight retardation of the photolysis rate. However, the experiment in
pure water containing humic acid does show a definite reduction in photolysis
rate. The absorbance of the humic acid solution at 366 nm is 0.63, indicating
that 75% of the incident light is absorbed by the humic acid. Thus, the
screening by humic acid accounts for only a fourfold decrease in rate, and
other processes must also be occurring to explain the twelvefold decrease in
rate observed. Some possible mechanisms are discussed in Section 9.5.1.
The half-life for direct photolysis of DBC in sunlight was calculated
according to the procedure of Zepp and Cline (1977), using the quantum yield of
2.8 x 10~3 measured at 366 and the absorption spectrum (Section 10.4.2). The
UV spectrum of DBC passes through a minimum at about 320 nm. While the experi-
185
-------
TABLE 10.9. RATE CONSTANTS FOR PHOTOLYSIS OF 7H-DIBENZO[c,g]CARBAZOLE
CO
a\
Irradiation
source
366 nm .
366 nm
366 ntn
366 nm
366 nm
366 nm
Noon sunlight,
late January
DEC
concentration^3
Solution3 (ng ml'1)
Pure water
Pure water, N2 purged
Lake Tahoe water
Coyote Creek water
Searsville Pond
water
Humic acid in
pure water
Pure water
9.1
9.1
10.3
9.2
8.2
11.5
11.3
Rate constant
Extent of (k x 103 sec"1)
reaction (%) p
84
84
71
89
78
56
78
1.52 + 0
1.5e
1.42 + 0
1.17 + 0
1.80 + 0
0.122 +
0.518 +
.10c»d
.09
.13
.22
0.005
0.032f
Solution contained 0.1% acetonitrile as cosolvent.
b1.00 yg ml"1 DEC in water = 3.74 x 10"9 M.
£
Standard deviation.
Quantum yield of 2.8 x 10~3 for disappearance of DEC at 366
o
One data point at 84% reaction in nitrogen purged solution.
fHalf-life of 24 minutes.
nm.
-------
ments at 313 nm suggest that the rate of photolysis of DEC below 320 nm may be
minimal, the calculation was carried out for the entire solar spectral region
from 297.5 to 380 nm. Such a procedure is permissible since the major contri-
bution to photolysis will be in the region above 320 nm where the DBG absorbance
and sunlight intensity are more intense than in the lower end of the solar
spectrum (e.g., 290-320 nm).
Figure 10.6 shows the variation in the photolysis half-life of DBG as
a function of the time of day for midsummer and midwinter. The half-life of
51 minutes calculated for midday in midwinter is in good agreement with the
measured half-life of 24 minutes on a clear day in late January. While the
photolysis half-lives do vary with the time of day, Figure 10.6 shows that
photolysis of DBG will be an important process in the environment, with half-
lives of less than a day.
10.5.1 Oxidation Rate
When the susceptibility of DBG to free radical oxidation was examined
using the AA-initiated oxidation screening reaction (see Section 6.3 in Part
I and Appendix B for discussion and procedure), DEC was completely oxidized
after the usual 100-hour reaction time. The experiment was then repeated
using a 9.1 ng ml"1 (3.40 x 10~8 M) DBG solution (0.1% acetonitrile in water)
containing 1.00 x 10"* M AA. The solution was analyzed for DBG at reaction
times of 4.0 and 8.3 hours. A rate constant term (k [R02']) of (2.5 ± 0.4)
x 10~5 M~l sec"1 was calculated from the data. Underxthe reaction conditions
at 50°C, this corresponds to a rate constant kox of 7.5 x 103 M"1 sec"1 (see
Section 6.3 in Part I). At 25°C, k is then 830 M"1 sec"1, and assuming
[R02«] = 10~9 M in the environment,°£he half-life for oxidation of DBG is
about 10 days. The free radical oxidation of DBG by R02* is clearly not
competitive with photolysis or sorption under aquatic environmental conditions.
10.5.3 Hydrolysis Rate
DBG contains no groups that are hydrolyzable. Therefore, no hydrolysis
studies were carried out.
10.5.4 Products from Chemical Transformation
Two primary products were found by HPLC analyses of the photolysis of
DBG in pure water. Both products were separated by HPLC and subjected to mass
spectrometry. The only distinguishing features of the spectrum of the major
product were an apparent molecular ion at m/e 299 and peaks at m/e 270 and
271, corresponding to loss of CHO and CO, respectively from the molecular ion;
this indicates the presence of at least one phenolic group (McLafferty, 1973).
DBG itself has a molecular ion of m/e 267. The spectrum of the major product
is compatible with that of a dihydroxylated DEC; however, no silylated deriv-
ative of this product could be obtained using 0,N-bis(trimethylsilyl)trifluoro-
acetamide.
187
-------
_l
<
I
I
5
7
6
6
8 9
4 3
TIME
10
2
11AM
1 PM
12 NOON
FIGURE 106. SEASONAL AND DAILY VARIATION OF PHOTOLYSIS
HALF-LIFE OF 7H-DIBENZO[c,g]CARBAZOLE
188
-------
It is difficult to rationalize why neither hydroxyl nor amine functions
would be silylated with this reagent. The rate of photolysis of DEC did depend
on water and not on oxygen, but the observed products could depend on a sub-
sequent reaction with oxygen (this possibility was not checked). Unless an
(unlikely) termolecular process is occurring involving two water molecules and
DEC, the product must be formed by initial reaction of water and the photo-
excited DBG, followed by a subsequent rapid reaction of the intermediate with
another water molecule or oxygen. Neither of these events seems likely, and
some additional structure determination appears necessary.
The second product could not be silylated, and gave an apparent molecu-
lar ion of m/e 271 with no characteristic fragmentation pattern to allow further
identification. On the basis of the m/e 271 molecular ion, the second product
could be a tetrahydro-DBC derivative, but such a photoreduction in water in
the presence of oxygen is without precedent. As with the major product, some
additional structure determination for the second product seems necessary.
10.6 BIODEGRADATION
10.6.1 Development of Enrichment Cultures
Enrichment studies were initiated with 1 yg ml"-'- of DEC in 5-liter
liquid volumes in 9-liter, aerated, bottle-fermentors. We examined water
samples from Coyote Creek (eutrophic) and from the aeration effluents from
sewage plants in South San Francisco, and at the Shell Oil Refinery in Martinez,
California. Since there was a good possibility that the Shell Oil Refinery
effluents contained minor discharges of DEC, this source was considered both
as a commercial discharge source and as a sewage treatment plant that possibly
contained organisms capable of biodegrading DEC.
Thus, studies with the Shell Oil Refinery effluents were initiated with
DEC as the sole added carbon source [in addition to the normal components in
the samples, phosphate buffer, and (Nlty^SO^]. Studies of water samples from
all three sources were also initiated with DEC and either carbazole or naph-
thalene (10 yg ml~l) as an additional source of carbon.
After one week of incubation of the 9-liter fermentors, additional
2-liter volumes of fresh samplings from the South San Francisco sewage effluent
and from Coyote Creek were added to the respective bottles to possibly expand
the range of types of organisms in the original samplings. Carbazole
(10 yg ml"-'-) was added once a week for three weeks. Naphthalene (10 yg ml~l)
was added three times a week for three weeks.
During the six weeks of this study, degradation of DEC was not observed
by I^C analysis of ethyl acetate extracts of broths that had been subjected to
three freeze-thaw cycles.
After three and six weeks of incubation of Shell Oil'Refinery and South
San Francisco sewage plant samples, transfers were made from the corresponding
9-liter bottles into shaker flasks containing DEC (1 yg ml~l) plus carbazole
(10 yg ml~l) and DEC (1 yg ml"1) plus naphthalene (10 yg ml"1). From the
189
-------
9-liter bottle fermentor with the Shell Oil Refinery sample in which DEC was
the only added carbon source, transfers were made to shaker flasks containing
basal salts medium with DBC (1 pg ml"!), glucose (50 yg ml"!), and yeast ex-
tract (10 yg ml~l).
All shaker flasks developed good growth. However, after up to 10
days of incubation, the fermented broths did not indicate any loss of DBC in
the shaker flask media.
10.6.2 Biodegradation Kinetics
Since enrichment cultures that degraded DBC were not obtained, biode-
gradation kinetic studies were not performed.
10.6.3 Discussion
It had been previously emphasized (Section 6) that enrichment proce-
dures have a propensity for favoring certain types of organisms. Our inability
to obtain enrichment biodegrading systems cannot be accepted as a positive
indication of recalcitrance of DBC to microbial biodegradation. With the high
sorption coefficients and the low solubility of DBC, it is possible that con-
ventional enrichment procedures do not provide enough "available" DBC for
organisms to utilize this substrate for growth in competition with other organ-
isms that have a more favorable predator-prey relationship in the milieu of the
primary enrichment process.
A personal communication from J. J. Perry (1977) indicated that a number
of fungi could transform benzo[a]pyrene to complex mixtures of organic acids.
A comparable situation could exist in the biodegradation of DBC. However,
our enrichment procedures do not favor the growth of filamentous fungi, and
we would not expect to observe transformation of DBC by fungi. Transformation
by fungi are most likely to be important in flowing waters such as streams and
may be significant in bottom sediments. In summary, DBC biodegradation may
take place slowly under aerobic conditions on sediments, but this pathway was
not investigated in these studies.
190
-------
11. LABORATORY INVESTIGATION OF BENZO[b]THIOPHENE
11.1 SYNOPSIS
The results of the laboratory investigations suggest that benzo [blthio-
phene (BT) is not likely to accumulate in an aqueous environment and that
volatilization is the major environmental pathway of BT in aquatic systems.
However, under some conditions, such as a deep lake where volatilization is
slow because the turbulence is also low, photolysis and perhaps biodegradation
may be significant.
The half-lives of BT in solution, as predicted by the one-compartment
environmental exposure model, and the steady-state concentrations of BT in
solution, on suspended solids, and on sediments near a point source in the
presence of a continuous discharge of 1 Ug ml"1 (1 ppm), as predicted by
the nine-compartment model are given below.
River
Pond
Half-life
in solution
(hour)
0.8
19
Concentration
in solution
ml"1)
0.98
0.024
Concentration
on suspended
solids
(yg g"1)
49
1.16
Concentration
on sediments
(VR g"1)
48
1.16
Eutrophic
lake
19
0.047
2.32
0.60
Oligotrophic
lake
140
0.119
6.00
2.15
The laboratory procedures used to estimate volatilization, photolysis^
and oxidation rates, and the partition coefficients on a natural sediment
and on a mixture of microorganisms were applied to BT. Biodegradation was
a problem because BT degradation occurred only in the presence of naphthalene
by what appeared to be a co-metabolic process. We know of no mathematical
formalism that can be used to express the rate of a co-metabolic process.
The upper limit for the biodegradation rate of BT was estimated to be one-
half that of quinoline (see Section 7.6.2).
t
Predicted by one-compartment model,
Average concentration in sediments,
191
-------
11.2 BACKGROUND
Thiophenes and benzothiophenes are present as minor constituents of
petroleum crudes, coal tars, shale oil, and synthetic fuels. Concentrations
as high as 5.4% have been found in Kuwait crudes. Since sulfur-containing
compounds represent a major processing and air pollution problem, BT and
dibenzothiophene (Part II, Section 12) were selected as examples of the
general class of thiophenes.
The general physical properties of BT that were found in the literature
are given in Table 11.1.
TABLE 11.1. PHYSICAL PROPERTIES OF BENZO[b]THIOPHENE
Structure
Molecular weight
Melting point (°C)
Boiling point (°C)
Solubility a(pg ml"1)
Vapor pressure at 20°C (torr/5
UV spectrum [Amaxnm(log e)] c
* Ll
in cyclohexane
289 (3.36)
291 (3.36)
298 (3.57)
1 pg
ppm)BT = 7.45 x 10~6 M
Measured in this study, Section 11.4.1.
^Estimated from the vapor pressure of naphthalene.
CHaines et al. (1956).
Literature data were not available on the solubility or ultraviolet
absorption spectrum of BT in water, the vapor pressure at room temperature,
the volatility from water, sorption by soil or sediments, chemical oxidation
or photolysis in water, or environmental concentrations. Mackay and Wolkolf
(1973) estimated that the half-life for volatilization of naphthalene from a
typical lake is 2.2 hours. Therefore, we assumed by analogy that volatiliza-
tion of BT might also be fast and that careful volatilization measurements
would probably be required in this study.
Hydrogen peroxide, peracids, and hydroperoxides are reported to oxidize
BT to the sulfoxide and subsequently to the sulfone derivative (Anashkina et
192
-------
al., 1972; Ponec and Prochazka, 1974; Ford and Young, 1965). Oxidation of
BT in alkaline permanganate solution gives o-carboxylic acid and benzenesul-
fonic acid (Erickson, 1954).
BT absorbs strongly at 297 nm (log e 3.5) with a sharp decrease in the
absorption spectrum going to the longer wavelengths (Raines et al., 1956).
In the absence of air, purified neat BT is reported to be unstable to both
daylight and mercury lamp irradiation and to form hydrogen sulfide, hydrogen,
and condensed ring products.
Early studies by Maliyantz (1935), Strawinski (1950, 1951), and Zobell
(1953) involved primarily anaerobic removal of organic sulfur from petroleum
as H2S. Kirshenbaum (1961) used aerobic conditions to convert the sulfur in
petroleum into sulfate. However, these investigations were of a general type
and did not deal with specific types of compounds. More pertinent to the
current studies are those with dibenzothiophene (Yamada et al., 1968; Nakatani
et al., 1968; Kodama et al., 1970, 1973; Gibson, 1975; and Hou and Laskin,
1975), or the studies with 2-thiophenecarboxylic acid and the studies with BT
(Sagardia et al., 1975). Walker and Colwell (1974) and Walker et al. (1976)
reported what may be biodegradations of benzothiophenes. These reports demon-
strated that aerobic bacterial degradative processes were generally complex
and frequently required the presence of other assimilable nutrients; in many
cases, there was incomplete conversion of the substrate.
In most of the above investigations with thiophene compounds, petroleum
hydrocarbons were present and they served as organic nutrients. Sagardia et
al. (1975) used a light petroleum oil. They could not conclude whether the
oil functioned to reduce the amount of benzothiophene in the aqueous phase
and in this way reduced the toxicity of BT, or if it was the classical type
of cometabolism first designated by this term in the studies of Leadbetter
and Foster (1959).
While we anticipated that volatilization might be an important environ-
mental pathway for BT, the literature data were insufficient for estimating
the importance of adsorption by sediments, chemical oxidation, photolysis,
or biodegradation. Therefore, screening studies were conducted for each of
these processes.
11.3 ENVIRONMENTAL ASSESSMENT
11.3.1 Summary of Laboratory Data
The constants obtained in the laboratory investigations of the trans-
formation and transport processes for BT are summarized in Table 11.2.
11.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of BT calculated for transformations and transport by
individual environmental processes are listed in Table 11.3. Although the
half-lives of BT vary fivefold among water bodies, the half-life for vola-
tilization is the shortest in all standing waters and only dilution and
193
-------
sorption are faster in streams. Since volatilization occurs at the air-
water interface and is linearly correlated with the oxygen reaeration rate
constant at low turbulence (Part I), loss of BT from natural water bodies
is expected to be roughly three times slower from sparingly mixed systems,
such as deep stratified lakes, than from fast flowing streams.
TABLE 11.2. SUMMARY OF BENZO[b]THIOPHENE LABORATORY DATA
Process
Sorption equilibrium
Partition coefficient
Sorotion
S = K S
s p w
Rate expression
K = 59 ± 2°
P
= 200
Rate constant at 25°C
Volatilization k [BT]
VL
k = (0.38 ± 0.08)k°
v v
Photolysis
Oxidation
Hydrolysis
Biodegradation
kp [BT]
k [RO
ox
NA
1r —
V
= $ (£Zxex) [BT]
'2'] [BT]
^max
YK
s
v<«
k =
ox
\ = °
•S,"1
.9 ± 0.7) x 10~7 sec'1
83 M"1 sec"1
.6 x 10" 6 ml cell"1 hr"1
On Coyote Creek sediment.
On laboratory mixture of microorganisms.
Q
See discussion in Part I, Section 5.3, and Appendix B.
Assuming 12 hours sunlight per day in late May.
GBT degradation was not observed under our experimental conditions unless
naphthalene was also present. This appeared to be a cometabolic process,
but there is no kinetic expression that can be used to express the rate
for this type of biodegradation. A rate constant one-half that of quino-
line (Section 7.6.2) was used.
The photolysis half-life or BT is 25 to 120 days in surface waters
(Table 11.3), assuming 12 hours of sunlight per day at the latitudes of San
Francisco, Denver, and New York. These half-lives are nearly minimum for
direct photolysis in natural waters during the late spring insolation. The
minimum half-life of six days suggested by the experiments with humic acid
(Section 11.5.1) is probably more realistic. Photolysis in clear lakes and
small streams should be twice as fast as in turbid waters.
194
-------
Oxidation and sorption are relatively unimportant as means of removing
BT in natural waters. The oxidation rate was very slow in laboratory experi-
ments, and extrapolation to environmental conditions suggests that oxidation
in aquatic systems will be unimportant compared with other processes. Simi-
larly, sorption is expected to generally remove less than 1% of the BT from
solution, and even a significant increase of the sediment loading of natural
waters would not change its relative role.
TABLE 11.3. TRANSFORMATION AND TRANSPORT OF BENZO [b] THIOPHENE
PREDICTED BY THE ONE-COMPARTMENT MODEL
Eutrophic Eutrophic Oligotrophic
Process River pond lake lake
Photolysis, half-life (hr) 1,200 2,900 3,500 600
Oxidation, half-life (hr) 105 10s 10s 105
\
Volatilization, 45 230 180 180
half-life (hr)
Hydrolysis NA NA NA NA
Biodegradation, >20 <20 20 Very long
half-life (hr)a
Half-life for all 14 19 19 140
processes, except
dilution (hr)
Half-life for all 0.8 19 19 140
processes, including
dilution (hr)
Amount BT adsorbed (mg m~3)b <5 <15 <2.5 <0.5
Percentage BT sorbed 0.5 1.5 0.2 0.2
rt
Based on the assumption that the biodegradation rate of BT in the presence of
alternative carbon sources will be one-half the biodegradation rate of quino-
line when quinolinr is the only carbon source.
1 pg ml~ BT is assumed in solution.
Although an enrichment culture for degradation of BT was developed in
less than a week when naphthalene was present, enrichment cultures were not
obtained in the absence of naphthalene. This result suggests that cometabo-
lism of BT may be rapid, perhaps as rapid as volatilization.
195
-------
In summary, the half-lives attributable to individual degradation
mechanisms summarized in Table 11.3 suggest that volatilization and perhaps
biodegradation will dominate the removal processes in typical ponds and lakes,
but will be a factor of 10 less important than dilution in short stream seg-
ments. Indeed, the effects of dilution are only important in streams where
the half-lives are 0.5 and 3.3 hours with and without dilution, respectively.
The net effect in all water bodies is a rapid movement of undegraded BT either
downstream or into the atmosphere.
11.3.3 Persistence
BT is unlikely to be persistent in the sense that this term has been
used in pesticide research. Even the extremely low concentrations predicted
in the computer simulations (next section) are improbable since BT biodegrades
so rapidly.
11.3.4 Mass and Concentration Distributions Calculated Using Computer Models
The distributions of mass and concentration of BT expected at steady
state during chronic discharge to each of four types of water bodies, as
calculated with the aid of the computer model, are summarized in Table 11.4.
The pseudo—first-order rate constants used in these simulations are presented
in Appendix A.
The mass retained in each of these systems represents less than 1% of
the BT in the inflowing waters, but of that 1%, 30% to 80% is retained in the
sediments. The highest retention of BT in sediments occurs in the oliogo-
trophic lake (0.78% of the total inflow) where biodegradation is minimal. The
lowest percentage of retention in the sediments occurs in the eutrophic lake
(0.32% of the total inflow). Because of the relatively small mass of sus-
pended particulates, the soluble fraction dominates within the water columns
of each system.
The interesting patterns with respect to the concentration of BT are
best seen in Figures 11.1 through 11.4. Concentrations rapidly attain steady-
state values in the well-mixed waters of the pond and river simulations and in
the surface waters of the lake simulations, but change relatively slowly in
the bottom waters of the lakes. Indeed, the concentrations in the deeper
waters of the oligotrophic lake (Figure 11.4) scarcely change in the three
to four weeks of simulated time following stoppage of discharge. Changes in
sediment concentrations are extremely slow in all water bodies, because of
the long times for scouring of sediments and exchange with the water column.
Concentrations of BT are highest in the suspended solids component of
all the simulations although, for simplicity, this is shown only in Figure
11.1. Changes in concentration of suspended solids parallel changes in the
concentrations of dissolved BT. Except in the river simulation, in which
the sediment concentrations were still increasing at a significant rate when
the simulated discharge was stopped, the concentrations of BT in the sediments
vve as much as twice as high as the concentrations of BT in solution.
196
-------
TABLE 11.4.
DISTRIBUTION OF BENZO[b]THIOPHENE IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations of 1 yig ml"1 ben'zo [b]thiophene
Pond
Mass Cone.
(kg) (ug ml'1)
Compartment 1
(surface water)
Solution 0.49 0.024
Suspended solids 0.007 1.16
Compartment 2
(surface water)
Solution — —
Suspended solids — —
Compartment 3
(surface water)
Solution — —
Suspended solids — —
Compartment 5
(bottom water)
Solution
Suspended solids —
Compartments 7-9
(sediment)3
Solution 0.006 0.024
Solids 0.78 1.16
Total mass 1.28
River Eutrophic lake
Mass
(kg)
294
1
289
1.44
285
1.42
—
—
0.72
97.2
969.8
Cone. Mass
(ug ml'1) (kg)
0.98 18.9
49 0.029
0.96 10.6
48 0.026
0.95 0.548
47.5 0.0013
0.242
0.0006
0.96 0.105
48 8.2
31.65
Cone.
(yg ml~l)
- OT047
2.32
0.004
0.20
0.002
0.10
9.7 x 10~5
0.0048
0.012
0.60
Oligotrophic lake
Mass
(kg)
29.9
0.075
120
0.299
10.7
0.0267
13.0
0.0326
0.0375
29.2
203.38
Cone.
(Ug ml-1)
0.119
6.00
0.0478
2.39
0.0428
2.14
0.0052
0.260
0.0430
2.15
aThe amounts given for solid and solution phases in the sediment compartments are estimated from the
adsorption partition coefficient for suspended solids and may be overestimated because it was assumed
that biodegradation of sorbed material does not occur.
-------
VO
03
I I I I I
50
100 150
TIME - hours
200
250
FIGURE 11.1 PERSISTENCE OF BENZO[b]THIOPHENE IN A PARTIALLY MIXED
TWO-COMPARTMENT POND SYSTEM
-------
FIGURE 11.2 PERSISTENCE OF BENZO[b]THIOPHENE IN A PARTIALLY
MIXED RIVER SYSTEM
199
-------
2 x 10
01
3.
I
UJ
Z
UJ
I
Q.
o
X
o
N
01
CQ
<
-------
4 x 10
10
-1
01
a.
I 10-3
10-
UJ
I
a.
O
o
N
z
UJ
CO
2 ID'5
z
UJ
o
z
o
10-6
10
-7
COMPARTMENT
1
10-2
SOLUTION
SEDIMENTS
I
I
I
100
200 300 400
TIME - hours
500
600
700720
FIGURE 11.4 PERSISTENCE OF BENZO[b]THIOPHENE IN A PARTIALLY
MIXED OLIGOTROPHIC LAKE
201
-------
As expected, the highest concentrations both iu solids and in solution
occur near the sources (compartment 1 in each simulation), where values range
from about 10~2 yg ml"1 in the pond simulation to roughly 0.9 yg ml"1 in the
river simulation when the concentration in the waste stream is 1 ug ml"1.
Concentrations on solids and in solution in the lake simulations tend to be
noticeably less in the deeper portions of the lake, although this is least
pronounced in the oliogotrophic lake.
11.3.5 Discussion
The significance of the concentrations simulated in this study cannot
be readily appraised with the available toxicological data; however, thiophene
and by analogy BT, can be expected to be roughly equivalent to a p_-cresol
with respect to toxicity (McKee and Wolf, 1963). If so, some risk to aquatic
organisms might be expected near discharges under the conditions simulated,
but the risk to these organisms would appear generally to be low because of
the rapid loss of BT from solution.
11.4 PHYSICAL PROPERTIES
11.4.1 Solubility in Water
The solubility of BT in water at 20.0°C was determined for five sam-
ples that were at room temperature before immersion and for six samples
that had been warmed to 32°C before equilibration at 20.0°C. The solubility
if the room temperature samples was 127.3 ± 2.5 (standard deviation) ug ml"1,
.and the solubility of the warmed samples was 132.9 ± 4.6 yg ml"1. The average
solubility for all samples was 130.3 ± 4.6 yg ml"1 (9,7 x lO"*1 M). Thus, BT
is slightly soluble in water (about five times as soluble as naphthalene).
LI.4.2 Absorption Spectrum
Preliminary measurements showed that the UV absorption spectrum of
BT in water is independent of pH in the range found in the environment (pH
5 to 9 was investigated). The absorption coefficients at wavelength inter-
ils above 297.5 nm in distilled, deionized water at pH 6.6 (nonadjusted)
ire reported in Table 11.5.
11.4.3 Volatilization Rate
The volatilization of BT was measured using the method of Hill et al.
^1976), which is described i" Appendix B, Section B.I.3. Table 11.6
contains the BT volatilization data obtained at seven different oxygen reaera-
tion rates. The BT volatilization rate constant k^T is plotted versus the
oxygen reaeration rate k~ in Figure 11.5.
202
-------
TABLE 11.5. ABSORPTION SPECTRUM OF BENZO[b]THIOPHENE
IN WATER AT pH 6.6?
Center of
wavelength
interval
(nm)
297.5
300.0
302.5
Q05.0
p?.5
310.0
312.5°
315.0
Absorption
coefficient
(1 mole"1 cm"1)
1793
395
130
30
13
7
3
0
JBenzo[b]thiophene concentration is 18.7 yg ml"1 (1.39 x lO"4*
in a 1-cm cell.
3The wavelength intervals are given in Appendix B, Table B.I.
M)
"The absorption coefficient at 313.0 nm is 1.
In Part I, Section 5.3, it was shown that at low reaeration rates, the
liquid film resistance is the rate controlling factor. Under these conditions,
it was shown that the ratio k^ /k^ is equal to the ratio of the molecular
diffusion constants D or the inverse of the ratio of the molecular diameters
(equation 11.1).
BT
v
= !BT = ^O_
D0 dBT
(11.1)
203
-------
1.2
1.0
I-
cc
O
K
N 0.6
13
O
oo 0.4
0.2
95% CONFIDENCE LIMIT
OF CORRELATION
LEAST SQUARES
FIT OF LOW
REAERATION
RATE DATA
BT
= 0.38 > 0.08
• I
I •
I • I Bars Represent 95% Confidence Limit of Estimates of k
k° — OXYGEN REAERATION RATE — hr"'
SA-4396-3
FIGURE 11.5 VOLATILIZATION OF BENZO[b]THIOPHENE FROM AQUEOUS SOLUTION
-------
TABLE 11.6. VOLATILIZATION RATE DATA FOR BENZO[b]THIOPHENE
Oxygen reaeration
rate
BT volatilization
rate
BT .. 0
0.062
0.26
0.40
1.16
1.4
4.<
8
7
i
7.19
± 0.
± 0.
± 0.
± 0.
± 0.
± o.
± o.
002a
01
03
03
09
07
11
0
0
0
0
0
0
1
.087
.04
.21
.42
.57
.70
.11
+
+
+
+
+
+
+
0.
0.
0.
0.
0.
0.
0.
a
001
004
002
015
003
007
015
1.
0.
0.
0.
0.
0.
0.
40 ±
15 ±
53 ±
36 ±
38 ±
14 ±
15 ±
a
0.06
0.02
0.04
0.02
0.02
0.004
0.008
Standard deviation .
A value of d for BT may be estimated from the kinetic theory of gases (Present,
1958), in which the molecular volume (based on hard spheres) of a compound is
between 1/3 to 1/2 of the critical volume. Using Lydersen's critical property
increments (Reid and Sherwood, 1966), we estimate the critical volume to be
379 ml mole"1. From this critical volume, the molecular diameter is estimated
to be from 7.4 to 8.4 A. Since the molecular diameter of oxygen is 2.98 A
(Tsivoglou, 1965), the expected ratio from equation (11.1) falls between 0.40
and 0.35. The experimental ratio of 0.38, obtained from a least squares
regression of the k^T/k^ ratio for experimental data obtained when k^ is less
than 1.5 hr"1 (see Figure 11.5), is in good agreement with the theoretical
estimate.
At higher reaeration rates, the gas-film resistance becomes important.
The data in Table 11.6 include only two high reaeration points. The experimental
error of these points is only a few percent, and it is unlikely that the two
high reaeration points also fall on the line that fits the data obtained at
reaeration rates less than 1.5 hr"1 (see Figure 11.5). However, since there
are only two data points, we are reluctant to estimate a value for
It is therefore difficult to determine whether the limiting equation
v
BT
Y
dBTH^
(11.2)
205
-------
kBT
V
k°
V
n -„ 1.3 x 10*
°'38 3 x 107
where H is the Henry's law constant, applies for the data. If this limiting
form does apply to the data, the two points suggest that the Henry's law
constant for BT is less than the Henry's law constant for oxygen, H° (3 x 107
torr). The value of the Henry's law constant for BT, HBT, can be estimated
by using equations (4) and (5) of Mackay and Wolkolf (1973). Using the
measured solubility of 130 yg ml"1 and assuming that the vapor pressure of
BT is the same as naphthalene (0.23 torr at 20°C), we can calculate that the
Henry's law constant for BT is 1.3 x 104 torr (230 torr M"1). Therefore, at
high reaeration, the ratio of k^T/k^ should be
= 2 x 10-* (11.3)
The ratio for the two high reaeration rate data points is about 0.15. The
cause of the lack of agreement is probably that the volatilization rate
ratio is in an intermediate turbulence range and should be expressed by a
more complex form (see Part I, Section 5.3). It is difficult to use this
expression because we do not have a good way to estimate either the diffusi-
vity of BT or the film thickness. It is also possible that the estimate of
HBT is incorrect, but it is not likely to be underestimated by three orders
of magnitude.
11.4.4 Sorption on Clay and Sediments
Two sorption isotherms were measured with a calcium montmorillonite
clay and two with a natural sediment collected from Coyote Creek, San Jose,
California. The isotherms were measured at two sediment or clay loadings
and with initial BT concentrations of about 4 and 8 yg ml~x. Precautions
were taken to minimize exposure to light and volatilization. The results
of these experiments are summarized in Table 11.7.
Very little sorption of BT took place on the calcium montmorillonite
clay. A small but statistically significant difference existed between the
concentration of BT in the blanks and the concentration in the supernatant
from the clay. However, there was no statistically significant difference
between the BT concentrations in the supernatant in the low and high clay
loading samples. This was true for both clay isotherms.
The data from the two calcium montmorillonite clay experiments were
used to estimate an upper bound of Kp = 17 for the sorption partition coeffi-
cient. It is possible that the small differences in BT concentration between
blanks and clay samples could be attributed to a volatilization loss, since
the workup of the clay samples involved one extra transfer and centrifugation
that was not carried out during the workup of the blanks. The fact that the
less of BT in the clay samples was independent of clay loading would seem to
support this possibility. This loss (expressed as percentage difference
from blanks) remained relatively constant (9.9%, 11.2%, 11.2%, and 7.5%) over
a wide range of clay loadings (100:1, 1000:1, 2000:1, and 5000:1 clay:BT by
206
-------
TABLE 11.7. SORPTION OF BENZO[b]THIOPHENE ON SEDIMENTS
to
o
Sediments
Ca-montmor-
illonite
clay
sorption
Coyote
Creek
sorption
Total
organic
carbon
(percent)
0.06
1.4
Sediment
concentration
(mg ml"1)
6.0
15
8.2
16
19
41
BT
concentration
in supernatant
(mg ml'1)3
2.0 + 0.1
2.1 ± 0.1
2.3 ± 0.0
3.7 ± 0.2
1.5 ± 0.1
2.3 ± 0.0
BT Partition coefficient, K c
p
concentration
on seditnent
o o
36.0 ± 21
8.9 + 5.2 11+6 -170 + 50
(aQ = 374 ± 96)
130.0 + 10
220.0 ±10 59+2 57 ± 8
100.0 ±10 (a = 3 ± 20)
120.0 ±10 °
Concentration measured in supernatant with population standard deviation.
Concentration on sediment calculated from supernatant concentration with population standard deviation.
!•»
"Calculated by linear least squares method; see Appendix B,Section B.I.4 for description of regressions.
Limits are 95% confidence limit.
-------
weight, respectively). It is also possible that the clayrBT loadings that
were used in the isotherms were not in the optimal range for measuring
adsorption.
Good sorption data were obtained, however, with Coyote Creek sediment
loadings of 2000:1 and 5000:1 sediment:BT by weight in those isotherms. The
partition coefficient of 59 estimated from these experiments with Coyote Creek
sediment suggests that sorption of BT was not strong, and comparisons with
volatilization and biodegradation (Section 11.6) supported this conclusion.
Since the environmental assessment suggested that less than 1% of the BT
would be sorbed, no additional isotherm measurements using other natural
sediments were made.
11.4.5 Biosorption
The sorption data obtained with 100 and 300 ng ml"1 solutions of BT
with living as well as heat-killed mixed populations of Azotobacter
beijerinckii ATCC 19366, Bacillus cereus ATCC 11778, Escherichia coli ATCC
9637, and Serratia marcescens ATCC 13880 are presented in Table 11.8. The
sorption partition coefficients for heat-killed cells are more than twice as
large as for living cells.
TABLE 11.8. BIOSORPTION OF BENZO[b]THIOPHENE
BY MIXED BACTERIAL CULTURES
Condition Initial level of Sorption
of cells BT (ng ml"1) coefficient3
Viable cells 100 122 ± 17
Viable cells 300 130 ± 22
Heat-killed cells 100 284 ± 92
Heat-killed cells 300 275 ± 71
(ug/g)/(pg/ml), based on dry weights of cells, average coefficients ±
standard deviation. Dry weights of viable cells = 1.02 ng ml"1;
heat-killed cells, 0.838 ne mi"1.
The sorption of BT by heat-killed cells is not surprising and is
probably associated with the lipid content of these cells. The sorption
coefficient of heat-killed cells, as compared with viable cells, appears to
be somewhat higher, as is indicated by the recoveries. These differences in
recovery yields could readily account for the differences in sorption coeffi-
cients, which are of a relatively low order and not of great significance for
a nollutant.
208
-------
11.5 CHEMICAL TRANSFORMATION
11.5.1 Photolysis Rate
Data for the photolysis of BT in pure water and in water containing
9.5 ug mi"1 humic acid are given in Table 11.9.
TABLE 11.9. RATE CONSTANTS FOR PHOTOLYSIS OF 1.0 Ug ml"1 BENZO [blTHIOPHENE*
Irradiation
source
Borosilicate
filter only
Sunlight, late Ma
Extent of Rate constant
Solution reaction (%) kp x 106 sec"1
Pure
,' PH 6
y| Pure
' pH 6
water,
.9
water,
.9
39
84
11
3.
5.
0.
50
12
689
± 0.
± 0.
± 0.
22b
28
072°
,d
313 nm
9.5 yg ml 1 humic
acid, pH 5.4
22
0.851 ± 0.013
1.0 yg ml 1 benzothiophene in water = 7.45 x 10 M.
Standard deviation.
Q
Calculated on basis of 12 hours of sunlight per day; to obtain average
rate constant for full calendar day (24 hours), divide rate constant by
two.
Half-life of 23 calendar days.
Direct photolyses using only the borosilicate lamp well as a light.
filter to eliminate wavelengths below 290 nm were carried out to 39% and 84%
consumption of BT; good first-order kinetic behavior was observed. Since BT
has only a very small absorption coefficient at 313 nm, the borosilicate
glass filter system was used since it provides greater light intensity than
the 313-nm filter system for these screening experiments. The reason for
the difference in the two rate constants in Table 11.9 is not known, but may
be due to a change in the light flux. With the first-order photolysis
behavior demonstrated by these experiments, direct photolysis of BT in sunlight
was carried out to 11% loss of BT, and the rate constant in Table 11.9 was
Calculated from a first-order plot of the sunlight photolysis data.
A quantum yield was not measured for the direct photolysis of BT
because volatilization of BT with a half-life of just a few hours appears to
be the dominant pathway of BT (see Section 11.3); sunlight photolysis is only
209
-------
1/100 as fast. Moreover, to measure the quantum yield of direct photolysis
at 313 nm would require several hundred hours irradiation time because of the
very small and uncertain value for the absorption coefficient for BT at this
wavelength. The time and effort required for this experiment did not seem
justified in view of the lesser importance of photolysis of BT as an environ-
mental pathway.
It was, however, of interest to estimate how the photolysis half-life
would vary as a function of the time of year by calculating the half-life of
BT in sunlight as a function of season using the procedure of Zepp and Cline
(1977). To make this calculation, a quantum yield for direct photolysis
must either be measured or estimated. The direct photolysis in late May sun-
light was found to have a measured half-life of 23 days. If a quantum yield
of 0.1 is assumed, half-lives of 32 days in midspring and 17 days in mid-
summer are calculated using the procedure of Zepp and Cline. The agreement
between the calculated and measured half-lives indicates that direct photoly-
sis of BT has a quantum yield of ^ 0.1.
The annual variation of half-life for direct photolysis of BT using a
quantum yield of 0.1 is given in Figure 11.6. The figure shows that the half-
life will vary from about 17 days during summer to over 350 days in winter.
The data in Table 11.9 show that the presence of humic acid promotes
the photolytic transformation of BT. If we assume, based on the previous
discussion, that direct photolysis of BT at 313 nm has a quantum yield of
0.1 and that the absorption coefficient at 313 nm has a maximum value of
unity, the photolysis half-life of BT at 313 nm is greater than 950 hours.
Thus, the acceleration of photolysis rate at 313 nm in the presence of humic
acid may then be as much as fourfold: the half-life for BT, when photolyzed
for 12 hours per day in sunlight in the presence of humic acid, may be about
6 days. Although the sensitized photolysis is a rapid process, volatilization
will still be dominant under most environmental conditions, and for this rea-
son no humic-acid-sensitized photolysis of BT was conducted in sunlight.
11.5.2 Oxidation Rate
The susceptibility of 1.2 ug ml"1 (8.8 x 10~6 M) BT to free radical
oxidation was examined using the AA-initiated oxidation reaction (see Part I,
Section 6.3 of this report and Appendix B for discussion and procedure).
This experiment was carried out at 50°C in pure water containing 1.0 x 10~/<
M AA. At the end of 100 hours reaction time, 60% of the BT had been consumed,
corresponding to a first-order rate constant (kox[R02"]) of 2.5 x 10~6 sec"1.
Under these conditions (see Part I, Section 6.3) this value corresponds to
a second-order rate constant kQX of 750 M"1 sec"1. At 25°C, kox would be
83 M"1 sec"1, and with the assumption that [R02*] = 10~9 M in the environment,
the half-life for BT toward oxidation in the aquatic environment is 96 days.
The free radical oxidation of BT by R02' is clearly not competitive with
volatilization in aquatic systems.
210
-------
ca
LU
360
320
280
240
200
I] 160
i
LL.
_l
x 120
80
40
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MONTH OF YEAR
S A -439 6-68
FIGURE 11.6 ANNUAL VARIATION OF PHOTOLYSIS HALF-LIFE FOR BENZO(b]THIOPHENE
-------
11.5.3 Hydrolysis Rate
BT contains no functional groups that are hydrolyzable; therefore,
no hydrolysis studies were carried out.
11.5.4 Products from Chemical Transformation
Since screening studies indicated that chemical transformation pro-
cesses were not competitive with volatilization in aquatic systems, detailed
kinetic and product identification studies were not carried out.
11.6 BIODEGRADATION
11.6.1 Development of Enrichment Cultures
Samples of Coyote Creek water and aeration effluent from the Palo
Alto Sewage Plant, which were aerated and buffered in our 9-liter fermentors,
did not produce BT-degrading systems when BT was the only carbon source added.*
Because of the volatility of BT from aqueous solutions that are aerated, BT
was added equivalent to 10, 10, 50, and 50 yg ml"1 at days 0, 2, 4, and 7,
respectively. Aliquots from the 5-liter volumes in the 9-liter fermentors
were transferred at 3, 7, 10, and 14 days to media composed of basal inor-
ganic salts, BT (50 yg ml""1), glucose (50 yg ml"1), and Difco Bacto yeast
extract (50 yg ml"1). This medium (100 ml) was contained in 500-ml Erlenmeyer
flasks, which were capped with aluminum foil or screw-on caps to prevent loss
of BT by aeration. BT-biodegradation was also not observed in these shaker
flasks.
BT-biodegrading systems were, however, obtained in the 9-liter bottle
fermentors after five to eight days if BT and naphthalene (each equivalent
to 10 yg ml"1) were added at two-day intervals. Transfers from the 9-liter
fermentors were made to shaker flasks with media containing BT, naphthalene,
and basal salts. Degrading systems were obtained with waters from Coyote
Creek and from the eutrophic pond near Searsville Lake in Woodside, Califor-*
nia, as well as with aeration effluents from the Palo Alto, South San Fran-
cisco, and Shell Oil Refinery wastewater treatment plants. These mixed
cultures could be readily adapted to grow on media containing BT and naphtha-
lene, each at 50 yg ml"1. The Lake Tahoe water sample did not yield a BT-
degrading system after six weeks of aerobic incubation and addition of BT
and naphthalene (10 yg ml"1 of each).
With all the above BT/naphthalene enrichment cultures, there was
generally 70% to 80% decomposition of BT in two to three days when the media
initially contained 50 yg ml"1 each of BT and naphthalene. It soon became
apparent that BT degradation was dependent on the presence of naphthalene.
In an effort to develop a system without this dependency on naphthalene
(that is, a system that would degrade BT when BT was the sole carbon source),
These data were obtained using a UV assay, as described in Appendix B
212
-------
we attempted to develop a biodegrading system with aeration effluents from
the Shell Oil Refinery. The following observations were made.
When a transfer was made from a medium containing BT and naphthalene
each at 30 yg ml"1 to a medium containing only 30 yg ml"1 BT as the
sole carbon source, no BT degradation took place.
When the culture mixture was grown in a medium containing BT and
naphthalene each at 50 yg ml"1, more than 80% of the BT was decom-
posed in 76 hours, but if tryptone (50 yg ml"1), or a mixture of
glucose (50 yg ml"1) and Difco Yeast Extract (10 yg ml"1) was sub-
stituted for the naphthalene, no BT degradation occurred.
When the culture mixture was grown with naphthalene (50 yg ml"1) as
the sole carbon source, it developed an optical density of 0.05 at
24 hours, but when BT and naphthalene were present (50 yg of each
ml"1) the optical density was 0.03.
When transfers were made to mineral salts/BT (5 and 10 yg ml"1)
no BT degradation took place in 64 hours. If at this time, naphtha-
lene (10 yg ml"1) was added, more than 90% of the BT was degraded
in 24 hours.
Growth was observed when the basal salts medium was supplemented
with BT (20 yg ml"1) and quinoline (50 yg ml"1), but no BT degra-
dation occurred in 7 days. The control with BT (20 yg ml"1) and
naphthalene (50 yg ml"1) had more than 90% destruction of BT in 36
hours.
The culture mixtures grown in salts/naphthalene (100 yg ml"1) and in
salts/BT (20 yg ml"1) plus naphthalene (100 yg ml"1) media for 22
hours were centrifuged, washed, and suspended in 0.05% phosphate
buffer (pH 7.0). Cells were rested at RT for 20 hours, centrifuged
again, and resuspended in salts media containing BT (20 yg ml ),
or BT (20 yg ml"1) plus naphthalene (20 yg ml"1). The following
results were obtained:
Additives to Additives to
growth media test media Residual BT
(yg ml"1) (yg ml"1)
BT
0
0
20
N
100
100
100
BT
20
20
20
N
0
20
0
7 hr
96
72
98
22 hr
88
1
92
20 100 20 20 37
Cultures grown on salts media supplemented with BT (50 yg ml x) and
naphthalene (50 yg ml"1) for 64 and 70 hours had residual BT values
of 25 and 1 yg ml"1, and cell counts of 7.2 x 107 and 13.2 x 107
cells ml"1, respectively.
213
-------
A BT-degrading system without dependency on the presence of naphtha-
lene could not be developed by repeated subtransfers on media with
gradual sequentially reduced naphthalene concentrations. When
naphthalene was reduced to 10 yg ml"1, 80% of the initial 30 yg ml"1
of BT was decomposed; but when naphthalene was at 5 yg ml"1, only 40%
to 70% of the initial 30 yg ml"1 BT was degraded.
• Replacing (NHiJaSOj, with NHi»N03 to reduce SO* in the basal salts
medium did not affect the results.
In our final attempt to develop cultures that could degrade BT when it
was the only carbon source in an inorganic salts medium, we incubated aeration
effluent from the Palo Alto Sewage Plant in our 9-liter aerobic fennentors
with initial additions of BT and naphthalene (10 yg ml"1 of each), BT and
quinoline (10 yg ml"1 of each), BT and glucose (10 yg ml"1 of each) with Difco
Bacto Yeast Extract (1 yg ml"1), and BT and Difco Bacto tryptone (10 yg ml"1
of each). At day 2, additional quantities of all the above components except
quinoline were added to the respective bottles. Subsequently, all compounds
or preparations as originally present were added at two-day intervals. Period-
ically, aliquots from the 9-liter fermentors were transferred to capped shaker
flasks containing inorganic salts media and the corresponding additives.
After 6 days, a BT-degrading system was obtained only in the bottle with BT
and naphthalene as additives. After 26 days of incubation of the 9-liter
fermentors and 13 days of incubation of the subtransfers in shaker flasks,
no other BT-degrading systems were observed.
11.6.2 Biodegradation Kinetics
Because no culture system was developed that could utilize BT as
the sole carbon source, no kinetic studies were conducted.
11.6.3 Identification of Metabolites
Six BT metabolites were tentatively identified by gas chromatography/
mass spectrometry analysis of ethyl acetate extracts of acidified fermenta-
tions in which BT and naphthalene were the only carbon substrates. The
products were assigned the following structures:
0
IV
(cis and trans
isomers)
214
-------
Subsequent studies have shown that products II and III are artifacts of
the analytical procedure.
11.6.4 Discussion
The fact that all our BT degrading systems depended on naphthalene
does not necessarily indicate that if we had used other hydrocarbons we
could not have obtained BT decomposition. Naphthalene was selected in our
studies because we felt it might be an inducer of enzymes that could also
degrade BT and because naphthalene would usually be present in larger amounts
if BT were discharged into aquatic systems.
In the enrichment studies it was apparent that BT had a toxic effect
on microorganisms, and this may be an important factor in our failure to
develop a culture system that would utilize BT as a sole carbon source.
Although our studies involved reservoirs of many types of organisms, as is
the realistic situation in nature, this does not preclude the possibility
that microorganisms exist or may be developed that could utilize BT as a
sole source of carbon. More realistic, however, is the fact that in the
environment many other pollutants coexist that could function in cooxidative
mechanisms to decompose BT.
The structures of the metabolites observed in these studies indicate
that the thiophene ring is susceptible to biological oxidation. These
compounds represent nearly all primary oxidations that could be expected
on the carbons and sulfur in the thiophene ring.
215
-------
12. LABORATORY INVESTIGATION OF DIBENZOTHIOPHENE
12.1 SYNOPSIS
The results of the laboratory investigation suggest that biodegradation
is the major environmental pathway for dibenzothiophene (DBT) in eutrophic
waters, while volatilization and photolysis dominate in oligotrophic waters.
Dilution is competitive with biodegradation in eutrophic lakes and ponds and
is several times faster in rivers. Sorption is important only in ponds where
heavy sediment loading is assumed.
The nine-compartment environmental exposure model predicted the following
steady-state concentrations of DBT in solution, suspended solids, and sediments
near point sources in the presence of a continuous discharge of 1 yg ml"1
(1 ppm) DBT:
Suspended
Half-life
(hr)
Solution
(yg ml-1-)
8.50 x 1CT3
1.32 x Hr4
3.72 x ICT4
solids
(yg g-1)
1.19 x 10
1.84 x 10~ x
5.20 x 10"1
Sediments
(yg g-1)
1.15 x 10
1.84 x 10"1
1.32 x 10"1
River 0.5
Pond 13
Eutrophic 13
lake
Oligotrophic 140 1.06 x 10~3 1.48 5.10 x 10"1
lake
12.2 BACKGROUND
Dibenzothiophene is found in coal tars and petroleum and is probably in
the effluent streams from coal gasification or liquefaction. Dibenzothiophene
is also representative of the large class of organic sulfur compounds, includ-
ing thiophenes, thienylsulfides, benzo(b)thiophenes, dibenzothiophenes, and
benzo(b)naphthothiophenes, which could present particular problems in the en-
vironment. The physical properties of dibenzothiophene (DBT) are listed in
Table 12.1.
The dibenzothiophenes present in coal and bitumens have recently been
discussed by Hayatsu et al. (1975) and Clugston et al. (1976), respectively.
*
Predicted by the one-compartment model.
216
-------
TABLE 12.1. PHYSICAL PROPERTIES OF DIBENZOTHIOPHENE
Structure
Molecular weight
Melting point (°C)a
Boiling point (°C)a
Vapor pressure at 20° (torr)
Solubility ?in water (pg ml"1)
UV spectrum (in ethanol) [X nm (log e)]a
max
1 pg ml-1 DBT = 5.43 x 10"6 M
184.27
99-100
332-3
2 x 10~3
1.11±0.09
234 (4.9)
256 (4.3)
262 (4.1)
303 (3.5)
325 (3.6)
CRC Handbook of Chemistry and Physics, 57th Ed., 1976.
Estimated from empirical equation for temperature range
110° to 330°C (Aubry et al., 1975).
Measured in this study (Section 12.4.1).
Concentration data for dibenzothiophenes in coal, coal tars, and petroleum
crude oils are somewhat controversial because thermal degradation of the na-
tional sulfur-containing polycyclic compounds may occur during analysis.
However, there is ample evidence that many of these compounds are polyalkylated.
In American Petroleum Institute studies on some crude oils, no dibenzothiophenes
were reported (C. T. Hou, personal communication), but Kuwait and South Louisi-
ana crude oils contained 3.3% and 0.4% dibenzothiophenes, respectively. These
crudes also contained 5.9% and 0.7% naphthalenes, respectively.
Data on the water solubility, volatilization, or sorption of DBT could
not be found in the literature.
Dibenzothiophene will not hydrolyze under the conditions prevailing in
environmental aquatic systems.
Data were not available for evaluating the importance of chemical oxida-
tion of DBT under environmental conditions. However, benzothiophene is oxidized
to the sulfoxide and sulfone by hydrogen peroxide, peracids, and hydroperoxides
(Anashkina et al., 1972; Ponec and Prochazka, 1974; Ford and Young, 1965), and
similar processes should occur with DBT. In our study of the oxidation of ben-
217
-------
zothiophene by peroxy radicals (see Section 11), the half-life for oxidation in
the aquatic environment was estimated to be over 10 years.
DBT absorbs strongly in the solar spectrum up to about 340 nm. No infor-
mation on the photolysis of DBT was found in the literature. Benzothiophene,
which has a long wavelength cutoff of about 310 nm (absorption coefficient of
7 M~l cm~*) in the solar spectrum, was found to photolyze in pure water with a
half-life of about 23 days.
The literature gave considerable evidence that Knecht (1962) obtained a
mixture of organisms that could utilize dibenzothiophene as a sole carbon source
in a medium lacking inorganic sulfur. These organisms belonged to the Arthro—
bacter and Pseudomonas genera and were obligately synergistic. Sulfate was one
of the end products of metabolism, and the addition of organic nitrogen sources
to the medium reduced the desulfurizing activity. Kuwait residuum, a fraction
high in organic sulfur, supported growth of the organisms but no detectable
reducible-sulfur accumulated in the medium. Knecht postulated that the initial
attack on the DBT was at the thioether linkage and cleavage of one of the ben-
zene rings. Walter et al. (1975) in studying microbial degradation of petro-
leum hydrocarbons concluded that aromatic nuclei containing sulfur were twice
as refractory as nonsulfur analogs.
Yamada et al. (1968) reviewed microbiological studies used to desulfurize
petroleum products and reported their first studies on the biodegradation of
DBT. They isolated six biodegrading strains: one was identified as Pseudo-
monas jianii. In a later report from this laboratory (Nakatini et al., 1968),
three more DBT degrading isolates were reported. In studies with mixtures of
these isolates, a decrease in biodegradation activity of DBT was noted with
some mixtures of active isolates. Their optimal medium contained 8.78% light
oil, 0.4% meat extract, and 0.46% DBT, at an initial pH of 7.3, and they used
a mixture of four isolates.
In two subsequent publications, Kodama et al. (1970, 1973) reported the
isolation and identification of 3-hydroxy-2-formyl-benzothiophene, dibenzothio-
phene-5-oxide, 3-oxo-2[3'-hydroxy-thianaphthenyl-(2)-methylene]dihydrothia-
naphthalene, trans-4[2-(3-hydroxy)-thianaphthenyl]-2-oxo-3-butenoic acid, and
the hemiacetal of this last product, as metabolites produced from DBT by P.
abikonensis or _P_. jianii. Their medium contained 4.0 g meat extract in an
inorganic salts medium that did not include any sulfates.
Laborde and Gibson (1975) used a Beijerinckia species mutant B836. The
parent strain, obtained by enrichment procedures with biphenyl, could completely
metabolize this compound when it was the sole carbon source. The mutant, ob-
tained by exposure of the parent to a mutating agent, had restricted diphenyi
metabolizing capabilities. When the mutant was grown on inorganic salts medium
to which DBT and Na succinate were added, it produced cis-l,2-dihydroxy-l,2-di-
hydrodibenzothiophene. This is consistent with Gibson's position that microbes
produce cis-dihydrodiols and mammalian systems produce trans isomers ,when aromatic
compounds are converted to dihydrodiols. Gibson (1975), in studies with the
parent and mutant strains, reported that in addition to the above cis-dihydrodiol
DBT was transformed to 1,2-dihydroxydibenzothiophene, trans-4[2-(3-hydroxy)-
thianaphthenyl]-2-oxo-3-butenoic acid, 3-hydroxy-2-formylbenzothiophene, and
d ibenzothiophene-5-oxide.
218
-------
Hou and Laskin (1975) isolated a strain of Pseudomonas aeruginusa
(ECR-8) that, in the presence of dodecane and tetradecane as hydrocarbons to
solubilize DBT (and perhaps as cometabolites), resulted in the formation of
4[2-(3-hydroxy)-thianaphthenyl]-2-hydroxy-3-butenoic acid and two other com-
pounds—probably a tetradecane ester of this acid and probably a chromophoric
derivative of this acid.
Walker and Colwell (1974) used sediments collected from an oil-polluted
creek at Baltimore Harbor in Chesapeake Bay to inoculate media contained 2 vol%
motor oil or mixed hydrocarbon substrate. They reported aerobic degradation of
benzothiophenes, dibenzothiophenes, and naphthobenzothiophenes in the range of
90% to 97%.
Kuritz et al. (1971) isolated four kinds of bacteria that produced H~S
from organic sulfur compounds such as thiophene, dimethylsulfide, 1-butanethiol,
polysulfides, crude oil, residue oil, and asphaltene when the organisms were
grown in the presence of polypeptone and FePO^ under anaerobic conditions. It
is very probable that DBT adsorbed on sediments may also be decomposed by these
types of organisms in fresh water systems.
Although the above cited studies clearly indicated that DBT is biodegrad-
able, no kinetic constants were available that could be used in our environmen-
tal assessment procedures.
12.3 ENVIRONMENTAL ASSESSMENT
12.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigations of the
transformation and transport processes for DBT are summarized in Table 12.2.
12.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved DBT calculated for individual transforma-
tion or removal processes in the one-compartment model are given in Table 12.3.
Biotransformation (biodegradation) is clearly the dominant transformational
pathway in eutrophic waters, but is roughly 2000 times slower than volatiliza-
tion in oligotrophic waters. Dilution is very nearly as important as biotrans-
formation in eutrophic lakes and ponds, and several times faster in rivers.
Sorption is important only in the pond simulation, which assumes heavy sediment
loading.
12.3.3 Persistence
Dibenzothiophene is not persistent in eutrophic waters, where half-
lives are predicted to be half a day. In oligotrophic waters, however, half-
lives should be about a week. The half-life of DBT in entire river segments
will not differ from the half-life of DBT in ponds and eutrophic lakes because
biodegradation is the dominant process in all these waters. However, DBT con-
centration in individual segments of a river will drop rapidly after a spill
because of dilution from incoming water.
219
-------
TABLE 12.2. SUMMARY OF LABORATORY DATA FOR DIBENZOTHIOPHENE
Process Sorption equilibrium Partition coefficient
Sorption3 S = K S Ko = 1380
s p w v
Rate expression Rate constant at 25 C
Volatilization13 kDBT k = (0.12*0.04)^
V
Photolysis0 k [DBT] = $(EZ, e, [DBT] k = (2.04 * 0.08) x 10~ 6sec"1
p A A p
Oxidation k [R02-][DBT] k <7.5 M"1 sec"1
OX OX
Hydrolysis NA
Biodegradation k,. = ^-%- k,, = (5.3 ± 0.4) x 10"7 ml cell"1 hr"1
DZ I K. D£
S
Measured on Coyote Creek sediment.
See Part I, Section 5.3, and Appendix B.
Assumes 12 hours sunlight per day in early March.
TABLE 12.3. TRANSFORMATION AND TRANSPORT OF DIBENZOTHIOPHENE
PREDICTED BY THE ONE-COMPARTMENT MODEL
Photolysis, half-life (hr)
Oxidation, half-life (hr)
Volatilization, half-life (hr)
Hydrolysis, half-life (hr)
Biodegradation, half-life (hr)
Half-life for all processes
except dilution (hr)
Half-life for all processes
including dilution (hr)
Amount DBT sorbed (mg iff3)
Percentage DBT sorbed
River
380
>105
140
—
13
13
0.5
1.4
12%
Eu trophic
pond
950
>105
720
—
13
13
13
4.2
42%
Eu trophic
pond
950
>105
580
—
13
13
13
0.7
6.5%
Oligotrophic
pond
190
>105
580
XLO4
140
140
0.7
6.5%
Estimates are the average photolysis rates on a summer day at 40° latitude.
Photolysis rates in midwinter are at least three times slower.
Assumes 10 ng ml"1 of dibenzothiophene in aqueous phase and a partition coef-
ficient of 1400.
220
-------
12.3.4 Mass and Concentration Distributions Calculated Using Computer Models
The pseudo-first-order rate constants used in these simulations are
presented in Appendix A. The distributions of mass and concentration of DBT
expected at steady state during chronic discharge to each of four types of
water bodies are given in Table 12.4. The changes in concentration during and
after discharge of DBT into the water bodies are shown in Figures 12.1 through
12.4.
The concentrations of DBT predicted to be in solution at steady state
are on the order of 10~3 to 10~5 pg ml"1, with concentrations on the order of
10~4 yg ml-1 being most common (Table 12.4). The patterns of variation in con-
centration within and between simulations generally resemble those for the pre-
viously studied compounds (Figures 12.1-12.4). Changes in concentrations on
suspended sediments parallel changes in the concentrations of dissolved DBT
(Figure 12.1). ;
Concentrations of dissolved DBT in the computer simulations generally
reach steady state, or very close to steady state, within 2 to 10 days of
start of discharge. Declines from those relatively high concentrations to the
lower steady states, which maintained by desorption from contaminated sediments
after simulated discharges stop, is slower, perhaps 2 to 5 times slower (Fig-
ures 12.1-12.4). The concentrations in the sediments build up slowly during
exposure and decline slowly when the input ceases. However, the corresponding
increases and decreases in the aqueous compartments are rapid. The oligotrophic
lake, where biodegradation is slow, differs from this pattern and the concen-
tration in compartments that are relatively far from the point of contamination
declines rather slowly.
12.3.5 Discussion
In eutrophic systems, biodegradation appears to be the dominant process
and the half-lives of DBT are similar in the river, pond, and lake. The concen-
trations of DBT in various locations of water bodies decrease at approximately
the same rate after the discharge stops, except in the oligotrophic lake, where
decreases in different compartments are substantially different because biodeg-
radation is very slow. For instance, the concentration in the lower part of
the oligotrophic lake (Compartment 5) shows an insignificant decrease after
discharge stops, yet the concentrations in the upper part of the oligotrophic
lake show observable decreases due to the losses from photolysis and volatili-
zation. The relatively slow decline in the concentration of dissolved DBT is
a result of the relatively low microbial population in the oligotrophic lake.
In conclusion, DBT is expected to persist longer in oligotrophic waters
than in eutrophic waters. In the lower part of the oligotrophic lake, DBT has
a half-life of 15 days. During spring and fall turnovers in the lake, DBT may
be brought up to the surface where photolysis and volatilization are important,
and half-lives will be shorter.
221
-------
TABLE 12.4. DISTRIBUTION OF DIBENZOTHIOPHENE IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations of 1 ng ml"1 dibenzothiophene)
to
to
to
Compartment 1
(surface water)
Solution
Suspended solids
Compartment 2
(surface water)
Solution
Suspended solids
Compartment 3
(surface water)
Solution
Suspended solids
Compartment 5
(surface water)
Solution
Suspended solids
Compartments 7-9
(sediment)3
Solution
Solids
Pond
Mass Cone.
(kg) (pg ml-1)
2.64xlO~3 1.32x10-" 2.
1.10xlO~3 1.84xlO-! 0.
.... — — 2
0.
2.
0.
_— — — __
—
3.30x10-5 1.32x10-" 6.
1.24X10"1 1.84X10-1 2.
River
Mass
(kg)
55
35
48
34
41
33
17 x lO-2
32xl02
Cone.
(yg ml-1)
8.
1.
8.
1.
8.
1.
8.
1.
50
19
26
15
03
12
23
15
xlO"3
xlO
xlO~3
xlO
xlO"3
xlO
__
—
xlO"3
xlO
Eutrophic lake
Mass
(kg)
9.31x
6. Six
6.67x
4.67x
2.91x
2.05x
5.91x
4.14x
8.24x
1.81
ID'2
10- 3
ID'2
io-3
io-3
10-"
10-"
10-5
10""
Cone.
(pg ml-1)
3.72x
5.20x
2.66x
3.73x
1.16x
1.64x
2.36x
3.31x
9.42x
1.32x
10-"
10- l
10-5
lO-2
10" 5
10-2
10-7
10-"
10-5
10-1
Oligotrophic lake
Mass
(kg)
2.65X10-1
1.85x10-2
1.10
7.70x10-2
9.20xlO"2
6.44xlO~3
5.80 x lO"2
4.06xlO"3
3.19xlO~3
6.95
Cone.
(yg ml-1)
1.06xlQ"3
1.48
4.40x10""
6.16X10-1
3.68x10-"
S.lSxlQ-1
2.32xlO-5
3.24x10-2
3.64x10""
S.lOxlO-1
The amounts given for solid and solution phases in the sediment compartments are estimated from the sorption
partition coefficient for suspended solids and may be overestimated because it was assumed that biodegradation
of sorbed material does not occur.
-------
10
-6
100 150
TIME - hours
200
250
FIGURE 12.1. PERSISTENCE OF DIBENZOTHIOPHENE IN A TWO-COMPARTMENT POND SYSTEM
223
-------
10-1
10
-2
10
-3
z
IU
I
Q-
o
o
N
UJ
CD
5
LL
O
Z
O
10"
a.
z
UJ
o
O
o
10-"
10
-7
I I
COMPARTMENT
xXX5/
DISCHARGE
STOPPED
— —SEDIMENTS
SOLUTION
TIME - hours
10
FIGURE 12.2. PERSISTENCE OF DIBENZOTHIOPHENE IN A PARTIALLY MIXED RIVER SYSTEM
224
-------
3 x 10"2
10'2
I I
COMPARTMENT
10
-3
10"
SEDIMENTS
SOLUTION
n-9
DISCHARGE
STOPPED
100
200 300 400
TIME - hours
500
600
700 720
FIGURE 12.3. PERSISTENCE OF DIBENZOTHIOPHENE IN A EUTROPHIC LAKE
225
-------
SEDIMENTS
SOLUTION
10
100
200 300 400
TIME — hours
500
600
700 720
FIGURE 12.4. PERSISTENCE OF DIBENZOTHIOPHENE IN AN OLIGOTROPHIC LAKE
226
-------
12.4 PHYSICAL PROPERTIES
12.4.1 Solubility in Water
The solubility of DBT in water at 28 ± 1°C was measured using the
method of Campbell (1930). The solubility was 1.11 ± 0.09 ug ml-1, based on
replicate determinations.
12.4.2 Absorption Spectra
DBT absorbs light strongly in the solar region to 350 nm. The absorp-
tion spectrum of DBT in 30% acetonitrile-70% water was measured from 210 nm to
350 nm and in 80% acetonitrile-20% water from 320 nm to 380 nm. The pH of the
first solution was 7.1; the pH of the second solution was not measured. The
absorption coefficients at wavelength intervals from 297.5 to 360 nm and at
wavelengths of 313 and ;366 nm are reported in Table 12.5.
TABLE 12.5 ABSORPTION SPECTRUM OF DIBENZOTHIOPHENE
IN 30% ACETONITRILE-70% WATER AT pH 7a
Center of wavelength Average absorption
interval" coefficient0
(nm) (M-1 cm"1)
297.5 1154
300.0 1224
302.5 1327
305.0 1499
307.5 1782
310.0 2025
312.5 2080
315.0 1939
317.5 1892
320.0 2119
323.1 2394
330 526d
340 13. ld
350 7.5d
360 Od
r\ 1
Dibenzothiophene concentration was 23.5 Mg ml '
(1.27 x 10-4 M).
The wavelength intervals are given in Appendix B,
Table B.I.
p
The absorption coefficients at 313.0 and 366 nm are
2080 and 0 M"1 cnr1, respectively.
Values were determined from a solution of dibenzo-
thiophene in 80% acetonitrile-20% water at a concen-
tration of 98.2 pg ml-1 (5.33 x 10~'* M) ,
227
-------
12.4.3 Volatilization Rate
The volatilization rate of dibenzothiophene was measured using the
method of Hill et al. (1976), which is described in Part I, Section 5.3. The
DBT volatilization data obtained at four different oxygen reaeration rates are
TTRT
shown in Table 12.6. The DBT volatilization rate constant kUDi is plotted ver-
sus the oxygen reaeration rate constant k° in Figure 12.5.
TABLE 12.6. VOLATILIZATION DATA FOR DIBENZOTHIOPHENE
Oxygen reaeration
rate, k° (hr-1)
4.49 ± 0.07
2.19 ± 0.43
0.432 ± 0.012
0.199 ± 0.014
DBT volatilization
rate, k°,BT (hr-1)
0.107 ± 0.001
0.0651 ± 0.0025
0.0474 ± 0.011
0.0318 ± 0.0037
, DBT ,, 0
k /k
v v
0.024 ± 0.002
0.030 ± 0.008
0.11 ± 0.03
0.16 ± 0.03
The critical volume of DBT was estimated using the critical volume
increments of Lydersen (Reid and Sherwood, 1966) to be 525 ml/g-mole. Since
the molecular volume is 1/2 to 1/3 of the critical volume (Present, 1955) , this
critical volume yields molecular diameter estimates of 9. 4 to 8. 2 A. If volatil
ization were completely controlled by liquid phase resistence, then Equation
(5.7) in Part I [Equation (12.1) below] can be used to estimate the ratio
k0,. Using the value of k° = 2.98 A (Tsivoglou, 1965),
kDBT Ho
- =0-34 ±0.02 (12.1)
For oxygen, mass transport is limited by liquid phase resistance for all k
< 0.6 hr"1. There are only two points on Figure 12.5 where k^ < 0.6. Using
these two points kDBT/k° = 0.12 ± 0.04, which is significantly different from
the prediction of equation (12.1).
The reason for the discrepancy is probably that the Henry's law constant
for DBT, which is estimated to be 335 torr M"1 using the method of Mackay and
Wolkoff (1973), is low compared with the constant for H° (3 x Wk torr M"1).
Since DBT has such a low Henry's law constant with respect to oxygen, the gas
phase resistance to volatilization becomes important for DBT at turbulence
levels where reaeration is still controlled by the liquid phase. The addition
of the gas phase resistance slows the volatilization rate of DBT and yields a
volatilization rate lower than predicted by the simple theory of equation (12.1).
228
-------
to
to
0.12
I
iu
0.10
0.08
g
I
- 0.06
5
O
fe
Q
fc
Q
0.04
0.02
_ II
0.5
T
™ *
J.
1.5
2.5
k° OXYGEN REAERATION RATE - hr'1
3.5
H
4.5
FIGURE 12.5. VOLATILIZATION OF DIBENZOTHIOPHENE FROM AQUEOUS SOLUTION
-------
12.4.4 Sorption on a Sediment
The sorption partition coefficient of dibenzothiophene on Coyote Creek
sediment was measured; the data are shown in Table 12.7. The experiment con-
sisted of duplicate samples at two DBT concentrations and three sediment load-
ings, in addition to blanks containing DBT only and sediment only. The concen-
trations of DBT on the sediment were calculated from the concentration differ-
ence between the supernatants and the appropriate blanks.
Additional isotherm measurements were not conducted since sorption
should not be an important pathway for dibenzothiophene.
TABLE 12.7. DIBENZOTHIOPHENE SORPTION
ON COYOTE CREEK SEDIMENT3
Sediment
concentration
(mg ml-1)
0.47
0.16
0.78
0.16
DBT
concentration
in supernatant
(yg ml-1 x 103)b
103 * 2
129 ± 5
64 ± 2
95 ± 4
DBT Partition coefficient, K
concentration LLS^
« « j ~ ~.n *.
/ -He a=0 a0/0
(yg g 1)c o °
120 ± 5
193 ± 31 1380 ± 130 1580 ± 540
71 ± 3 a0 = -2 ± 54
153 ± 27
Total organic content = 1.4%.
Concentration measured in supernatant with population standard deviation.
"Concentration on sediment calculated from supernatant concentration with
population standard deviation.
LLS = linear least squares regression; see Appendix B, Section B.I.4, for
explanation of regressions; limits are 95% confidence limits.
12.4.5 Biosorption
A saturated solution of DBT in water was prepared by coating the inside
of a clear 9-liter bottle with a solution of DBT in acetone and evaporating the
acetone with a stream of N£• Highly purified water was added to this bottle,
and the resultant suspension was shaken. The suspension was centrifuged to
provide a clear supernatant solution of DBT. The biosorption portion coeffi-
cient was measured using this solution and our mixture of four types of bac-
teria. Biosorptions were conducted in Corex tubes placed in a rotor drum.
The centrifuged cell pellet and clear supernatant were extracted with ethyl
acetate, and the sorption coefficients were calculated for both viable and
heat-killed cells (Table 12.8).
The sorption coefficient with heat-killed fells was signifii-antlv
higher than with viable cells. No data are available to explain this
230
-------
phenomenon at this time, but it is apparent that the sorption is not
necessarily dependent on metabolic prof-esses.
TABLE 12.8. BIOSORPT10N OF DIBENZOTHIOPHENE
?P' Cell
No.
1 Viable
Viable
type
cells
cells
Concentration3
(g liter"1)
1
1
Heat-killed cells 0
2 Viable
Viable
cells
cells
Heat-killed eel
; i
i i
Ls 0
.108a
.iooa
.92
.054
.054
.847
Initial DBT
concentration
(ng ml"1)
59
118
118
120
350
350
(1
(1
(2
(1
(1
(2
Sorption
coefficient
.54
.55
.81
.64
.35
.99
±
±
±
±
±
i
0.
0.
0.
0.
0.
0.
20)
37)
60)
42)
21)
47)
x 103
x 103
x 103
x 103
x 103
x 103
Dry weight of cell suspension.
12.5 CHEMICAL TRANSFORMATION
12.5.1 Photolysis Rate
Photolysis of DBT is slow in the solar region with a half-life of
greater than 5 days in summer. The photolysis rate of DBT is unaffected by
oxygen but is accelerated by the presence of humic acid. Photolysis experi-
ments were conducted with 0.50 pg ml"1 (2.71 x 10"6 M) DBT in pure water and
in natural waters from several sources; these waters also contained 1% added
acetonitrile to increase the solubility of DBT. Data for the photolysis of DBT
in sunlight and at 313 nm are given in Table 12.9. First-order kinetic behavior
was found for photolyses carried out to beyond one half-life. The quantum yield
at 313 nm for direct photo trans format ion of DBT in pure water is 5.0 x 10"4.
The data in Table 12.9 show that photolysis rates of DBT at 313 nm in
the three natural waters are about the same (within experimental error) as the
rate in pure water. Two additional photolysis experiments (not shown in Table
12.9) were carried out to 50% conversion at 313 nm on a pure water solution of
0.5 yg ml-1 DBT. One of the solutions had been degassed by three freeze-thaw
cycles. The absence of oxygen did not affect the photolysis rate. The experi-
ment at 313 nm in the water containing 810 ug ml""1 humic acid absorbance of 0.46
at 313 nm theoretically should have reduced the rate 38% due to a light screen-
ing effect, the humic acid is in fact weakly promoting photolysis of DBT at 313
nm.
The sunlight photolysis experiments confirm this conclusion since DBT
in the humic acid solution is photolyzed over twice as fast as in pure water.
DBT photolyses in sunlight are more rapid in humic acid solution than in pure
231
-------
TABLE 12.9. RATE CONSTANTS FOR PHOTOLYSIS
OF 0.5 yg ml-1 DIBENZOTHIOPHENEa
Irradiation
source
313 nm
313 nm
313 nm
313 nm
313 nm
„ , h Extent of
Solution" . ,„.,
reaction (%)
Pure water
Lake Tahoe
Searsville
water
pond water
Coyote Creek water
Humic acid
(8 yg ml-1)
60%
37
33
37
55
Rate constant
(Kp x 106 sec"1)
2
2
2
2
1
.32 ±
.55 ±
.38 ±
.92 ±
.93 *
0.
0.
0.
0.
0.
22°
34
27
66
11
,d
in pure water
Sunlight,
early March
Sunlight,
early March
Pure water
Humic acid
(8 ug ml-1)
35
63
2
4
.04 ±
.60 ±
0.
0.
08e
04e
,f
,g
in pure water
r-l i r
1.00 pg ml"1 Dibenzothiophene in water = 5.43 x 10 M.
Solutions contained 1% acetonitrile as cosolvent.
Standard deviation.
Quantum yield for disappearance of dibenzothiophene was 5.0 x lO"1*.
Calculated assuming 12 hours of sunlight per day in early March; to
obtain average rate constant for full calendar day (24 hours), divide
rate constant by two.
fHalf-life of 7.9 days.
8Half-life of 3.5 days.
water because humic acid absorbs more extensively in the solar spectrum than
DBT and transfers some fraction of the energy to DBT. The humic acid photosen-
sitized process then dominates over any screening effect of humic acid in the
less extensive absorption of DBT in the solar spectrum.
The half-life for direct photolysis of DBT in sunlight as a function
of the time of year was calculated by the procedure of Zepp and Cline (1977)
using the quantum yield of 5.0 x 10 measured at 313 nm and the measured UV
spectrum of DBT. The data are plotted in Figure 12.6. The half-life of about
12 days calculated for direct photolysis in sunlight in early March is in good
agreement with the measured half-life of about 8 days.
232
-------
to
U)
28
26
24
22
20
18
i,.
I
UJ
LL 14
10
8
6
4
2
0
MEASURED
I
JAN FEB MAR APR MAY JUN JUL
MONTH OF YEAR
AUG
SEP
OCT
MOV
DEC
FIGURE 12.6. ANNUAL VARIATION OF PHOTOLYSIS HALF-LIFE OF DIBENZOTHIOPHENE
-------
12.5.2 Oxidation Rate
The susceptibility of 0.50 yg ml"1 (2.7 x 10~6 M) DBT to free radical
oxidation was examined using the AA-initiated oxidation reaction. (See Part I,
Section 5.4.2 of this report and Appendix B for discussion and procedure.)
The experiment was carried out at 50.0°C in 1% acetonitrile in water contain-
ing 1.0 x 10"4 M AA. A solution of DBT without AA was maintained under iden-
tical reaction conditions as a control. At the end of 70 hours reaction time,
analyses of both solutions showed that about 50% of the DBT had been lost, and
no early-eluting (polar) products were observed in the reverse-phase HPLC anal-
ysis, as were found in the photolysis experiments. We conclude that DBT did
not oxidize and that the observed losses were probably due to volatilization.
In order to estimate a maximum value for the oxidation rate constant
kQX, it was assumed that the experimental error in the analysis was i5%. A
first-order rate constant for oxidation of DBT is then less than 2 x 10~7 sec"1
at 50°C. Under our reactions, this corresponds to a second-order rate constant
kox of <68 M"1 sec"1. At 25°C, kox is then <7.5 M"1 sec" , and with the assump-
tion that [R02-] = 10~9 M in the aquatic environment, the oxidation half-life of
DBT is over 3 years. The free radical oxidation of DBT is clearly not competi-
tive with volatilization or photolysis under environmental conditions.
12.5.3 Hydrolysis Rate
Since DBT contains no groups that are hydrolyzable, no hydrolysis studies
were carried out.
12.5.4 Products from Chemical Transformations
Photolysis of DBT gave five early-eluting, primary products in the HPLC
analysis. Attempts to characterize these products using GC/MS were unsuccessful.
However, when the reaction solution was degassed using freeze-thaw cycles be-
fore photolysis, the yield of the third eluting product was diminished and the
fifth eluting product was increased. These results suggest that both water and
oxygen may be involved in some product-determining steps. The photolysis reac-
tion products require further study.
12.6 BIODEGRADATION
12.6.1 Development of Biodegrading Cultures
Because naphthalene and anthracene could be expected to be present in
natural materials containing DBT, enrichment studies were initiated with DBT,
naphthalene in combination with DBT, and anthracene plus DBT as the organic
substrates added to the six water samples used in most of these studies. These
compounds were added at 10 pg ml-1. NHijNC^ was used in place of (NHi^SOi^ with
the phosphate buffers to dilute water samples 4:5. All enrichments were ini-
tiated with 5-liter volumes of diluted waters in the 9-liter fermentors. Incu-
bations were at 25°C, except for studies with Lake Tahoe water and cultures,
which were conducted at 15°C.
234
-------
Table 12.10 Indicates the times (days) at which biodegradation of DBT
was observed and the successes in developing cultures that would grow on basal
salts medium containing DBT as the sole carbon source. Biodegrading cultures
were developed successfully in all 9-liter fermentors containing DBT with
naphthalene or anthracene. When DBT was the only substrate added, biodegrada-
tion of DBT was observed within six weeks for all but two waters (Coyote Creek
and Lake Tahoe). Where degrading cultures developed with DBT as the sold carbon
substrate in these 9-liter fermentors, breakdown of DBT was observed much later
than in the samples containing naphthalene.
Transfers from the 9-liter fermentors were made to screw-capped Erlen-
meyer flashs containing basal-inorganic-salts medium containing DBT, DBT plus
naphthalene, or DBT plus anthracene, to correspond with the additives to the
9-liter fermentors. The basal salts medium was modified by substituting ni-
trate or chloride salts where sulfates were customarily used. Flasks were incu-
bated in rotary action shakers. Subtransfers were made at two- or three-day
intervals to media with.decreasing amounts of naphthalene or anthracene, if
these were used originally, and to media with increasing levels of DBT (up to
30-50 yg ml-1).
Degradation of DBT was accompanied by color changes in the media to
colorless to yellow to orange to a red insoluble product or products. Cell
counts in these media after two to three days incubation were (0.8 ~ 2) x 10s
cells yg"1 DBT. These yields were approximately 10% of those obtained with
methyl parathion or quinoline as the sole carbon source.
With the effluent from the Shell Oil refinery in the 9-liter fermentor
with DBT as the only added carbon source, degradation of DBT was observed
within four days. Transfers made at four, six, and ten days to basal salts/DBT
media did not result in DBT degradation in the shaker-flask fermentations. At
13 days, DBT equivalent to 10 yg ml"1 was again added to the 9-liter fermentor,
and a transfer made at 17 days resulted in a mixed-culture system that could
degrade DBT in the basal salts/DBT medium.
With cultures that grew on basal salts/DBT plus naphthalene media and
that were derived from Coyote Creek, aeration effluent from the South San Fran-
cisco sewage plant, and the effluent from the Shell Oil Refinery, several trans-
fers were necessary in these media before cultures could be obtained that would
utilize DBT as a sole carbon source.
Cultures were preserved in the vapor phase of liquid nitrogen after
suspension in 5% DMSO.
12.6.2 Biodegradation Kinetics
Pseudo-first-order kinetic studies were conducted with a DBT degrading
culture system derived from the eutrophic pond in Woodside, California. Two
liters of two-day-old fermentation broth developed in three transfers on basal
salts/30 yg ml-1 DBT medium were used. The final broth prepared by a 100% di-
lution of the prior culture with 30 yg DBT ml-1 medium was incubated only 18
hours and, at harvest, it contained 0.8 yg DBT ml-1. The combined broths
235
-------
TABLE 12.10. DEVELOPMENT OF DIBENZOTHIOPHENE
BIODEGRADING ENRICHMENT CULTURES
10
U)
Source of water sample
Aeration effluent from
Palo Alto Sewage Plant
Eutrophic pond,
Wood, CA
Coyote Creek, San Jose, CA
Aeration effluent from
South San Francisco plant
Aeration effluent from
Shell Oil Refinery, Martinez, CA
Lake Tahoe, CA
Compounds added
to water sample
DBT
DBT
DBT
DBT
DBT
DBT
DBT
DBT
DBT
DBT
DBT
DBT
+ naphthalene
+ anthracene
4- naphthalene
+ naphthalene
+ naphthalene
+ naphthalene
-t- naphthalene
Days for first
biodegradation
5
8
4
17
8
2
21
3
17
32
Development of
culture on salts
medium with DBT
Yes
Yes
Yes
Yes
Yes
Noa
Yes
Yesa
Yes
Yes
Nob
No
Two attempts at transfers from 9-liter fermentors.
Grew on salts medium with DBT plus naphthalene.
-------
were filtered through sterile Whatman No. 4 filter paper to remove insoluble
red metabolite(s). The filtrate was centrifuged, cells were washed twice with
basal salts medium, the final cell pellet was suspended in 0.05 volume of basal
salts medium, incubated on a shaker for 4 hours, centrifuged and again resus-
pended in 0.05 volume of basal salts medium. The stock suspension was used
at two dilutions in the kinetic studies with initial concentrations of 1000 ng
DBT ml"1. The stability in viable cell counts during the kinetic studies is
indicated in Table 12.11.
TABLE 12.11. VIABLE CELL COUNTS (cells ml" )
DURING PSEUDO-FIRST-ORDER BIODEGRADATION
KINETIC STUDIES WITH DIBENZOTHIOPHENE
Time
(hr)
o
io
120
150
180
220
Series
1
3.7 x 106
3.5 x 106
3.6 x 106
3.8 x 106
—
—
1.9
1.8
1.8
1.9
2.0
2
x 106
x 106
x 106
—
x 106
x 106
There was a short lag period before biodegradation of DBT was optimally
developed (Figure 12.7). The pseudo-first-order rate constants (k^) calculated
from the slopes of the plots of In DBT concentrations versus time were 1.85 hr"1
and 1.05 hr"1 for degradations initiated with 3.7 x 106 and 1.9 x 106 cells ml"1,
respectively. Consequently, the respective second-order rate constants were
5.0 x 10~7 and 5.5 x 10~7 ml cell"1 hr"1. Average k-„ value equals (5.3 * 0.4)
x 10-7 ml cell"1 hr-1.
12.6.3 Metabolites
The residues from the above kinetic studies were refrigerated for two
days "before they were extracted under neutral and acidic conditions, with ethyl
acetate. The concentrated extracts were analyzed by GC and HPLC. Preliminary
observations indicated better resolution in our GC system and more products in
the extract made from acidified broths than from the neutral broth. By GC, at
least 10 products were observed in the acid-extracted fraction. This fraction
was converted to the trimethylsilyl (TMS) derivatives and analyzed by GC-MS.
Structures for seven metabolites are indicated in Figure 12.8 and the
mass spectra for Compounds A, B, C, D, E, F, and G or their TMS derivatives
are presented in Figures 12.9 through 12.15. Compound A has been confirmed
as dibenzothiophene-5-oxide by comparison of the mass spectrum with that of a
reference sample. The identification of Compound B as 1,2-dihydroxy-l,2,3,4-
tetrahydrodibenzothiophene is suggested from our MS data, from the structures
237
-------
1000
800
600
400
£
01
c
I 200
Z
O
I—
ct
^-
ui
O
I !00
ui
Z
UJ
X
Q.
O
80
fe
N
Z
LU
CO
60
40
20
A
O
I
I
40
80 120
TIME — minutes
160
200
FIGURE 12.7. DIBENZOTHIOPHENE BIODEGRADATION IN BATCH FERMENTATION
WITH HIGH CELL COUNTS
Initial cell concentrations: O3.6 x 106 cells ml"1
A 1.9 x 106 cells ml'1
238
-------
of Compounds C, D, E, and F, and from the observation by Gibson (1975) that
c 1 s-1,2-dihydroxy-1,2-d ihyd rob enzot hiophene was a metabolite of DBT.
Compounds C and D were converted to their bis TMS derivatives and the
mass spectra of the separate GC fractions were interpreted as those from the
cis and trans isomers of 4[2-(3-hydroxy)-thianaphthenyl]-2-oxo-3-butenoic acid.
Compounds E, F, and G were assigned the structures 3-hydroxy-2-formylbenzothio-
phene, 3-hydroxy-benzothiophene, and 2,3-diketobenzothiophene, respectively.
The insoluble red product removed by filtration on Whatman No. 4 filter
paper in the process of preparing the culture for kinetic evaluations was not
studied.
ono
DBT
COMPOUND B
COMPOUNDS C AND D
(cis and trans isomers)
COMPOUND A
COMPOUND E
COMPOUND G
COMPOUND F
FIGURE 12.8 METABOLISM OF DIBENZOTHIOPHENE
239
-------
M
100-
90-
80-
Z 60-
LU
S 50-
LU
p 40-
5
i2 30-
20-
10-
. ,,|
I '
|
I
I i
, il,l,| L-liLiI
|""I»»|»«|WW|"
Ilin
II I
If 1 III 1 1
N! Hi . tin ill liiiillllLihiitiillLniii j iMnii.i
, ILiii
1 1
1
111,1,, ,,,!
"'' T '"' 1
M-
I
||, |
N
20
-0
I
I, .,i
'" I"1"
0
OoO
?
i
0
1, ,
' 1 ' 1 ' 1 ' 1 • 1 ' 1 ' 1 • 1 • 1 • 1
SO
100
150
200
m/e
250
300
FIGURE 12.9 MASS SPECTRUM OF COMPOUND A, DIBENZOTHIOPHENE-5-OXIDE
-------
10
100
90-
80-
£ 70-
i
g 60-
g
Ul
P 40-
<
U> 30-
oc
20-
10-
0-
90-
80-
t 70-
| 60-
>
H 40-
g 30-
20-
10-
n-
lill L ii 1 1 ll..,.l.i,i..lli.i.,.i..,.i
M-OTMS
rr~TT7~- 275
ll I
III! II 1 1 111 till .ll 1 . ll 1 , ll It 1 ll II 1 ,1. 1 ll . 1 1. 1 1 .1 II .11 III
50 100 150 200 250 300
m/e
OTMS
X£X .A^OTMS
M^SAJ
M (TMS2)
364
M-CH3
350
400
m/e
FIGURE 12.10 MASS SPECTRA OF COMPOUND B, BIS-TRIMETHYLSILYL ETHER
-------
275
to
K)
100-
90-
80-
t 70-
V)
£ 60-|
u 50-
| J
ID
CC 30'
20-
10-
0-
100-
90<
80-
60-
50
40
30
20
10
0
LU
..ui.li I ii..|ini|ilii|i.i.|jri|il imimn.m.!
•••—I—• .•—,-•,.-,—, r...,....f
SO
100
M-CH3
150
200
250
m/e
OTMS
C-OTMS
ISOMER 1
M
392
(TMS2)
M-COOTMS
300
\+»l(*,1
350 400 m/e
FIGURE 12.11 MASS SPECTRA OF COMPOUND C, BIS-TRIMETHYLSILYL DERIVATIVE, ISOMER 1
-------
100-
90-
80-
£ 70-
W
RELATIVE INTEh
u te at o>
0 0 O O
till
20-
10-
90-
80-
£ 70-
ul 60-
t-
- 60-
lil
I 4°"
530-
oc
20-
10-
0-
I. i . i li
1 • 1 • 1
\
1
50
}ll §
i f • \
V
M-CH3
It
. , -. | 1 1 I |
350
M-COOTMS
275
, ,
1 • i" • ' i "• " i i "• " '( " '""i •"" T"1 r""' — r""1" (• • ' i i -- •• - 1™ • -| - -•— |" • |- •••- -|- i •• • ••( • j • i
100 150 200 250 300
m/e
OTMS 0
rj£^s A £-C-OTMS
t^^l^ ^JL^J o
S ^v^
liomer 2
, (TMS2)
2
l|.
». ,
400 m/o
FIGURE 12.12 MASS SPECTRA OF COMPOUND D, BIS-TRIMETHYLSILYL DERIVATIVE. ISOMER 2
-------
>
55
LU
111
to —
*• £
*• 5
DC
100-
90-
80-
70-
60-
50-
40-
30-
20-
10-
JllL ItJl
1 * 1 " 1
TMS
73
I, IU,,J,L,I .J«(
III I-. .1 ,l.,l ....da.. .1.11 ..... ... 1
50 100 150 200
m/e
III 1 1 lu
-^^ OTMS
fy^*vi >j
S
o
M
250
. .1. .
1 • 1 • 1 1 • '1 • 1 "• 1 ""•- 1 • 1
250 300
FIGURE 12.13 MASS SPECTRUM OF COMPOUND E, TRIMETHYLSILYL ETHER
-------
10
en
100-
90-
80-
H TO-
OT
2 60-
20-
10-
0-
1 "•
60
M
222
M-CH3
207
100
150
200
250
300
m;e
FIGURE 12.14 MASS SPECTRUM OF COMPOUND F, TRIMETHYLSILYL ETHER
-------
136
cn
1-
S
i
01
H
3
Ul
cc
100-
90-
80-
70-
60-
60-
40-
30-
20-
10-
„ III II,!,'
i ' * i ' i
E
10
,. .Ill || ||
UMIHI-llllllHUllJIIIIIIjlllll..!.,!!!..!..!!...!!!!
0 100
8
HIl.ilJ.I i ll InUiii
I1 I"1 I""1
o
'OCXo
M
164
III II Illll II
160 200 260 300
m/e
FIGURE 12.15 MASS SPECTRUM OF COMPOUND G
-------
12.6.4 Discussion
The development of DBT biodegrading mixtures of cultures by our enrich-
ment process revealed several interesting features that may exist when contami-
nants containing DBT are introduced into aquatic systems. When more readily
biodegradable and chemically related compounds that coexist with DBT in petro-
leum or coal tar products were added, DBT biodegrading culture systems were more
readily developed. In two cases, the addition of only DBT to the water sample
did not result in DBT biodegradation. The difficulties encountered in several
instances in developing cultures that could degrade DBT when transfers were made
from media containing DBT plus naphthalene also indicate the complexity of eval-
uating the biodegradability of a compound. Undoubtedly, induction and cometab-
olism are important phenomena that operate in natural waters.
The kinetic rate constants developed with a mixed culture system indi-
cate that if such cultures exist or can be developed, DBT is readily degraded.
Since the thiophenes exhibit toxic and phyiological activities, a knowl-
edge of the characteristics of the metabolites can be of considerable signifi-
cance. Although not all the observed metabolites have been studied, six chemical
entities have been tentatively identified and the structure of one product has
been confirmed by comparing its mass spectrum with a reference sample.
Basically, our data are in agreement or related to those of Kodama
et al. (1970, 1973), Laborde and Gibson (1975), Gibson (1975), and Hou and
Laskin (1975). Our determination of Compound B as a 1,2-dihydroxy-l,2,3,4-
tetrahydrodibenzothiophene as compared with the 1,2-dihydroxy-l,2-dihydrobenzo-
thiophene reported by Gibson (1975) may be due to differences in the cultures
involved or to possible reduction at the 3,4 positions occurring between the
time of the kinetic study and the extraction, when the 1,2-dihydro compound
may have functioned as a hydrogen acceptor. The observation of the cis and
trans forms of 4[2-(3-hydroxy)-thianaphthenyl)-2-oxo-butenoic acid] may be due
to a cis-trans isomerase (Kodama et al., 1973) or to conversion under GC condi-
tions. Bioconversions to Compounds E, F, and G are logical. It is possible
that 2,3-dihydroxybenzothiophene is formed as an intermediate between Compounds
F and G and that this dihydroxy compound or Compound G may undergo ring fission.
Compound G was not previously reported as a metabolite of DBT. The insoluble
red product present at these high dilutions of substrate may be similar to the
Compounds A and B reported by Hou and Laskin (1975) or may be dimeric forms of
DBT metabolites.
Many questions remain regarding DBT degradation pathways and these could
be examined by using mixed cultures, pure cultures, or mutant cultures; by iso-
lating metabolites at ambient temperatures; and by conducting material balances.
The naturally occurring alkylated derivatives of DBT would of course have differ-
ent metabolic pathways because of the influences of the alkyl groups. In some
cases, the alkyl groups would be attacked first and such compounds could be more
readily removed from water columns. On the other hand, branched alkyl substitu-
ents on both benzene rings could increase the resistance of such compounds to
biodegradation.
247
-------
13. LABORATORY INVESTIGATION OF METHYL PARATHION
13.1 SYNOPSIS
The results of our laboratory studies suggest that biodegradation is
the major environmental pathway for methyl parathion (MP) in eutrophic systems.
Sorption, photolysis, and hydrolysis are important and perhaps dominant trans-
formation processes in oligotrophic systems. There is some evidence that the
rate of hydrolysis of MP is accelerated by suspended sediments, but conclusive
measurements were not conducted. The rate of volatilization of dissolved MP
is not significant compared with the rates of chemical and biodegradation
processes.
P_-Nitrophenol was identified as a biodegradation product but was rapidly
degraded in the one mixed culture system examined. 0-Methyl-0-p_-nitrophenyl-
thiophosphoric acid and p_-nitrophenol are the major products from neutral-
(pH <8) and base-catalyzed hydrolysis (pH >8), respectively; the same products
were found in the photohydrolysis of MP. Transformation of MP to paraoxon was
not observed.
The results obtained from the nine-compartment computer model supported
the conclusions reached on the basis of the laboratory rate data. They also
suggested that physical sorption and biodegradation may be significant in
bottom sediments. The model predicted the following steady-state concentra-
tions of MP in solution, suspended solids, and sediments near point sources in
the presence of a continuous discharge of 1 pg ml"1 (1 ppm) methyl parathion.
^ Suspended
Half-life Solution solids Sediments
(hr) dig ml"1) pig g"1) (ug g"1)
River 0.6 0.989 49.46 49.1
Pond 27.3 1.2 x 10~2 0.60 0.60
Eutrophic
lake 28.3 0.06 3.00 0.51
Oligotrophic
lake 151.6 0.098 4.90 1.40
Predicted by one-compartment model.
248
-------
13.2 BACKGROUND
Table 13.1 gives the physical properties and nomenclature of methyl
parathion, which is a broad spectrum insecticide widely used in cotton cul-
tivation in the United States. It is highly toxic to humans and most other
nontarget organisms and may be sorbed by ingestion, by skin contact, or by
inhalation. It is nonpersistent, requiring 3 to 10 applications per season
on cotton and 1 to 4 applications per season on other crops. It is a serious
hazard to persons or animals entering treated areas within 48 hours of appli-
cation (von Rumker et al., 1974). However, methyl parathion disappears rapidly
in both soils and water and does not appear to have had significant environmen-
tal impacts away from the sites of use.
TABLE 13.1. PHYSICAL PROPERTIES OF METHYL PARATHION
: - _
/r\\ H
02N(()/-0-P-(OCH3)2
Molecular weight 263.12
Melting point 36°C
Vapor pressure at 20°C 9.1 x 10~3 torr
Solubility in water, ug ml"1 50 ng ml"1 or 2.1 x 10~4 M
Chemical name 0,0-Dimethyl-0-j>-nitrophenyl-
phosphorothioate
Tradenames Dalf, Folidol M., Metron,
Nitox 80, Putron M,
Tekovaisa
Previous assessments of the environmental fate of methyl parathion have
relied heavily on analogies to ethyl parathion and have consisted of a simple
documentation of the apparent rate of disappearance from fields and selected
water samples. King and McCarty (1968) reported sorption data on soils for
both MP and parathion. Saltzman and Yaron (1972) reported sorption data for
parathion on both clays and soils. All these workers found that the sorption
partition coefficient, IL>, varied from 5 to 30, that K^ was roughly propor-
tional to the soil organic content, and that n in the Freundlich isotherm
equation was approximately 1.
Limited information was available for assessing the chemistry of methyl
parathion in aquatic environments. However, these data, along with those
reported on parathion indicated that chemical transformation might be important
under some environmental conditions.
249
-------
The hydrolysis of methyl parathion has been studied under a variety of
conditions as summarized below.
Conditions Half-life Reference
"pH about 6," room temperature 1.7 weeks Cowart et al. (1971)
pH 12, 37.5°C 2.1 hours Jaglan and Gunther (1970)
pH 6 buffer, 70°C, 20% ethanol 8.4 hours Ruzicka et al. (1967)
Low concentrations of copper ion have been reported to catalyze the hydrolysis
of parathion and paraoxon (Ketalaar et al., 1956). The hydrolysis products of
methyl parathion itself are reported to be _g-nitrophenol and dimethyl phosphoric
acid. It was recently reported that hydrolysis of parathion in acidic or
neutral solutions also gives the dealkylated ester, 0-ethyl-0-p_-nitrophenylthio-
phosphoric acid, and ethanol (Weber, 1976).
Oxidation of parathion esters using permanganate, chlorine, silver oxide,
ozone, and other chemicals has been reported (Gunther et al., 1968, 1970; Comma
and Faust, 1972). The latter paper reports an acceleration of the oxidation
rate with increased pH values. The products identified in these reactions were
paraoxon and ja-nitrophenol.
While no work on the photochemistry of methyl parathion in aqueous solu-
tion has been reported, literature on the phototransformation of parathion is
relevant because the structures are similar and the UV spectra of the two com-
pounds are almost identical (the UV spectra of both parathion and methyl para-
thion show Amax 274 nm, log e 4.0 and Xmax 350 nm, log e 2.4) (Sandi, 1959).
Paraoxon and j>-nitrophenol have been reported to result from photolysis of
parathion (Cook and Pugh, 1957; Grunwell and Erickson, 1973). The latter paper
also reported that parathion did not oxidize in the presence of photogenerated
singlet oxygen. Contrary to the above findings, paraoxon was not found in a
study on the photolysis of methyl parathion on soils (Baker and Applegate,
1970).
Although the above information indicated that methyl parathion would
undergo hydrolysis, photolysis, and possibly oxidation, no reliable data were
available for evaluating the rates or kinetics of these reactions in aquatic
environments. Furthermore, the information available was both incomplete and
confusing with regard to the identity and yields of products obtained in the
respective transformation reactions. Since paraoxon and methyl paraoxon are
more toxic than the parent thioesters themselves, particular attention was
directed toward determining whether methyl paraoxon is formed from oxidation
or photolysis of methyl parathion.
250
-------
Recent research studies conducted in the laboratories of Hsieh and
Sethunathan^ clearly indicate the biodegradability of parathion. Lichtenstein
and Schulz (1964) in studies with soils and Graetz et al. (1970) using pure
cultures concluded that the breakdown of parathion involves reduction of nitro
groups to aminoparathion (0,0-diethyl-0-p_-aminophenylphosphorothioate).
In their publication on pathways of microbial metabolism of parathion,
Munnecke and Hsieh (1976) discussed the many potential intermediate aromatic
products and diethylthiophosphoric acid formed in the decomposition of para-
thion by a mixed bacterial culture (minimum of nine isolates) fermentation with
technical parathion. The intermediates were determined largely by the degree
of the aerobic state of the fermentation. p_-Nitrophenol and diethylthiophos-
phoric acid were primary degradation products in the oxidative pathway.
Munnecke and Hsieh (1976) indicated that, in most studies on biodegradative
mechanisms, insufficient substrate was used for identification of metabolites.
A characteristic of their parathion degradation system was the release of
small amounts of parathion hydrolase into the medium. They also observed that
their hydrolase preparation had an optimal pH of 9.2. All studies with mixed
cultures indicated that biodegradation of parathion proceeded until no aromatic
ultraviolet-absorbing material remained. The parathion hydrolase was studied
in greater detail by Munnecke (1976) for possible pesticide disposal in plant
operations. This investigation included MP and five other pesticides.
The correlation of parathion fixation on soils with the soil organic
content (Kliger and Yaron, 1973; D. P. H. Hsieh, private communications, 1976)
suggest that methyl parathion may be sorbed by cellular material, including
microbial cells. This is environmentally important because if a substance is
sorbed by nonbiodegrading organisms (or sediment), its availability for decom-
position by organisms that are capable of biodegrading it is decreased.
Also, digestion of free floating cellular material containing the substance
by cells that only partially degrade it, or do not degrade it, could result in
food chain magnification. The alga Chlprella pyrenoidosa has been reported to
biodegrade parathion (Zuckerman et al., 1970). The very fact that algae can
sorb parathion poses a problem for food chain magnification or the deposition
of the pesticide in the detritus and eventual degradation in the sediment.
Gregory et al. (1969) reported the uptake of parathion by Euglena gracilis
(a "flagellate"), two species of algae, and two species of ciliates. Their
data implied a sorption partition coefficient of 2.2 x 103 when their Euglena
were exposed to 1 yg ml"1 parathion for 7 days. Kortus et al. (1971) studied
the uptake of 32P-labelled parathion by Euglena gracilis and observed 69% up-
take of the pesticide in 15 minutes.
A Midwest Research Institute report to the EPA (1975) reviewed the sorp-
tion of parathion by microbes and indicated that parathion sorption may be
independent of life processes since the sorption characteristics of live and
Munnecke and Hsieh (1974, 1975, 1976), Hsieh and Munnecke (1972).
^Siddaramappa et al. (1973), Sethunathan (1973), Sethunathan and Yoshida (1973)
251
-------
dead organisms showed little difference. This report also reviewed the effects
of parathion on microorganisms and parathion biodegradation.
Several reports were previously published on the biodegradation of MP
(Lichtenstein and Schulze, 1964; Yasuno et al., 1965; Maleszewska, 1974;
Naumann, 1967). These include studies on waters and soils. However, these
studies are not of the type that could be used in a multidisciplinary overall
assessment of the fate of MP in oligotrophic, eutrophic, and waste waters.
13.3 ENVIRONMENTAL ASSESSMENT
13.3.1 Summary of Laboratory Data
\
The rate constants obtained in the laboratory investigations of the
transformation and transport processes for MP are summarized in Table 13.2.
TABLE 13.2. SUMMARY OF METHYL PARATHION LABORATORY DATA
Process
Sorption equilibrium
Partition coefficient
Sorption
Volatilization
Photolysis
Oxidation
Hydrolysis
S = K S
s p w
K
= 50a
P
Rate expression
k [MP]
k [MP] = !-*• ^
Ic, „ — j_. / x JLU ug cej-j. nr
Average for all sediments
See discussion in Part I, Section 5.3, and Appendix B.
•»
"Early January.
JAt pH < 8.
252
-------
13.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved MP calculated for individual transformation
or removal processes following a spill are listed in Table 13.3. Virtually
all the MP is predicted to remain in solution and is predicted to be rapidly
transformed by microbes in eutrophic waters. Photolysis is predicted to domi-
nate transformations in oligotrophic waters. Half-lives of roughly 6.5 days
and 1 day are expected in oligotrcphic and eutrophic waters, respectively.
Dilution effects are important only in rivers.
13.3.3 Persistence
Methyl parathion is not a "persistent" compound in eutrophic waters
because it is rapidly transformed in solution. In clear, biologically unpro-
TABLE 13.3. TRANSFORMATION AND TRANSPORT OF METHYL PARATHION
PREDICTED BY THE ONE-COMPARTMENT MODEL
Eutrophic Eutrophic Oligotrophic
River pond lake lake
Photolysis, half-life (hr)a 340 850 850 170
Oxidation, half-life (hr) >2 x 10s >2 x 10s >2 x 105 >2 x 10s
Volatilization, half-life (hr) >4 x 10s >5 x 106 >2 x 106 >2 x 106
Hydrolysis, half-life (hr) 1700 1700 1700 1700
Biodegradation, half-life (hr) 40 40 40 >10,000
Half-life for all processes >7 28.5 28.5 157.5
except dilution (hr)
Half-life for all processes 0.6 27.3 28.3 151.6
including dilution
Amount methyl parathion 5 15 2.5
sprbed (mg m~ )k
Percentage methyl parathion 0.5% 1.5% 0.25%
sorbed
2.5
2.25%
Estimates are the average photolysis rates on a summer day at 40° latitude.
Photolysis rates in midwinter are at least three times slower.
b
1 pg ml"1 of methyl parathion in aqueous phase and partition coefficient of
50 are assumed.
253
-------
ductive (oligotrophic) waters, however, MP will be more persistent because of
the small microbial populations.
13.3.4 Mass and Concentration Distributions Calculated Using the Nine-
Compartment Computer Model
The pseudo-first-order rate constants used in these simulations are
presented in Appendix A. The distributions of mass and concentration of MP
expected at steady state during chronic discharge to each of four types of
water bodies are given in Table 13.4. Figures 13.1 to 13.4 show the dynamics
of changes in concentration for the pond, river, and lake simulations. Changes
in the MP concentrations in suspended solids parallel changes in concentration
of dissolved MP, and consequently are shown only for the pond simulation to
simplify the figures.
Despite the marked tendency of MP to remain in solution, roughly half
of the MP is in the sediments at steady state (Table 13.4), under conditions
of continuous discharge, with the possible exception of oligotrophic lakes.
In all simulations, concentrations are 10 to 100 times higher on the suspended
solids and in the sediments than in solution. After discharges are stopped in
the simulations, MP concentrations decline most rapidly in the aqueous phases
of the river and eutrophic lake simulations. The concentrations scarcely
change in the sediments of any of the simulations because the scouring rate of
the sediments is low (t, /2 ^ 1700 hr).
13.3.5 Discussion
The major influx of MP into natural waters should occur shortly after
application either as sprays, drift, or run-off, possibly with a large fraction
sorbed to soil particles. If so, the degradation rates should most nearly
resemble the pond simulation, which assumes large populations of metabolically
active microbes and high turbidity (Figure 13.1). The much smaller influx of
MP that can be expected in winter, when biodegradation rates are suppressed,
probably enters mainly in the sorbed form. If so, most of the MP is probably
removed from the water column by sedimentation before undergoing substantial
degradation, and then degrades slowly by hydrolysis or biodegradation^
The fraction that enters as a solution during the winter months prob-
ably undergoes moderate degradation by photolysis and hydrolysis before being
transported to the sediments. If so, degradation in eutrophic waters during
the winter should roughly approximate the rates for compartment 1 of the oligo-
trophic lake simulation, which shows the maximum contribution of abiotic
degradation mechanisms (Figure 13.4). Given these assumptions, biodegradation
should account for roughly 1000 times more degradation of MP in actual waters
than any other form of degradation, and most of the degradation should occur
in the summer.
The rapid attainment of steady-state values with respect to degradation
rates and MP accumulation in solution and on solids shown by the simulations
is surely excessive, perhaps a factor of ten too rapid, because of the assump-
tion of frequent mixing with each compartment. However, the sequence in which
254
-------
TABLE 13.4. DISTRIBUTION OF METHYL PARATHION IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations of 1 ug ml"1 Methyl Parathion)
Pond
Mass Cone.
(kg) (ug g'1)
Compartment 1
(surface water)
Solution 0.75 0.38 x 10" '
Suspended solids 0.10 x 10"' 1.74
Compartment 2
(surface water)
Solution — —
Suspended solids ~ —
Compartment 3
(surface water)
Solution
Suspended solids — --
Compartment 5
(bottom water)
Solution
Suspended solids
Compartments 7-9
(sediment)8
Solution 0.95 x 10"2 0.38 x 10"'
Solids 1.14 1.74
River
2
1
2
1
2
1
7
9
Mass
(kg)
.96 x 102
.47
.93 x 102
.46
.91 x 102
.45
-
.34
.84 x 102
(
9.
4.
9.
4.
9.
4.
9.
4.
Cone.
Ug g"1)
87 x 10"1
90 x 10
79 x 10"1
80 x 10
71 x 10"'
83 x 10
—
79 x 10-1
86 x 10
1.
3.
2.
0.
1.
0.
0.
0.
0.
1.
Eu trophic
Mass
(kg)
51 x 10
79 x 10~a
08 x 10
62 x 10~*
31
32 x 10"2
60
15 x 10~2
14
11 x 10
6
3
1
0
5
0
2
0
0
0
Lake
Cone.
(ug g~l)
.06 x 10"2
.03
.00 x 10"*
.50
.24 x 10~3
.26
.40 x 10""
.012
.16 x 10~2
.82
Oligotrophic Lake
3
0
1
0
1
0
1
0
0
2
Mass
(kg)
.0 x 10
.75 x 10-'
.30 x 102
.32
.12 x 10
.28 x 10"1
.05 x 10
.26 x 10"1
.37
.92 x 10
Cone.
(ug g"1)
1.20 x 10~l
6.00
5.20 xlO-2
2.60
4.50 x 10"2
2.25
4.20 x 10"3
0.21
4.42 x 10"'
2.14
aThe amounts given for solid and solution phases in the sediment compartments are estimated from the sorption partition coefficient for
suspended solids and may be overestimated because it was assumed that biodegradation of sorbed material does not occur.
-------
10
en
cn
50
100 150
TIME — hours
200
250
FIGURE 13.1 PERSISTENCE OF METHYL PARATHION IN A TWO-COMPARTMENT POND SYSTEM
-------
UJ
O
z
o
o
I ,n-2 _
o
I
en
<
Q.
I
t-
LU
10'5 —
10'
TIME - hours
FIGURE 13.2 PERSISTENCE OF MrTH> L PARATHION IN A PARTIALLY
MIXED RIVER SYSTEM
257
-------
100
200
300
400
500
600
700 720
TIME — hours
FIGURE 13.3 PERSISTENCE OF METHYL PARATHION IN A EUTROPHIC LAKE
258
-------
3 x 10"1
10-'
10'2
10-3
2
O
10-4
I
UJ
<
nc
Ul
u
§
O
10's
10'7
COMPARTMENT
1
100
SOLUTION
SEDIMENTS
I
I
200
300 400
TIME - hours
500
600
700 720
FIGURE 13.4 PERSISTENCE OF METHYL PARATHION IN AN OLfGOTROPHIC LAKE
259
-------
the contents of MP, and hence the degradation rates of MP, in the various com-
partments reach steady state in the simulations should be accurate. The
predicted sequence shows steady states appearing first in the well-mixed
aqueous compartments near the source, then in the more remote aqueous compart-
ments, and last in the sediments, which are not only remote from the source,
but are characterized by a high mass of solids and hence change slowly.
On balance it is probable that MP concentrations under the chronic,
low influx conditions assumed will be below toxic levels for vertebrates
(LC50 generally greater than 5 yg ml~l) except in oligotrophic lakes, or
possibly, in eutrophic lakes during winter when microbial activity is depressed
by low temperatures. However, concentrations in all waters may be toxic to
invertebrates (LC50 of 2 ng ml"1 to 1 ug ml"1) under the conditions assumed.
The risk should be consistently greater in shallow waters, but even here it
should be fairly small since degradation appears to be virtually complete
within weeks in both soil (Knutson et al., 1971) and contaminated waters
(Yasuno et al., 1965; Hirakoso, 1968; Eichelberger and Lichtenberg, 1971).
13.4 PHYSICAL PROPERTIES
13.4.1 Solubility in Water
The solubility of MP in water is 50 yg ml"1 (AAPCO, 1959)-
13.4.2 Absorption Spectrum
The absorption spectrum was measured in the solar region (295 to
800 nm). The spectra at pH 5.1, 6.8, and 8.9 were identical within experi-
mental error (± 2%). The data reported in Table 13.5 are the average absorp-
tion coefficients as described in Part I, Section 5.2, and Appendix B.
13.4.3 Volatilization
The volatilization rate of MP was measured by the method of Hill et
al. (1976), which is described in Appendix B. The oxygen reaeration rate
ranged from 5 to 23 hr"1, which corresponds roughly to the turbulence of
a waterfall. Each experimental point for the MP concentration is the average
of replicate analyses of replicate samples and was corrected for evaporation
of water. Within experimental error, which is about 5%, there was no decrease
in MP concentration in 148 hours. ,
The MP volatilization data were fit to an exponential decay equation,
using the Hewlett-Packard Model 65 routine. The volatilization rate constants
were -0.013 day 1 for beaker A and +0.0017 day"1 for beaker B. Since these
rate constants are equal to zero within experimental error and since the ratio
of the volatilization rate constant to the oxygen reaeration rate constant is
less than 4 x 10 , we concluded that volatilization was not an important
environmental pathway for MP.
260
-------
TABLE 13.5. ABSORPTION SPECTRUM OF METHYL PARATHION IN WATER3
Center of
wavelength interval
(nm)
297.5
300.0
302.5
305.0
307.5
310.0=
312.5
315.0
317.5
320.0
323.1
330.0
340.0
350.0
360.0
370.0
380.0
390.0
400.0
410.0
Average
absorption coefficient0
(M~l cm'1)
6040
5460
4930
4310
3700
3210
2760
2290
1920
1630
1310
933
568
374
244
145
82
45
9
0
3Methyl parathion concentration 7.81 x 10 ^ mole liter
Wavelength intervals are given in Appendix B, Table B.I.
CThe absorption coefficients at 313.0 nm and 366.0 nm are 2670 and
185, respectively.
261
-------
13.4.4 Sorption on Clay and Sediments
Methyl parathion sorptlon isotherms were measured on six natural sedi-
ments and Ca-montmorillonite. Preliminary kinetic experiments were conducted
using sediments from Coyote Creek, Navarro River, and Searsville Pond with no
pH adjustment. In measuring isotherms starting from 7.5 and 8 yg ml"1 sediment,
we measured concentrations of MP in the supernatant after 16 and 22 hours of
equilibration with the sediment. For all three sediments, no significant dif-
ference in supernatant concentration was detected at 16 and 22 hours, and it
was concluded that sorption was complete in 16 hours. All subsequent isotherm
measurements used 16-hour contact times to determine sorption partition
coefficients.
The partition coefficient data, calculated by linear least squares
regression using supernatant concentrations only, are summarized in Table 13.6.
The partition coefficient for MP on Searsville sediment was also calculated
using the nonlinear least squares regression (see Appendix B )
giving essentially the same value for Kp. Some parathion data have been
included for comparison (Saltzman and Yaron, 1972). There are no striking
differences in the sorption characteristi -s of these sediments. The Spearman
rank correlation test (Langley, 1971) showed no correlation of the partition
coefficient with either total organic carbon (TOC) or cation exchange capacity
(CEC). Also, these sediments came from different regions of the United States
and from several types of water bodies. The TOC and CEC values also vary over
a wide range.
Saltzman and Yaron (1972) observed that the value of Kp in the Freund-
lich isotherm for parathion increased as the organic content of the sediment
increased. Therefore, our observation that the partition coefficient was
fairly constant and independent of the organic content of the sediments studied
here (Table 13.6) was surprising. However, the concentrations of MP used in
this study were considerably lower than the parathion concentrations reported
by Saltzman and Yaron (1972) and this may explain the differences.
In one experiment, a sorption isotherm of MP on Coyote Creek sediment
was measured in the same way as above except that the sediment was subjected
to three freeze-thaw cycles before exposure to MP. The freeze-thaw cycle
breaks up cellular material that might be contained in the sediment. No effect
on the partition coefficient was observed, suggesting that the effect of cellu-
lar material on the overall value of the partition coefficient is small with
Coyote Creek sediment.
Sorptioi'i Kinetics—A study of the rate of sorption of MP was conducted
with Coyote Creek sediment. Initial conditions were 3.5 yg ml"1 MP.and sedi-
ment levels of 5.9 mg ml"1 and 11.9 .Tig ml"1. The pH was adjusted to 6 at the
begiuning of the experiment. The coucentration of MP in the supernatant was
measured every 15 minutes for the first 2 hours and then less frequently for
a total contact period of 27 hours. Replicate samples and replicate analyses
were used for each measurement.
The data illustrated in Figure 13.5 show a rapid drop in MP followed
by a slower drop that continues for at least 24 additional hours. It is not
262
-------
possible to determine from these data whether the drop continues after 24 hours
or whether an equilibrium is established. The concentrations in the superna-
tant measured at 1 hour and estimated for 16 hours of contact give partition
coefficients of 39 and 83, respectively, which are comparable to the Coyote
Creek sediment isotherm run previously (K = 51 ± 5, Table 13.6).
The slowly decreasing part of the curve in Figure 13.5 was analyzed
assuming that the decay was first order in MP and sediment concentration,
TABLE 13.6. METHYL PARATHION SORPTION ON SEDIMENTS
Sediments
Ca-
Montmor-
illonite
clay
sorption
Coyote
Creek
sorption
Des Moines
River
sorption
Navarro
River
sorption
Oconee
River
sorption
Searsville
Pond
sorption
Montmor-
illonited
I Sediment
Total organic ; cone.
carbon (%) ; (mg ml"1)
8.0
0.06 16
4.2
13
7.5
1.4 15
3.7
11
7.7
0.6 15
3.8
12
8.1
0.4 16
3.7
11
7.9
0.6 17
4.7
13
7.7
3.8 16
4.2
13
MP
concentration
in supernatant3
(ug ml"1)
5.7 ± 0.3
4.7 ± 0.2
3.2 ± 0.1
2.6. ± 0.1
5.5 ± 0.4
4.5 ± 0.1
3.2 ± 0.1
2.3 + 0.2
5.8 ± 0.2
4.7 ± 0.1
3.2 ± 0.1
2.4 + 0.1
6.2 ± 0.5
4.9 ± 0.3
3.5 ± 0.3
2.6 ± O.I
6.3 ± 0.2
5.3 ± 0.1
3.2 ± 0.0
2.7 ± 0.0
5.5 ± 0.4
3.7 ± 0.1
3.1 ± 0.1
1.9 ± 0.2
MP
. Partition
concentration
on sedimentb LLSC
(Vg g *) a0 = 0
280 ± 40
200 ±10 46+2
160 ± 20 (a0
100 ± 10
270 ± 50
200 ±10 51 + 5
210 ± 40 (a0
140 ± 20
230 ± 20
190 ±10 42+2
160 ± 20 (a0
120 ± 10
310 ± 60
230 ±20 48+3
160 ± 70 . (a0
130 ± 10
140 ± 20
120+3 51 ± 5
89+9 (ao
73 ± 3
310 + 50
270 ±10 60 ± 4
190 ± 30 (a0
160 + 10
100
coefficient, K
NLLSC
a0 f 0 Sy, only
49 ± 6
- -11 ± 27)
28 ± 12
= 99 ± 49)
29 ± 4
= 59 ± 19)
41 ± 12
= 30 ± 59)
17 ± 3
=34 ±3)
38 ± 8 65 ± 4
= 100 ± 30)
Soil"
12
Concentration measured in supernatant with population standard deviation.
Concentration on sediment calculated from supernatant concentration with population standard deviation.
CLLS = linear least squares; NLLS = non linear least squares. See Appendix B,..Section B.I.4 for description o
regressions. Limits are 95% confidence limit.
Data for parathion (Saltzman and Yaron, 1972).
263
-------
= ka[MP](Sed)
(13.1)
where:
[MP] = MP concentration (yg ml"1)
(Sed) = sediment concentration (mg liter"1)
The result of the analysis was ka = 2.0 (± 0.2) x 10~6 ml pg 1 MP hr 1. This
yields a pseudo-first-order decay constant for MP of 0.012 hr""1 when the sedi-
ment concentration is 5900 mg liter"1 and 0.022 hr"1 when the sediment concen-
tration is 11,900 mg liter"1. The half-lives are 58 hours and 29 hours,
respectively.
The slow disappearance of MP could be caused by biological activity,
sediment-catalyzed hydrolysis, or continued sorption by a mechanism different
from the one that obtains in the first hour. Biodegradation is unlikely
since MP solutions inoculated with Coyote Creek water show a lag of 5 days
for adaptation before any degradation occurs (Section 13.6).
No sediment
5.9 fig ml"1 Coyote Creek sediment
11.9 fjg ml"1 Coyote Creek sediment
I I I I
FIGURE 13.5 CONCENTRATION OF METHYL PARATHlON IN SUPERNATANT
264
-------
The following observations indicate that clay-catalyzed hydrolysis is
also unlikely. Although no analyses for hydrolysis products were performed
during our kinetic experiment, evidence for hydrolysis in other experiments per-
mits an estimate of sediment-catalyzed hydrolysis rates. In the experiment to
measure a desorption isotherm (described below), MP (26 yg ml"1) was in contact
for 70 hours with Coyote Creek sediment at 12,100 mg liter"1. An analysis of
the supernatant showed, at most, 10% hydrolysis of MP to a single product, 0-
methyl-O-p-nitrophenylthiophosphoric acid ester.''" £-Nitrophenol was not
observed in the supernatant. If we assume that the hydrolysis rate is first
order in both sediment and MP, this is equivalent to a half-life of 980 hours
if the sediment concentration were 5900 mg liter"1 or 490 hours for a concen-
tration of 11,900 mg liter"1. These half-lives are much longer than the half-
lives for MP disappearance observed during the adsorption kinetic experiment
(58 and 29 hours). The half-life for the solution-phase neutral hydrolysis was
shown to be about 2100 hours (Section 13.5). There does appear to be about a
fourfold increase in the neutral hydrolysis rate of MP in the presence of
Coyote Creek sediment under the conditions of the experiment. However, neither
neutral nor clay-catalyzed hydrolysis seems to be fast enough to account for
all the observed slow decrease in MP. The data suggest that about 95% of the
decrease in MP is due to continued sorption. The rapid initial sorption could
be due to organic material and readily accessible clay surfaces. The slow
sorption may involve diffusion into the interlaminar portion of the clay frac-
tion of the sediment.
Desorption—Desorption of MP was measured by first contacting Coyote
sediment at 12,100 mg liter""1 with MP at 26 yg ml"1 at pH 6. A 70-hour contact
time was chosen in an attempt to ensure full sorption equilibrium. The parti-
tion coefficient that resulted from this contact was 60, based on the assump-
tion that all MP that had disappeared from the supernatant was adsorbed into
the sediment. Analysis of the supernatant indicated that approximately 10%
hydrolysis to 0-methyl-0-£-nitrophenylthiophosphoric acid ester occurred during
the contact period. If the hydrolysis is included but we assume that the
monoester is not strongly sorbed, then the MP partition coefficient is 47.
Either of these estimates is consistent with previous determinations for MP
sorption onto Coyote Creek sediment.
The sediment containing its sorbed MP was washed and then resuspended
in water and shaken for eight hours. No significant hydrolysis would occur in
this brief contact period. Figure 13.6 shows the supernatant concentrations
determined during the contact period. Desorption equilibrium was obtained in
4 hours of contact. The data do not show any changes in supernatant concen-
tration after four hours, indicating that desorption is complete.
Satlzman, Mingelgrin, and Yaron (1976) demonstrated that oven-dried kaolinite
accelerates the hydrolysis of both parathion and methyl parathion. Although
the presence of small amounts of sorbed water increased the hydrolysis rates,
larger amounts of water reduced the effect. No experiments on clay suspended
in water containing dissolved MP were reported.
This product should not be strongly sorbed because it is an acid and should
be very soluble in water.
265
-------
Figure 13.6 also shows the supernatant concentrations that would be
expected if desorption had the same partition coefficient as sorption (Kp =
51 ± 5) and if all the sorbed MP could be desorbed (corrections have been made
for material lost during the washing operation). Sorption-desorption of MP
on this sediment is not reversible. If we assume that the partition coeffi-
cient for sorption and desorption is the same, and about 50, these results
suggest that approximately 50% of the sorbed MP was irreversibly sorbed and
unavailable for desorption.
13.4.5 Biosorption
The results of three experiments with viable heat-killed (100°C for
15 minutes) bacterial cell mixtures are given in Table 13.7. The bacterial
mixtures contained equal optical densities (2) of Azotobacter beijerinckii
ATCC 19366, Bacillus cereus ATCC 11778, Escherichia coli ATCC 9637, and
Serratia marcesens ATCC 13880. The recoveries of MP in sorptions with viable
cell mixtures incubated at 25°C averaged 76%. Since no yellow coloration due
to p-nitrophenol was observed, the degradation by this cell mixture was not
restricted to a hydrolysis to p_-nitrophenol. With viable cells incubated at
2°C and with heat-killed cells incubated at 25°C, recoveries of 100% were
li
°o^ E 3
5 > :*-
D I
a
MP CONCENTRATION AT EQUILIBRIUM
IF THE DESORPTION COEFFICIENT
WERE 51
MEASURED SUPERNATANT
CONCENTRATION
4 5
TIME - hr
8 9
SA-4396-7
FIGURE 13.6 DESORPTION OF METHYL PARATHION FROM COYOTE CREEK SEDIMENT.
Loss of MP due to hydrolysis is included in the calculations.
266
-------
obtained. With viable and heat-killed cells incubated at 25°C, the sorption
coefficients were comparable, indicating that sorption by these organisms was
not enzymatically mediated.
TABLE 13.7. SORPTION PARTITION COEFFICIENTS3 OF
METHYL PARATHION ON A MIXED BATERIAL POPULATION
Initial MP
in mixture
(ng ml'1)
100
300 :-
Viable
cells
at 25°C
502 + 30
468 ± 20
Heat-killed
cells
at 25°C
359 ± 90
384 ± 47
Viable
cells
at 2°C
200 ± 86
242 ± 95
aThe values of Kp are the averages plus or minus standard deviation
for three experiments and are based on the dry weight of cells-
which were 916, 1037, and 1027 pg ml"1 in the three experiments.
These sorption coefficients are considerably higher than those obtained
in the sediment studies. This suggests that, after the algae "bloom" of eutro-
phic lakes, the sediments may sorb more MP because of the increased microbial
counts and detritus in the sediment. Both the MP levels and the bacterial-cell
concentrations used in these studies were much higher than would be expected
in the natural state, and the levels of MP in microbial cells would be much
lower.
13.5 CHEMICAL TRANSFORMATION
Methyl parathion was transformed by both hydrolysis and photolysis to
give the same two products: £-nitrophenol (£NP) and 0-methyl-0-p_-nitrophenol-
thiophosphoric acid (MNTP). Although not identified in our HPLC analytical
procedure, dimethylthiophosphoric acid and methanol are expected to be the
complementary products to £NP and MNTP, respectively. In the hydrolysis
studies, MNTP was the major product from reactions at less than pH 8, with
increased yields of £NP at higher pHs. The products from photolysis MP were
the same as those from hydrolysis in the same waters, but the yields of £NP
were higher. Details of product yields and the chemistry are presented in
Section 13.5.4.
13.5.1 Photolysis Rate
Data for the photolysis of MP in buffered and natural waters containing
1% acetonitrile are given in Table 13.8. Although the data are insufficient to
make a firm conclusion, it appears that the photolysis rate is slightly accel-
erated by the presence of phosphate and borate salts in moderate concentrations,
267
-------
especially at pH 8. Similar buffer catalysis effects were also found in the
hydrolysis studies (Section 13.5.3). The apparent acceleration of photolysis
rate in buffer solutions is probably not important for an environmental assess-
ment, however, since ionic strengths in most natural freshwater systems are
quite low. The data in Table 13.8 indicate that 1.7 x 10~A is a best value
for the quantum yield for photolysis of MP in aqueous solution.
The photolysis rate c ustants of MP in natural wat.ers are only slightly
smaller than those in the 1ightly buffered, distilled waters, the maximum
decrease being 40% for the Aucilla River water. As ciscussed in Part I, Sec-
tion 6 2 of this report, the presence of natural mat^r-als in waters may retard
photulvsis rai.es by Tight screening or by quenching processes The natural
materials may also promote phototransformations by photosensii:ization or photo-
initiated free radical processes.
On the basis of the limited number of experiments and the data in Table
13.8, it is impossible to give definitive reasons for the rate constant
decreases found. Thus, the rate constants for. Searsville Pond and Lake Tahoe
TABLE 13.8. RATE CONSTANTS FOR PHOTOLYSIS OF 26 yg ml"1 METHYL PARATHION
Irradiation
source
313 nm
313 nm
313 nm
313 nm
313 nm
313 nm
313 nm
313 nm
313 nm
366 nm
366 nm
Sunlight,
early January
Solution13
(buffer or source)
pH 8.0, 0.01 M borate
pH 5.0, 0.01 M acetate
pH 8-7, 0.001 M phosphates6
pH 5.0-5.4, 0.001 M phosphates6
pH 5.25, 9.5 yg ml~l humic acid
in pure water
pH 7.6, Lake Tahoe water
pH 7.8, Coyote Creek water
pH 7.5, Searsville Pond water
pH 6.07 Aucilla River water
pH 8.0, 0.067 M phosphates
pH 5.0, 0.067 M phosphates
pH 5.0, 0.067 M phosphates
Extent of
reaction (%)
82
64
37
20
59
67
69
57
60
91
54
57
kp x 107
73.8 ±
32.3 ±
46.7 ±
45.1 ±
36.2 ±
33.3 ±
46.8 ±
32.1 ±
28.0 ±
128.0 +
50.2 ±
7.89 ±
sec 1
3.1C
2.0d
2 9^
s'.Qf .
1.1
1.4
1.7
1.0
0.7
1.6
4.78
0.54h
1.0 pg ml"1 methyl parathion in water = 3.80 x 10 6 M.
Solutions contain added 1% acetonitrile; composition of buffer solutions are described in
Appendix B.
Standard deviation.
Quantum yield for MP disappearance at 313 nm is 1.2 x 10~*.
pH shifted due to insufficient buffer concentration.
Quantum yield for MP disappearance at 313 nm is 1.7 x lO"1*.
Quantum yield for MP disappearance at 366 nm is 2.5 x 10"*.
Calculated assuming 8 hours of sunlight per day in early January; to obtain average rate constant
for full calendar day (24 hours), divide rate constant by three.' Rate constant corresponds to half-
life of 30 calendar days.
268
-------
waters (absorbances at 313 are 0.07 and 0, respectively) are about 75% of those
calculated after taking into account a light screening effect. On the other
hand, the photolysis rate constants for the distilled water-humic acid, Coyote
Creek and Aucilla River experiments (absorbances of 0.56, 0.06, and 0.65 at
313 nm, respectively) are 45%, 10%, and 18% greater than those rate constants
obtained after correcting for light screening.
While more experiments are required to confirm these results, these
data along with the variations in product yields found in the experiments
(see Section 13.5.4) indicate that the chemistry occurring may be quite com-
plex. Further investigations of the photochemistry of these systems should
be revealing and valuable for a better understanding of the effects of natural
materials in aquatic photochemistry.
The half-life for the direct photolysis of MP as a function of the
time of year was calculated by the procedure of Zepp and Cline (1977), using
the quantum yield of 1.7 x 10"* and the measured UV spectrum of MP (Section
13.4.2). Allowing for a slight buffer effect on the rate constant (and
therefore the quantum yield) at 366 nm in Table 13.8, the quantum yield is
assumed to be independent of wavelength. The variation in half-life over the
year, summarized in Figure 13.7, shows that the half-life will vary from 8
days in summer to about 38 days in winter. The latter is in excellent agree-
ment with a 30-day half-life measured in an experiment conducted in early
January sunlight.
13.5.2 Oxidation Rate
The susceptibility of 26.0 ]_ig ml"1 (1.0 x 10" 4 M) methyl parathion
to free radical oxidation was examined using the AA-initiated oxidation
reaction (see Part I, Section 6.3 and Appendix B for discussion and procedure),
The data for the experiments carried out for 100 hours at 50.0°C in 1% ace-
tonitrile in water at pH 5 and pH 8 are given in Table 13.9.
TABLE 13.9. FREE RADICAL OXIDATION OF METHYL PARATHION AT 50°C
Solution
Buffered pH 5
Buffered pH 8
Initial cone.
(x 105 M)
10.0
10.0
MP consumed
with added AA
(x 105 M)
4.9
5.8
MP consumed
without AA
(x 10s M)
4.7
6.0
326.0 yg ml"1 MP = 1.00 x lO"4 M.
b
1.00 x 10~A M 4,4'-azobis(4-cyanovaleric acid) in 1% acetonitrile in
buffered water (0.01 M acetate or borate). Reaction time was 100 hours.
269
-------
40
32
T>
I
UJ
u. 24
_i
ul
16
JAN
MAR
MAY JUL
MONTH OF YEAR
SEP
NOV
SA-4396-12
FIGURE 13.7 ANNUAL VARIATION OF PHOTOLYSIS HALF-LIFE
OF METHYL PARATHION
These experiments demonstrated that the free radical oxidation of MP
is not competitive with hydrolysis at 50°C. The loss of MP shown in Table
13.9 can be completely accounted for by hydrolysis processes (see Section
13.5.3). As in the photolysis experiments, no methyl paraoxon was detected.
If we assume that less than 3% of MP consumed was oxidized in 100 hours
(based on the precision of the measurements), then the second-order rate
constant kox is <3 M"1 sec"1 at 25°C (see Part I, Section 6.3).
The absence of oxidation products is fully consistent with the rela-
tive unreactivity of esters and pentavalent phosphorus toward oxyradicals.
Lack of formation of methyl paraoxon in this experiment suggests that some
other circumstances are necessary to convert a phosphorus-sulfur bond to a
phosphorus-oxygen bond and that reports of its formation need to be examined
270
-------
carefully to clarify the special oxidizing conditions apparently necessary for
this reaction. In this regard, it is interesting that oxidation of thiophos-
phoric acid esters to phosphoric acid esters by ozone has been reported
(Gunther et al., 1970). Such a reaction could account for the field obser-
vations of the formation of paraoxon, which may then leach to ground and sur-
face waters.
13.5.3 Hydrolysis Rate
Rate constants for the hydrolysis of MP at 40.0°C as a function of pH
are given in Table 13.10. These data show that the hydrolysis rate is inde-
pendent of pH to about pH 8 where the hydroxyl anion catalyzed reaction (also
referred to as specific base catalysis; Frost and Pearson, 1961) becomes
important. Buffer solutions at 0.067 M in total phosphate were used in
initial studies but were replaced when catalysis by hydrogen phosphate dianion
became evident. Acetate and borate buffers were then used successfully with
no catalysis problems (see Table 13.10 for data and Appendix B for buffer
compositions).
The rate constants for the hydrolysis of MP in several natural waters
are listed in Table 13.11. The data from experiments in distilled water
containing 9.5 yg ml"1 humic acid and the four natural waters are in good
agreement (in all of these exepriments the pH was less than 8). The data in
Tables 13.10 and 13.11 show no catalysis in these natural waters when com-
pared with experiments in buffered solutions. The satisfactory agreement
among the first-order rate constants at the different methyl parathion con-
centrations (Table 13.11) verified that the hydrolysis reaction is first order
in methyl parathion. The pH rate profile for methyl parathion at 40°C in the
buffered and natural waters is given in Figure 13.8.
The detailed kinetics of hydrolysis have been discussed in Part I,
Section 6.4, of this report. Figure 13.8 shows that MP hydrolysis is not an
acid-catalyzed process above pH2 and that the OH" catalyzed reaction is impor-
tant only above pH 8. The first-order rate constant k^ is then given by
k = k + kBHT] (13.2)
To determine a value of kjj at any pH and temperature, we determined the
temperature dependence of k^j and kg independently. Rates of hydrolysis
were measured at pH 5 and 10, where the respective neutral and base-promoted
processes are dominant (see Figure 13.8). These data are given in Table
13.12. The measured rate constant kg at pH 10 still required a slight
adjustment for a contribution from kjg before kg could be calculated. Table
13.12 also gives the Arrhenius expression (temperature dependence) for kjj and
kg calculated from this data.
271
-------
TABLE 13.10. FIRST-ORDER RATE CONSTANTS V
FOR HYDROLYSIS OF METHYL PARATHION3 H
AS A FUNCTION OF pH
PH
3
5
6
7
8
9
10
11
.00
.00
.00
.00
.00
.00
.00
.25
6
6
8
7
9
16
67
747
kHx 107
(sec"1)
.80
.45
.08
.44
.27
.37
.70
(±
(±
(±
(±
(±
(±
(±
(±
0.
0.
0.
0.
0.
1.
1.
26)
33)
52)
24)
17)
39)
1)
b
c
d
e
f
166)
"3
Experiments used 26 pg ml"1 (1.0 x 10"'' M) methyl parathion
at 40°C. Solvent was 1% acetonitrile in buffered distilled
water. See Appendix B for composition of buffer solutions;
pH buffer solutions designated (b) in footnotes c, e, and f
were used in experiments where 0.076 M total phosphate buf-
fers, designated (a), were found to show buffer catalysis.
Standard deviation.
Solution pH 5(b) used; rate constant in solution pH 5(a)
was 8.47 x 10"7 sec"1.
High value probably due to buffer catalysis (see footnotes
c, e, f).
Solution pH 7(b) used; rate constant in solution pH 7(a)
was 14.3 x 10~7 sec"1.
Solution pH 8(b) used; rate constant in solution pH 8(a)
was 16.6 x 10"~7 sec"1.
272
-------
The rate parameters for the neutral hydrolysis (kN) are Ea = 23.4 ± 0.9
kcal mole"1 logCA/sec"1) = 10.10 ± 0.64; for the specific base catalyzed pro-
cess (kB), Ea = 22.0 ± 0.2 kcal mole"1 with logCA/M"1 sec"1) = 14.13 ± 0.17.
From the Arrhenius expression for kN (Table 13.12), kN at 25°C is 9 x 10~8
sec"1, corresponding to a half-life of 89 days for MP in aqueous solutions
below pH 8. The calculated half-life at 15°C is about 1 year in the same pH
region. At pH 9, which is found in some lakes and ponds, half-lives of 45
and 180 days are calculated from the data in Table 13.12 at 25° and 15°C,
respectively.
13.5.4 Products From Chemical Transformation
Hydrolysis and photolysis of methyl parathion produce two sets of
common products, which are detected in our LC analysis scheme by the £-nitro-
phenyl moiety. These; products are 0-methyl-0'-p_-nitrophenylthiophosphoric
acid (MNTP) and £-nitrophenol (p_NP). The £NP product, which has been reported
in previous studies, was identified by its HPLC retention time compared with
an authentic sample. An early eluting product was found in experiments under
acid reaction conditions. To identify this product, we conducted a large-
TABLE 13.11. FIRST-ORDER RATE CONSTANTS kH FOR
HYDROLYSIS OF METHYL PARATHION IN NATURAL WATERS3
— — — — — -
Natural water source
Lake Tahoe
Lake Tahoe
Coyote Creek
Distilled water with 9.5
7
7
7
5
PH
.6
.6
.8
.25
k'H x 10 7
(sec"1)
12.
6.
7.
5.
41
31
54
99
(±
(±
(±
(±
1.
0.
0.
0.
44)b'°
36)
21)
53)d
ppm humic acid°
Aucilla River 6.07 6.10 (± 0.81)
Searsville Pond 7.5 6.83 (+ 0.08)
Except where noted, experiments used 26 ng ml
(1.0 x lO"*1 M) methyl parathion at 40°C. Solvent was 1%
acetonitrile in filtered natural water.
Standard deviation.
0.26 ug ml"1 methyl parathion was used.
2.60 ug ml"1 methyl parathion was used.
273
-------
scale hydrolysis at pH 5.5 and 50.0°C and analyzed the mixture at 5 half-lives
of methyl parathion. The reaction mixture contained 95% of the early eluting
product with very small amounts of £-nitrophenol and methyl parathion. The
reaction mixture was lyophilized to concentrate the reaction products. The
mass spectrum of the product corresponded to that of the monodemethylated
methyl parathion (parent m/e = 249). The NMR spectrum of this product in D20
confirmed the assigned structure as 0-methyl-0-p_-nitrophenylthiophosphoric
acid, 3.65 8 (d, 3 protons), 7.6 8 (q, 4 protons). Although they were not
identified in our HPLC analyses, dimethylthiophosphoric acid and methanol are
expected to be the complementary products to p_NP and MNTP, respectively.
-3
-5
o
Q
-7
19 *-
<
X
193
10
pH
12
TA-327522-28R1
FIGURE 13.8 pH PROFILE FOR METHYL PARATHION AT 40°C.
Points with bars: buffered distilled water (Table 13.11);
other points (Table 13.10), • distilled H2O with humic
acid, AAucilla River, D Lake Tahoe, 4 Searsville Pond,
• Coyote Creek.
274
-------
TABLE 13.12. TEMPERATURE DEPENDENCE OF RATE CONSTANTS FOR NEUTRAL
AND BASE CATALYZED (kB) HYDROLYSIS OF METHYL PARATHION
Temp. Rate constants
Reaction (°C) kH x 107 (sec"1) kN x 107 (sec~l)
Neutral (measured 40.0 6.45 (± 0.33) 6.45
at pH 5.00)
50.0 18.2 18.2
70.0 170.0 (± 2.8) 170.0
to
S kB x 103 (M-1 sec'1)
Base-catalysed
(measured at pH 10.00) 20.0 5.83 (± 0.41) 5.36a
40.0 67.7 (± 1.1) 61. 6a
51.0 220 (± 3.4) 198a
Arrhenius expression
log kN = 10.10 - 23400/4. 58T
log kB = 14.13 - 22000/4. 58T
Rate constant is corrected for the contribution of kjj at the given temperature; the value of kjj is calcula-
ted from the Arrhenius expression.
-------
TABLE 13.13. PRODUCT BALANCE FOR HYDROLYSIS AND PHOTOLYSIS OF METHYL PARATHION
to
Hydrolysis
Reaction
Buffer sol.
Buffer sol.
Buffer sol.
Buffer sol.
Buffer sol.
Buffer sol.
Buffer sol.
Buffer sol.
Buffer sol.
Lake Tahoe
solution
pH 3.00
pH 5.00
pH 6.00
pH 7.00
pH 7.00
pH 8.00
pH 9.00
pH 10.00
pH 11.25
Coyote Creek
Aucilla River
Searsville
Pond
%' Rxn
69
42
51
88
66
75
63
68
66
73
24
27
pNPb
0
0.01
0.05
0.10
0.02
0.36
0.33
0.50
0.31
0.05
0.06
0.05
0.03
MNTPC
1.51
0.40
0.56
0.69
0.69
0.61
0.41
0.04
-
0.62
0.66
0.58
0.35
(2
(0
(0
(1
(1
(0
(0
(0
(
(0
(0
(0
(0
yield3
,d
.26)
.60)
.84)
.04)
.04)
.92)
.62)
.06)
~ )
.93)
.99
.87)
.53)
Totald
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
51
41
61
79
71
97
74
54
31
67
72
63
38
(2.
(0.
(0.
(1.
(1.
(1.
(0.
(0.
(0.
(0.
(1.
(0.
(0.
26)
61)
89)
14)
06)
29)
95)
56)
31)
98)
05)
92)
56)
Photolysis yielda (313 nm)
% Rxn pNPb MNTPC.d Totald
46 0.15 0.07 (0.10) 0.22 (0.25)
64 0.83 0.19 (0.29) 1.02 (1.12)
38 0.71 0.17 (0.26) 0.88 (0.97)
58 0.30 0.63 (0.95) 0.96 (1.25)
58 0.11 0.44 (0.66) 0.55 (0.77)
54 0.41 0.19 (0.29) 0.66 (0.70)
Note: Product yield data are from HPLC analysis (UV detector) optimized for separation of methyl parathion during kinetic
experiments. The £NP and MNTP elute earlier in the chromatogram and in some cases the baselines for £NP and MNTP are not
well resolved; the MNTP data may also'include other material that elutes at the same time. The MNTP and £NP may then be in
error by ± 50% of the values listed.
Based on methyl parathion consumed =1.
£NP • £-nitrophenol.
c
MNTP = 0-methyl-0-£-nitrophenylthiophosphoric acid.
Since a pure reference sample of MNTP was not available, the yield of MNTP was calculated by assuming that the HPLC response
factor (Rj) was the same as that of methyl parathion; the values in ( ) assume that the Rf for MNTP is 1.5 times that of methyl
parathion.
-------
. + CHgOH
OCR 3
pNP
Although no specific studies were conducted to determine the conditions
favoring the selectivejformation of either set of products, some information
on this matter is available from analysis of the data obtained during kinetic
measurements. Yields d>f jgNP and MNTP calculated from these data are given in
Table 13.13. These data are considered only semiquantitative inasmuch as the
analytical separation was optimized for MP. While the precision of the 2.NP
yield data is a few percent, the precision of the MNTP yield data may be ± 20%
in some instances due to poor baseline resolution. However, MNTP is definitely
a primary product because the MNTP peak is observed to increase as the reac-
tions proceed. In the reaction in which MNTP was isolated, it was found to be
the major product (>95%) in agreement with the low yields of £NP actually
measured in the hydrolysis kinetic experiments below pH 8 (Table 13.13).
While pNP has been reported as a product from hydrolysis of MP, MNTP
has not been reported. Recently the monoethyl ester analog to MNTP was
reported to result from the hydrolysis and biodegradation of parathion (Weber,
1976). The data in Table 13.13 indicate that the MNTP is produced mainly by
neutral hydrolysis and £NP is produced by base-catalyzed hydrolysis (> pH 8).
These results are consistent with results for other phosphate esters in which
hydroxide effects displacement on phosphorus with loss of the most acidic
alcohol, whereas water displaces on carbon with loss of the phosphate group
(HLlgetag and Teichman, 1965; Hudson. 1965; Vernon, 1957; Kirby and Warren,
1965).
From the data in Table 13.12 it is evident that photolysis of methyl
parathion produces greater yields of p_NP than hydrolysis in water from the
same source. As in the hydrolysis experiments, a higher pH seems to favor
formation of £NP. The yield data for MNTP and £NP in the natural water
photolysis experiments also suggest that more subtle effects on the selec-
tivity of product formation may be operative. From the data available it
appears that the photolytic process involves a light-catalysed hydrolysis in
which both water and hydroxide preferentially displace a p_-nitrophenoxy group
through an excited state in which displacement at phosphorus or at the Ar-0
bond is possible. If MNTP is shown to be a significant environmental hazard,
some studies directed toward the selective product formation may be warranted.
Since paraoxon is known to be more toxic than parathion, considerable
effort was directed toward determining whether methyl paraoxon was a product
from the chemical transformation of MP. The HPLC analytical procedure used
277
-------
would have detected methyl paraoxon formed in greater than 4% yield based on
original MP present. No methyl paraoxon was found in the experimental work
undertaken. Since paraoxon is more stable to hydrolysis than parathion at
pHs less than 7, it is unlikely that this result was due to instability of
paraoxon under the reaction conditions (Comma and Faust, 1972). Although
the data of Grunwell and Erickson (1973) are suspect for a number of reasons,
their reported yield of paraoxon from photolysis of parathion is 3%, which is
below the detectable limits of the experiments reported here. Thus, while
methyl paraoxon may be a product, it must be formed in low yields (<4%).
13.6 BIODEGRADATION
13.6.1 Development of Biodegrading Cultures
Water samples used included all of those listed in Part I, Section
7.2. With all samplings of eutrophic water bodies and sewage effluents,
enrichment systems that could rapidly decompose MP were developed in the
9-liter bottle fermentors. These were generally subtransferred several times
to basal salts/MP/glucose/yeast extract, and they were finally able to grow
on media in which the only added organic carbon was MP. Up to 300 yg ml"1 MP
could be completely metabolized per day, as determined by the disappearance
of the UV absorbance of the ]D-nitrophenol moiety.
The development of enrichment systems generally followed the pattern
of no change in MP concentrations for 3 to 10 days and then, within 24 to 36
hours, there was total disappearance of MP. In subtransfers, there were
transient periods when the fermentation liquor would become yellow for 1 to
3 hours. (The yellow color was probably due to p_-nitrophenol, which was iden-
tified as an intermediate product of MP, as described below.) In contrast to
other water samples, the initial fermentation mixture of sewage effluent from
the Monsanto Plant in Anniston, Alabama, with 100 yg ml"1 MP became yellow
within 20 to 30 seconds after addition of MP.
In three instances, when the first 9-liter bottle fermentations were
set up with 5-liter volumes of water sample, buffer, and MP, 50-ml aliquots
of these mixtures were removed and incubated at 25°C in 250-ml shaker flasks.
Invariably, the initial breakdown of MP was markedly delayed in the shaker
flasks and, in one instance with a Coyote Creek sample, no MP degradation was
observed after three weeks of shaker flask incubation. Aeration was not a
limiting problem because dissolved oxygen concentrations were essentially all
at saturation levels in all fermentations. In an attempt to obtain a better
understanding of this phenomenon, we diluted the contents in a 9-liter bottle
fermentor 100-fold with centrifuged Coyote Creek water (to reduce bacterial
counts) plus phosphates and (NIU^SO/, to represent the same total number of
organisms that would be present in 50-ml aliquots for the 250-ml Erlenmeyer
shaker flasks. In this instance, the diluted mixture, which contained
10 yg ml"1 MP, developed a biodegradative system before the undiluted fermen-
tations containing 1 or 10 yg ml"1 MP. Considering that the 50 ml of mixture
in each Erlenmeyer flask contained 40 ml of water sample, these observations
suggest caution in classifying a chemical as nonbiodegradable when small scale
studies are involved with perhaps 0.5% water or small soil additions as
sources of natural microbial flora.
278
-------
The 24-hour growth of the Coyote Creek enrichment cultures on basal
salts/100 pg ml"1 MP medium generally resulted in complete loss of MP (deter-
mined by the decrease of the UV absorption of MP) and an optical density (OD)
of approximately 0.02. This OD is much lower than would be obtained with the
same level of a good carbon nutrient. The culture mixture had a tendency to
aggregate and viable counts were difficult to obtain, but were on the order
of 107 to 108 cells ml"1. Also, these cells were very small. The ability of
these cultures to degrade MP was affected by transfers to fresh nutrient
broth containing MP. In one case, a transfer in 0.1 diluted nutrient broth/
50 pg ml"1 MP resulted in a cell population that, when used as a 1% inoculum,
produced a more rapid degradation of 100 pg ml"1 MP in the inorganic salts
medium. However, after three transfers in 0.1 nutrient broth/50 pg ml"1 MP,
the resultant culture complex did not degrade MP in 24 hours in the 100
pg ml"1 MP basal salts medium.
With Lake Tahofe water (containing some sediment), no MP degradation
was observed in a fermentor containing MP at 10 pg ml"1 and incubated at
15°C for 6 weeks. '
13.6.2 Biodegradation Kinetics
Biodegrading culture systems developed with Coyote Creek water were
used in four types of kinetic studies: batch fermentations with low-count
inocula of rested cells, continuous-feed fermentations in chemostats, cascade
fermentations with freshly developed biodegrading systems sequentially trans-
ferred into buffered freshwater samples, and batch fermentations with high
microbial populations.
The enrichment "culture" used in the kinetic experiments was streaked
out on agar plates containing only basal salts plus MP. Many types of cul-
tures were observed, and some were selected and streaked radially, from the
outer edge of 50 pg ml"1 MP/salts agar plates to the center of the plate. On
incubation some colonies grew well, the entire length of the streak, some
grew only close to the center, and some did not grow. These types of results
would be expected if mixtures of biodegrading organisms contained cultures
that could not metabolize MP directly but could grow on metabolic products
released in total or in part by organisms capable of digesting MP or products
closer to MP in the biodegradative pathways. It is also possible that the
growth of some of the secondary organisms may be due to their ability to grow
on products excreted by other organisms and is not directly related to the MP
biodegradative pathway.
Another indication of the complexity of .these mixed culture systems
was that generally, after three washings with 0.05% phosphate pH 7 buffer to
prepare resting cells, viable counts were stable over 18 to 27 hours at room
temperature, but when an aliquot of one resting cell lot was gently shaken
at 4°C during the 20- to 27-hour period, the viable count increased from
3 x 109 to 1.4 x 1011 cells ml"1. This nearly fiftyfold increase could be
attributed to the autolysis or release of some metabolites that readily
served as substrates for psychrophilic organisms in the culture mixture.
Between 27 and 43 hours, the number of viable cells in the resting cell mix-
ture maintained at room temperature also increased from 2.4 x 10 to 2.5x10
279
11
-------
cells ml 1. It is possible that the psychrophilic or other organisms even-
tually increased in numbers at room temperature, utilizing excreted or
autolysis nutrients from other organisms, but this interpretation has no
experimental support at this time.
Batch Fermentations with Low-Count Inocula of Rested Cells—Experi-
mental results obtained with diluted inocula of rested cells in batch fer-
mentations are presented in Tables 13.14 and 13.15 and in Figure 13.9.
During the first 16 to 20 hours, there was no utilization of MP, but as is
indicated in Figure 13.9, the increases in viable cell counts were essentially
the same for flasks containing 0, 1, and 30 yg ml"1 MP. The data for flasks
containing 3, 10, and 20 yg ml"1 MP were similar. These increases in bac-
terial counts are attributed to microbial interactions in mixed culture
systems. It is also possible that carbon nutrients were introduced from the
atmosphere or from the distilled water (Kayser et al., 1975). After this
first phase of growth, MP degradation was initiated and there was a second
growth phase that was most evident for flasks containing initial concentra-
tions of 10, 20, and 30 yg ml"1 MP.
TABLE 13.14. CELL COUNTS IN BATCH FERMENTATIONS WITH LOW COUNT
INOCULA OF RESTED CELLS
Time
(hr)
0
4
8
12
16
20
24
28
32
36
40
Cell
0
9.0
9.8
15
74
1300
—
2000
—
—
1500
counts x 10 3 at different MP
1
9.0
6.0
13
53
1700
2100
2000
2900
2900
1500
2400
3
9.0
6.3
14
73
1300
2000
2000
2600
2300
1900
2400
10
9.0
6.5
14
63
2000
2400
2900
3000
2900
3100
5700
levels (yg
20
9.0
6.4
12
63
1800
2600
2600
3300
3000
3500
8600
ml"1)
30
9.0
5.9
14
210
3800
3900
4300
6800
7400
8000
16000
280
-------
TABLE 13.15. MP UTILIZATION IN BATCH FERMENTAITONS WITH LOW COUNT
INOCULA OF RESTED CELLS
Time
(hr)
0
24
28
32
36
40
44
Methyl parathion
1 3
S ASa S AS S
1.0 0.0 3.0 0.0 10.
9.
9.
8.
0.96 0.04 2.8 0.2 6.
0.86 0.14 1.6 1.4 4.
0.43 0.57 0.3 2.7 0.
concentration (ye
10
AS
0
8
2
5
3
1
9
0
0
0
1
3
5
9
.0
.2
.8
.5
.7
.9
.1
ml"1)
20
S
20.0
19.0
17.9
17.6
14.8
11.7
7.5
AS
0
1
2
2
5
8
12
.0
.0
.1
.4
.2
.3
.5
30
S
30
29
27
24
22
19
7
.0
.0
.8
.3
.1
.0
.0
AS
0
1
2
5
7
11
23
.0
.0
.2
.7
.9
.0
.0
aAS = initial MP concentration (S0) minus the MP concentration at time t
(S).
From the MP biodegradations (AS) given in Table 13.15, In AS was
plotted versus time, and linear relations were obtained (Figure 13.10).
Table 13.16 presents values for MP utilizations (AS) and viable cell counts
(X) at 40 hours, as well as calculated values for the slopes at the curves
(y ) and kjj. The X and AS values at 40 hours were determined from plots of
the data in Tables 13.14 and 13.15. Data for flasks initiated with 1 and 3
yg ml"1 were not included because the ym and k^ values were of questionable
significance.
Although this procedure did not correct for the initial growth phase
without MP utilization, this was not considered in the average k^ value of
(1.7 ± 0.3) x 10~~7 yg MP cell"1 hr"1. If X were determined at the time of
initiation of MP utilization, AS/AX = Y would be lower and the kb values
would be somewhat higher. For example, in the 30 yg ml"1 studies, k^ would
be approximately 30% higher.
Figure 13.11 presents MP degradation versus time in these batch
shaker flask studies. The points at which ^S was equal to 5 yg ml"1 are
indicated, and the slopes of the curves at these points along with corre-
sponding S values were used to determine Ks by equation (7.18), Part I. For
degradations of 5 yg ml""1 of MP in flasks originally containing 30, 20, and
10 yg ml"1, the times were 32.4, 35.2 and 38.5 hours, respectively. The
rates of decompositions or slopes corresponded to -0.72, -0.70, and -0.62 hr 1!
281
-------
107
"L 10
I
O
o
LU
O
10s
10*
103
30 fjg ml"1 MP EXPERIMENT
CELL COUNTS FOR
30 fig ml"1
MP EXPERIMENT
CELL COUNTS
NO MP
EXPERIMENT
D
CELL COUNTS FOR
1 fig ml"1 MP
EXPERIMENT
1 fjg ml"1
MP EXPERIMENT
i
30
20
10
0.5
0.1
E
CD
I
O
I
I
8 12 16 20 24 28 32 36 40 44 48
TIME-hours
SA-4396-25
FIGURE 13.9 CELL COUNTS AND MP CONCENTRATIONS IN BATCH
FERMENTATIONS WITH LOW COUNT INOCULA OF
RESTED CELLS
282
-------
a.
I
a
01
a
a:
o
01
a
z
o
I
(-
cc
a.
I
01
30
20
10
1.0
0.5
0.2
0.1
0.05
0.02 -
INITIAL fig ml'1 MP
30
I
I
24 28 32 36 40 44
INCUBATION PERIOD - hours
SA 4396 26
FIGURE 13.10 METHYL PARATHION UTILIZATION IN BATCH FERMENTATIONS
WITH LOW COUNT INOCULA OF RESTED CELLS
283
-------
TABLE 13.16. SUMMARY OF DATA AND KINETIC CONSTANTS IN BATCH FERMENTATIONS
WITH LOW COUNT INOCULA OF RESTED CELLS
Zero time MP
concentration AS at 40 hra X at 40 hr ^m T»
(yg ml"1)
30
20
10
(yg ml 1)
12.5
8.2
5.8
(cells ml"1)
1.3 x 107
5.4 x 106
5.0 x 106
(hr"1)
0.144
0.133
0.139
(yg cell""1 hr r)
1.4 x 10"7
2.0 x 10"7
1.6 x 10~7
Average (1.7 ± 0.3) x 10"
aFrom plots of data in Table 13.15.
From plots of data in Table 13.14.
c AS
Calculated by the equation, k, = y —-.
b m X
respectively. The Ks values were 1.1, 1.1, and 1.0 (average 1.1) yg ml"1,
using the combined data from the 30 and 10, 30 and 20, and 20 and 10 yg ml"1
fermentations, respectively. This calculated value of Ks justifies the
assumption that Ks is significantly less than S for the above calculations of
kb in the 10, 20,_and 30 yg ml"1 degradations reported in Table 13.16. The
kb2 was 1.3 x 10~7 ml cell"1 hr"1.
Continuous-Feed Chemostat Fermentations—The data for MP biodegrada-
tion studies conducted in a chemostat under continuous fermentation conditions
are presented and analyzed in Table 13.17. The average cell yield was
(5.9 ±5.1) x 105 cells yg"1 MP. When S/y was plotted versus S in Figure
13.12, Ks and ym values were calculated as 2.6 yg ml"1 and 0.61 hr"1, respec-
tively (Lineweaver-Burk procedure described in Part I, Section 7.3). The
calculated value kb = ym/Y = (2.0 ± 1.7) x 10~"6 yg MP cell"1 hr"1. The
variations in cell yields are understandable because even more complexities
were involved than in the batch studies described above. The kj.,2 value was
7.7 x 10-7 ml cell-1 hr-1.
In these studies, chemostats were thoroughly shaken to dislodge
organisms attached to solid surfaces and the contents were transferred daily
to sterile chemostats. The problem of the effect of cells adhering to the
reaction vessel walls and to the associated hardware on the kinetics has
been well recognized in the past* and has led to the development or
Zobell (1943), Larsen and Dimmick (1964), Atkinson and Fowler (1974),
Pawlowsky and Howell (1973a,b), Pawlowsky et al. (1973), Gaudy et al. (1967),
and Williamson and McCarty (1976a,b).
284
-------
to
CD
Ul
16
20 24
TIME — hours
28
32
36
40 44
SA-4396-27
FIGURE 13.11 METHYL PARATHION DEGRADATION IN BATCH FERMENTATIONS WITH LOW
INOCULUM OF RESTED CELLS. Initial MP concentrations at 10 ( ®). 20 ( • )
and 20 ( D ) fjg ml'1
-------
TABLE 13.17. DATA FROM CONTINUOUS FEED CHEMOSTAT FERMENTATIONS
Specific
growth
rate (y)
(hr)
0.5
0.45
0.3
0.38
0.4
0.4
0.53
0.2
Outflow MP
concentration (S)
(yg ml"1)
12.7
5.97
2.96
4.58
4.84
6.63
13.2
0.98
Microbial
count
(cells ml"1)
2.7 x 107
4.7 x 106
1.45 x 107
6.6 x 106
5.6 x 107
y
(calculated)
25.4
13.3
9.9
12.05
12.1
12.05
24.0
4.9
Cell
(cells
7.3
1.3
4.3
2.5
yield (Y) kb
yg-1 MP) (yg MP cell"1 hr"1)
—
—
x 105 8.4 x 10~7
x 10= 4.7 x 10"6
—
x 105 1.4 x 10~6
x 10= 2.5 x 10'6
14 x 1QS 4.4 x 10~7
Average (2.0 ± 1.7) x 10"6
-------
30
S 20
f
10
-2
8
ml"1
10
12
14
SA-4396-28
FIGURE 13.12 LINEWEAVER-BURK PLOT OF SPECIFIC GROWTH RATES
AND SUBSTRATE CONCENTRATIONS FOR CONTINUOUS
FEED CHEMOSTAT FERMENTATIONS OF METHYL PARATHION
adaptation of a "wall growth factor" parameter in kinetic evaluations.*
This fermentation phenomena can play an important role in biodegradation.
The degree to which MP is completely converted to cellular material can be
affected by the feed rate. At higher feed rates some excreted metabolites
may be washed out that can be utilized for growth by other organisms.
Pawlowsky and Howell (1973a,b) in studies on mixed culture biooxidations of
phenol observed that different organisms were favored at different dilution
rates. Because the times involved in chemostat experiments were so much
greater than division times of the organisms, more opportunities were avail-
able for changes in the character of microbial populations. It was difficult
to replicate cell counts because of the continuous appearance of small colo-
nies that became more obvious with increased incubation of plate count agar
plates.
Pawlowsky and Howell (1973a,b), Howell et al.
(1971), and Chi et al. (1974).
(1972), Topiwala and Hamer
287
-------
Cascade Batch Fermentations with Fresh Coyote Creek Water Medium and
a Freshly Developed Biodegradation System — Figure 13.13 includes the data
for the initial adaptation of an MP biodegradation system -and the subsequent
sequential cascade transfer of this system into flasks containing fresh
Coyote Creek water media. In this experiment the 1% (vol) inoculum was
adequate for complete degradation of MP in 24 hours in the first two series
of flasks, but it was inadequate in the third and fourth series. This could
be attributed to interactions within the crude MP biodegrading system and to
interactions among the bacteria present in the fresh Coyote Creek water media.
These fresh waters would also contain protozoa that could prey on bacteria
effective in MP degradation. The data in Figure 13.13 are typical. In other
studies where larger inocula were used, no decrease in MP biodegradation at
24 hours was observed in the fourth series of flasks.
The rate constant kfe was determined to be (0.83 ± 0.49) x 10~7 yg MP
cell"1 hr"1, using the procedure used for calculating the rate constant on
batch fermentations with low count inocula of rested cells. The counts used
in these calculations were undoubtedly high because the biodegrading system
had not been subtransferred on media containing MP as the only carbon source
and contained organisms not involved in MP degradation. For this reason,
the k^ value is undoubtedly low.
It must be recognized that in addition to the predator-prey relation-
ships, the water from this eutrophic stream contains other compounds and
diauxic growth may have occurred.
Batch Fermentation with Large Microbial Populations — The data obtained
in this batch fermentation, starting with 1.8 x 108 cells ml"1 (Xo) and 20
yg ml"1 MP (S) are presented in Figure 13.14. Since the plot of In S versus t
is a straight line function with a decreasing slope, the decrease in MP
concentration is first order in MP. In one hour, 97.5% of the MP was
degraded.
The pseudo-first-order biodegradation rate constant calculated from
the plot of In S versus t was 3.69 hr"1 (see Part I, Section 7.3). When this
rate constant is divided by the cell numbers (X0 = 1.8 x 108 cells ml"1), a
second-order biodegradation rate constant, kjj2> dependent on both S and X
can be calculated. Under these conditions
' - 2 X 10~8 m -
1.8x
In this experiment, it was assumed that Xo » YS0. That these con-
ditions have been met in this experiment is substantiated by calculating
YS0 = 1.2 x 107 cells ml"1, using the yield factor 5.9 x 10s cells yg MP"1
derived in continuous fermentations in a chemostat. This yield of cells is
more than one order lower than the concentration of cells at the initiation
of the experiment, and with conventional techniques it would not be possible
to detect increased cell growth.
288
-------
to
CO
10r«
§ 6
I 2
ui
9-LITER BOTTLE
B C
FLASK I FLASK 2 FLASK 3 FLASK 4
TIME - days
Viable Bacterial Counts: A, 5.0x103, B, 4.6x105, C, 8.3x105, D, 1.0x106, E, 4.4x106, F, 5.0x104, G, 1.2x107,
H, 1.3x105, I, 2.3x107, J, 2.6x105, K, 1.2x107, L, 1.3x105, M, 1.7x107
SA-4396-29
FIGURE 13.13 CASCADE BATCH FERMENTATION OF METHYL PARATHION
-------
100
80
60
40
I
H
O
til
O
2
O
O
O
I
oc
<
Q.
I
01
20
10
1
0.8
0.6
0.4
ANALYSIS BY GCA; UV o
I
10
20
30 40
TIME — minutes
50
60
SA-4396 30
FIGURE 13.14 BATCH FERMENTATION WITH A HIGH POPULATION OF
MP-DEGRADING ORGANISMS
290
-------
The kinetic data obtained for MP biodegradation are summarized in
Table 13.18. Considering the diversity of techniques and the fact that these
were heterogenous mixtures of organisms, the agreement is good.
13.6.3 Identification of p-Nitrophenol as a Metabolite
During the growth of enrichment culture systems capable of biodegrading
MP on basal salts/MP media, there was generally a variable period in which the
solutions became yellow and then became colorless. Also, as was indicated
previously, when MP was added to aeration effluent from the Monsanto Plant in
Anniston, Alabama, the suspension quickly became a clear yellow. These obser-
vations suggested that £-nitrophenol was an intermediate in the biodegradation
of MP as had been observed in the biodegradation of parathion (Munnecke and
Hsieh, 1976). j
The trimethylsilyl derivative of _p_-nitrophenol was isolated from an
ethyl acetate extract of a Coyote Creek enrichment culture and identified by
GC/MS (Figure 13.15) by comparison with the spectrum obtained from an authen-
tic sample of _p_-nitrophenol. The relatively low levels of MP in the fermen-
tation and the complexity of our microbial populations may account for our
not detecting other aromatic metabolites or the other primary hydrolysis
product, dimethylthiophosphoric acid.
Recently, Daughton and Hsieh (1976) reported the toxicity and resis-
tance to biodegradation of diethylthiophosphoric acid formed in the microbial
degradation of parathion; with the organisms used in these studies, £-
nitrophenol was totally mineralized.
13.6.4 Discussion
The development of MP biodegrading cultures by enrichment procedures
utilizing eutrophic waters and aeration effluents from sewage or wastewater
treatment plants indicates that MP can be biodegraded by organisms present
in many areas. The failure to develop a biodegrading system with waters from
Lake Tahoe does not prove that the lake lacks any MP biodegrading organisms.
It may be related to the results obtained with small samplings of microflora.
It is also possible that our particular enrichment process did not favor
development of other types of organisms that could degrade MP.
The very rapid development of the yellow color attributed to £-
nitrophenol when MP was added to aeration effluent from the Monsanto Company's
parathion MP manufacturing facility and the development of the yellow color
when some cultures were subtransferred on MP/inorganic salts media suggest that
the esterase involved in the hydrolysis of the phosphate-0-p_-nitrophenol can
be an exoenzyme. This had been reported for parathion by Munnecke and Hsieh
(1976).
The high kinetic rate constants determined in these biodegradation
studies also support the concept that MP should not be considered a serious
291
-------
TABLE 13.18. SUMMARY OF KINETIC CONSTANTS FOR METHYL PARATHION BIODEGRADATION
kb Ks kb2
(yg MP cell"1 hr"1) (yg MP ml"1 = yg ml"1) (ml cell"1 hr"1)
Batch fermentation with low count
inocula of rested cells
Continuous feed chemostat fermentation
(1.7
(20
± 0.3) x 10"7
± 17) x 10~7
1.1
2.6
1.3
7.7
x 10~7
x 10" 7
Cascade batch fermentation (0.83 ± 0.49) x 10~7 1.8a 0.46 x 10~7
u
_D
Batch fermentation with a high pbpula- — — 0.2 x 10"
tion of MP-degrading organisms
Average 2.4 x 10
~7
3.
Average value of Ks from first two experiments.
Calculated from a first-order rate constant measured at a microbial poplation of 1.8 x .10~8 cells ml"1.
-------
100-
90-
80-
70-
60-
50-
40-
30-
20-
0-
50
rt.nl
]. 100 1
j'
,
1 i i 'f"i""r •• |- ••••••( i'"i-i -i |"i c i i i i | '|
50 200 250 300
SA-4396-24
FIGURE 13.15 MASS SPECTRUM OF THE TRIMETHYLSILYL DERIVATIVE OF p-NITROPHENOL
environmental pollutant, provided the natural flora of a water body does not
sorb the pesticide to a significant degree for biomagnification up the food
chain. Munnecke (1976) recently reported that a crude cell extract from a
mixed bacterial culture grown on parathion hydrolyzed this pesticide 2450
times faster than 0.1N sodium hydroxide at 40°C and suggested that this
product could be used for disposal treatment of a number of organophosphate
insecticides.
The lack of observation of any aromatic degradation product other
than £-nitrophenol and the rapid degradation of this metabolite by aerobic
mixed culture systems in aqueous systems suggest minimal possibilities for
aromatic MP biodegradation products creating serious pollution problems.
293
-------
14. LABORATORY INVESTIGATION OF MIREX
14.1 SYNOPSIS
The results of our laboratory studies suggest that volatilization and
sorption are the dominant environmental pathways of mirex. Photolysis of mi-
rex occurs only very slowly; hydrolysis and oxidation were not observed. Bio-
degradation was not observed under either aerobic or anerobic conditions.
The nine-compartment environmental exposure model predicted the follow-
ing steady-state concentrations of mirex in solution, suspended solids, and
sediments near point sources in the presence of a continuous discharge of
1 x 10-5 pg ml-1 (10 ppt) mirex.
Half-life in the
aqueous phase*
(hours)
River 0.83
Pond 420
Lake 1480
Predicted steady-state concentrations
Suspended
Solution solids Sediments
(pg ml"1) (yg g"1)
0.29
0.29
0.05
8.6 x 10-2
7.6 x 1Q-"
1.4 x 10-2
7.7 x 10-2
1.6 x 10-"
1.7 x 10-3
14.2 BACKGROUND
Mirex is the generic name for 1,1A,2,2,3,3A,4,5,5,5A,5B,6-dodecachloro-
octahydro-l,3,4-methano-2H-cyclobuta[c,d]pentalene , which is shown in Table
14.1. The general physical properties are also given in Table 14.1.
Mirex is used as a pesticide to control the fire ant in the southeastern
United States and the big-headed ant in Hawaiian pineapple fields, and as a
fire retardant in polymers. The term mirex has been applied to both the pure
chemical as well as the fire ant bait. It has also been used as a trade name
by Allied Chemical Company and Hooker Chemical Corporation. Dechlorane is
the name used by Hooker Chemical Corporation for its fire retardant. All
references to mirex in this report will refer to the pure chemical, which is
assumed to have been leached from the bait into the aqueous environment.
Significant concentrations of mirex have been found in the sediment of
Lake Ontario as a result of previous manufacturing activities by Hooker
"^Predicted by one-compartment model.
294
-------
Chemical. It has been found in Spring Creek, Pennsylvania, as a result of man-
ufacturing by Nease Chemical Company. An excellent review of the chemistry and
toxicology of mirex was published by Alley (1972). More recent research has
shown that bioaccumulation in organisms can be significant (Gripe and Living-
stone, 1977; Tagatz et al., 1976).
Average mirex concentrations in fish from contaminated waters generally
appear to be less than 1 ug g-1 but may be as high as 4 yg g"1 (Suta, 1977).
While the significance of these concentrations to animals in the wild is un-
clear (National Academy of Sciences, 1972), there is reason to suspect that mi-
rex may be a source of Kepone, which is a reported carcinogen. The established
tolerances for mirex in foods are 0.1 yg g"1 in meats, eggs, and milk fat and
0.01 ug g"1 in all other agricultural commodities (Suta, 1977).
TABLE 14.1. PHYSICAL PROPERTIES OF MIREX
Structure
•ci,
Molecular weight
Melting point (°C)
Solubility in water at
22°C (pg ml-1)
546.0
485 (decomposition)
Vapor pressure at 50°C 6 x 10 6
(torr)
70 ± 10C
1.00 pg ml"1 (1 ppt) mirex in water = 1.83 x 10"1 M
Pleasured in this study, see Section 14.3.1.
Recently, agreements have been made to phase out aerial applications of
mirex bait (Chem. Eng. News, 1976; Holden, 1976), and the use of mirex may be
sharply reduced in the near future. However, its extremely long persistence in
the environment suggests that it may continue to be an environmental problem.
The low chemical reactivity of mirex is described in the 1972 review by
Alley. Several papers have been published describing the products from photol-
ysis of mirex, but it is not known how environmentally relevant these studies
were and no data were given to assess how fast photolysis occurs in the aqueous
environment (see Sections 14.5.1 and 14.5.4 for discussion). While mirex has
295
-------
been generally found to be resistant to biodegradation, it has been reported
to be biodegraded under anaerobic conditions in sewage sludge (Andrade and
Wheeler, 1974; Andrade et al., 1975). A mirex metabolite in monkeys has also
been identified (Stein et al., 1976). Recently products of mirex have been
identified in samples from field locations where large amounts of mirex were
deposited and left exposed in the environment for 12 and 5 years (Carlson et
al., 1976). Hollister et al. (1975) reported that several genera of algae
concentrated mirex from very dilute solutions. A Chlorococcum sp. concentrated
mirex 10,000-fold from a 10 ppt solution. Biomagnification into flatfish from
mirex-containing sediment was reported by Koblinski and Livingston (1975).
While the chemical and physical characteristics of mirex suggest that it
should be highly resistant to biological and chemical transformation in aque-
ous environments and be strongly sorbed on bacteria, algae, and sediments, a
careful study and integrated assessment of the environmental pathways of mirex
is required. Our study was designed to measure the physical properties of
mirex and to obtain limiting estimates of the rate constants for chemical and
biological transformation processes of mirex in solution in aquatic environ-
ments .
14.3 ENVIRONMENTAL ASSESSMENT
14.3.1 Summary of Laboratory Data
The rate constants obtained in the laboratory investigations of mirex
are summarized in Table 14.2. The sorption partition coefficient measured on
our test mixture of four bacteria species is (4.8 ± 0.5) x 10 .
TABLE 14.2. SUMMARY OF MIREX LABORATORY DATA
Process Sorption equilibrium Partition coefficient
Sorption S = K S K = 460,000 ± 110,OOQ2
r s p w P
Rate expression Rate Constant at 25°C
Volatilization13 k [mirex] k = (6.3 ± 1.8) x 10~2 k°
vl v v
Photolysis0 k [mirex] = $ (£Z..e,) [mirex] k = <5 x 10~ see"
p A A p
Oxidation k [R09-] [mirex] k <30M~1 sec"1
ox ^ ox
Hydrolysis k, [mirex] k, <1 x 10~ sec"1
„. , , . d , Umax , n
Biodegradation k, „ = • — k, „ = 0
*a
On Coyote Creek sediment.
See discussion in Part I, Section 5.3, and Appendix B.
First-order rate constant; see discussion in Section 14.5.1.
Biodgradation was not observed in the enrichment procedures.
296
-------
14.3.2 Environmental Assessment Using the One-Compartment Model
The half-lives of dissolved mirex calculated for individual transfor-
mation or removal processes following a spill are listed in Table 14.3. Vola-
tilization and sorption appear to be the only important processes. Photolysis
of mirex in solution was found to be much slower than sorption or volatiliza-
tion and will not affect the transport of mirex by other processes. No bio-
degradation was detected in experiments that lasted up to six months. While
this implies that acclimated cultures are improbable, species that can degrade
mirex may yet be found.
TABLE 14.3. TRANSFORMATION AND TRANSPORT OF MIREX
PREDICTED BY THE ONE-COMPARTMENT MODEL3
I Eutrophic Eutrophic Oligotrophic
Process 1 Stream pond lake lake
Photolysis, half-life (hr) > 8,000 > 8,000 > 8,000 > 8,000
Oxidation, half-life (hr) > 10,000 > 10,000 > 10,000 > 10,000
Volatilization, 500 700 1,980 1,980
half-life (hr)
Hydrolysis,
half-life (hr)
Biodegradation,
half-life (hr)
Half-life for all
processes, except 500 700 2,000 2,000
dilution (hr)
Half-life for all
processes, including
dilution (hr)
Amount mirex sorbed (mg m~3)
Percentage mirex sorbed
0.83
0.3
97%
420
0.9
99%
1,480
0.15
94%
1,480
0.15
94%
a
-1
The initial concentration of mirex was assumed to be 10 pg ml (10 ppt).
Sorption half-lives have not been measured, but the times needed to
reach equilibrium are generally less than half an hour, roughly two orders of
nagnitude faster than volatilization. Since the partition coefficient of mirex
Ls greater than 10^, the amount of mirex removed from the water by sorption is
expected to be high, with 93% to 98% of total mirex in the various systems
Located in the sediments even if one assumes that mirex enters only in the dis-
solved form. However, volatilization may cause significant losses of dissolved
nirex from laboratory containers during stirring or transfer.
297
-------
14.3.3 Persistence
Sorption is expected to significantly affect the persistence of con-
centrations of dissolved mirex after release to natural waters, rapidly bring-
ing the concentrations of dissolved mirex helow the limits of detection.
While desorption will maintain relatively low concentrations of dissolved
mirex in ensuing weeks, the concentrations should be well below 6 x 10~s pg g~~*.
the limit of solubility. Since these low concentrations will be present
until the contaminated solids are buried by uncontaminated sediments and may
later be uncovered by dredging or natural scouring, any statements regarding
the persistence of mirex should be made very carefully.
Although variations in the persistence of mirex in eutrophic and
oligotrophic waters should be scarcely measurable, concentrations will
initially decline more rapidly in eutrophic waters. However, since eutrophic
waters have higher concentrations of suspended solids than oligotrophic
waters, the amount of mirex subject to eventual desorption should be greater
in eutrophic waters.
14.3.4 Mass and Concentration Distributions Calculated Using Computer Models
Table 14.4 summarizes the distributions of mass and concentration of
mirex expected at steady state during chronic discharge to each of four types
of water bodies, as calculated with the aid of the computer model. The
pseudo-first-order rate constants used in these simulations are presented in
the appendix.
The steady-state concentrations of mirex in solution in the simulated
stream segments imply a removal from solution of approximately 10% per
kilometer. The reduction of concentrations of dissolved mirex in the stream
=5imulation is due primarily to sorption followed by sedimentation. Hence, even
in smaller and thus more slowly flowing and less turbulent rivers than we have
assumed in the simulation, very nearly the same amount of mirex will be removed
within the same river reach.
The steady-state concentrations of mirex in solution are roughly 10~7
to 10~9 ug ml"1 in pond and lake simulations. This is roughly 100 to 10,000
times less than the input concentrations (Table 14.4). In the pond simulation,
the steady-state concentration of mirex is estimated to be 0.01% of the concen-
tration in inflowing waters. Within the lake simulations, the concentrations
of mirex in the surface waters are about 100 times lower than the concentra-
tions in inflowing waters.
The model predicts that virtually all the mirex is in the bottom
sediments at steady state, a pattern in agreement with the available field data
(Suta, 1977). Since the model assumes that mirex does not degrade in the
sediments, the estimated values of mirex in sediment are probably upper limits
298
-------
TABLE 14.4. DISTRIBUTION OF MIREX IN VARIOUS AQUATIC SYSTEMS AT STEADY STATE
(input concentrations of 10 pg ml"1 mirex)
to
v£>
VO
Compartment 1
(surface water)
Solution
Suspended solids
Compartment 2
(surface water)
Solution
Suspended solids
Compartment 3
(surface water)
Solution
Suspended solids
Compartment 5
(bottom water)
Solution
Suspended solids
Compartments 7-9
(sediment)
Solution
Solids
Total mass
Pond
Mass Cone.
(kg) (yg ml-1)
5.7 x 10-8 2.9 x 10~9 8
4.5 x 10-6 7.6 x 10-" 2
— — — _ 7
2
7
2
— —
—
7.2 x 10-11 2.9 x 10-9 2
1.7 x 10-3 7.6 x 10-*1 1
1.7 x 10-3 1
River
Mass
(kg)
.7
.6
.9
.3
.2
.1
.0
.5
.5
x 10-5
x 10~3
x 10-5
x 10-3
x 10-5
x 10-3
__
—
x 10- 6
x 10- *
x 10- 1
Cone.
(Mg ml'1)
2
8
2
7
2
1
2
7
.9
.6
.6
.6
.4
.0
.6
.7
x 10-7 1
x 10-a___l
x 10-7 7
x 1C'2 1
x 10-7 4
x 10-2 6
7
1
x 10-7 8
x 10-2 6
6
Lake
Mass
(kg)
.2
..JL
.0
.0
.2
.3
.8
.1
.4
.5
.5
x 10-7
_x 10-*
x 10-5
x 10-3
x 10-6
x 10- 5
x 10-7
x 10-5
x 10-8
x 10-2
x 10-2
4
1
2
8
1
5
3
8
9
4
Cone.
(Vg ml-1)
.9
.4
.7
.0
.7
.0
.1
.8
.6
.7
x 10-8
x 10- 2
x 10-8
x 10-3
x 10- 8
x 10-3
x 10-10
x 10-5
x 10-9
x 10-3
The amounts given for solid and solution phases in the sediment compartments are estimated
from the sorption partition coefficient for suspended solids and may be overestimated because
it was assumed that biodegradation of sorbed material does not occur.
-------
for the concentrations that may be expected in natural environments. However,
this is unlikely to be a significant source of error and should be less than
the uncertainties introduced by variations in the properties of actual water
bodies.
The most significant patterns are seen in the pond and lake simulations
(Figures 14.1 and 14.2), where concentrations of mirex in the solution,
suspended solids, and sediment phases are predicted to be insensitive to
termination of discharge following a period of chronic pollution since most of
the mirex would be concentrated on the sediments. The exception is the concen-
tration of dissolved mirex in streams (Figure 14.3), which is essentially con-
trolled by dilution.
14.3.5 Discussion
Despite the uncertainties inherent in extrapolating laboratory data to
the field, it is evident that approximately 95% of mirex should be rapidly
trapped in suspended solids and sediments after discharges to the water are
started. The mirex concentrations in the water column are too low to be de-
tected in practice but are expected to be insensitive to the stoppage of dis-
charge. Therefore, field sampling and control efforts should be focused on the
.sediments where mirex is accumulated.
Although volatilization has been observed in the laboratory, it is be-
lieved that the rate of escape of mirex into the atmosphere from natural waters
as a result of volatilization will be very low because the amount of mirex
predicted to be in solution is so much lower than the amount that is sorbed.
The most significant uncertainties lie in the potential biological ef-
fects and in the estimation of the ultimate concentration of the mirex, since
this is determined largely by the properties of the sediments and dilution
rates. Since the partition coefficient of mirex is probably sensitive to the
organic content of the sediments, the ultimate concentration of mirex in water
bodies with a long history of pollution with mirex may vary by one or two
orders of magnitude among water bodies. However, the significance of the
variations cannot be appraised with the available toxicological data. We can
only say that the solution concentrations will be quite low shortly after dis-
charges stop.
14.4 PHYSICAL PROPERTIES
14.4.1 Solubility in Water
The solubility of mirex in pure water at room temperature (22 ± 2°C)
was measured using the procedure for low solubility substrates described in
Appendix B, Section B.I.I. Saturated solutions of mirex were prepared in a
5-gallon carboy by coating the walls with mirex, filling it with water, and
equilibrating for several days.
300
-------
UJ
o
H
io-3 ~
ID"4 t=7
10'B —
o>
o>
§
10-'
OC
UJ
O
8 10'7
X
Ul
cc
10'9
10
•10
L-l i I—I 1 1—I 1 — 1-—I
__L1_ *
w /
- I
SEDIMENTS
50
SOLUTION
II I I I
1 I I I T
SUSPENDED SOLIDS
DISCHARGE STOPPED
I I I I I I I I I I I
100 160
TIME - hours
200
250
FIGURE 14.1 PERSISTENCE OF Ml REX IN A TWO-COMPARTMENT POND SYSTEM
-------
u>
o
COMPARTMENT
j r
-1
T
Ol
a.
o
z
Ul
o
§
X
(11
tr
—- SEDIMENTS
SOLUTION
100
200
300 400
TIME - hours
&00
600
700 720
FIGURE 14.2 PERSISTENCE OF MIREX IN A PARTIALLY MIXED LAKE
-------
6 x10-
1CT
- I I
~ COMPARTMENT >
S's'
/ / /
/ //
4///
///
///
///
/// i
'/ l
'.
10
-7
10-
10'
,.10
SEDIMENTS
, SOLUTION
456
TIME - hours
10
FIGURE 14.3 PERSISTENCE OF MIREX IN A PARTIALLY MIXED RIVER SYSTEM
303
-------
Several experimental problems were encountered during the analysis of
these solutions as well as the solutions from the volatilization and sorption
studies. Solutions prepared in the carboy contained suspended solid mirex,
which could be removed by centrifugation at 7500 rpm for 60 minutes. We also
showed that a signifcant amount of the mirex will volatilize from the solution
into the available air space. This loss was minimized by keeping the carboys
full (i.e., replacing volume removed with water and allowing reequilibration).
All samples were extracted with 15% methylene chloride in hexane,
dried over Na2SO^, and concentrated by a factor of about 1000 on a steam
bath, resulting in removal of the methylene chloride as an azeotrope. The
remaining hexane solution was analyzed by electron capture GC. Considerable
interferences from impurities in both the solvent and the water made the
analysis difficult. No interfering peaks occurred at the elution time of
the mirex.
The samples for the solubility study were equilibrated for 1 week,
centrifuged at 7500 rpm for 60 minutes, extracted, and analyzed as described
above. The solubility of mirex in pure water is 70 ± 10 pg ml"1 (1.3 x
10"11 M), based on three determinations. In another experiment, a single
sample equilibrated for two weeks gave the same value (70 pg ml"1).
14.4.2 Absorption Spectrum
The solubility of mirex in water, 50% water/50% acetonitrile, 100%
acetonitrile, and 100% methanol was so low that no absorption could be de-
tected above 295 nm, even in 10-cm cells. However, since there were several
reports in the literature that mirex photolyzed under various conditions (see
Section 14.5.1), we decided to measure the uv absorption spectrum of mirex in
a solvent (hexane) where the solubility would be significantly higher.
At a concentration of 5 x 10~2 M, the absorption above 290 nm was
still below our detection limit of 0.005 absorbance unit. Therefore, the ab-
sorption coefficient is less than 0.1 M~1cm~1 in the wavelength region greater
than 290 nm. There is the possibility of a solvent shift between hexane and
water, but since mirex is nonpolar, this shift should be minimal.
14.4.3 Volatilization Rate
The rate of volatilization of mirex from pure water was measured using
the method of Hill et al. (1976), which is described in detail in Part I
We estimated a Henry's law constant Hc ^ 8 x 103 torr M"1 using an estimate of
the vapor pressure of mirex at 20°C of 1 x 10~6 torr, and the solubility of
mirex in water given in Table 14.1 (70 pg ml"1), following the method of
Mackay and Wolkoff (1973). Using this value, it can be shown that if a volume
of saturated mirex solution is equilibrated with an equal volume of mirex-free
air, approximately 40% of the mirex will volatilize into the air space.
304
-------
and Appendix B of this report. The solutions of mirex were prepared and
analyzed in pure water using the procedures described in Section 14.4.1. The
mirex solution was centrifuged before the experiment was initiated. The
initial mirex concentration was 86 ± 10 pg ml~l*.
Since there was a good possibility that the mirex would sorb onto the
membrane and probe of the oxygen analyzer, our basic volatilization procedure
was modified. The oxygen reaeration rate was measured in a separate beaker
that was prepared at the same time except that it contained no mirex. It was
stirred at the same rate as the mirex-containing beaker, as measured by the
depth of the vortices. Each time a sample was withdrawn for mirex analysis, an
equal volume was withdrawn from the second beaker. The oxygen-reaeration rate
was measured in the second. beaker in the usual way; the mirex-containing beaker
was not subjected to the nitrogen purge. At the end of the experiment the
beaker was emptied andj extracted; 11% of the amount of mirex originally present
was found sorbed to the beaker walls. This amount was within the experimental
error of the procedure;
The average oxygen reaeration rate, kv, over 2.5 days was 0.86 ± 0.13
hr"1. The rate constant for volatilization of mirex, k^, over the same time
period was (5.39 ± 0.65) x 10~2 hr"1. The ratio, kM/fcO, is 6.3 ± 1.8) x 10 2.
14.4.4 Sorption on Natural Sediment
The sorption partition coefficient of mirex was measured on Coyote
Creek sediment. All solutions containing mirex were centrifuged before the
isotherm measurements. The concentration of "mirex in the supernatants after
sorption and centrifugation was measured as described in Section 14.4.1. The
sediment was extracted with two 10-ml portions hexane, and the hexane was con-
centrated and analyzed as before. The initial concentration of mirex was al-
ways less than the saturation concentration of 70 pg ml"1. Duplicate samples
at two mirex concentrations and two sediment concentrations plus blanks were
prepared. The results are summarized in Table 14.5.
The wide range of recoveries is explained in part by the experimental
difficulties discussed in Section 14.4.1, interferences introduced with the
addition of sediment, and extra handling involved in the isotherm measurements.
The second point is very high, however, for reasons that are not apparent. In
the linear least squares (LLS) regression analysis, elimination of this point
reduces the 95% confidence interval by one-half, but the partition coefficient
is virtually unchanged. The nonlinear least squares (NILS) regression, which
compensates for flask effects (Appendix B, Section B.I. 4) allows use of all the
data and gives the best estimate of Kp with the narrowest confidence interval.
Even the lower bound of this value shows very strong sorption of mirex by
sediments.
This concentration is within the experimental error of the solubilility limit.
305
-------
TABLE 14.5. SORPTION OF MIREX ON COYOTE CREEK SEDIMENT
en
Mirex
Sorbent concentration
concentration in supernatant
(yg ml-1) (pg ml"1)
3.0
3.0
12
3.0
3.0
12
12
0
0
1
4
5
1
1
.33
.33
.0 ±
.5 ±
.2 ±
.0 ±
.0 ±
± 0.5
± 0.5
0.5
1.4
0.71
0.5
0.5
Mirex
concentration
on sediment
(yg g-1)
0.
2.
0.
i.
i.
0.
0.
91
3 ±
41
7 ±
6 ±
53
69
±0.09
0.43
± 0.07
0.53
0.37
± 0.05
± 0.05
Partition coefficients
(Kp x 10~ 5)
Recovery LSS^ NLLSa
(%) an = 0 a ?* 0 S and S
0 o w s
69 3.7 ± 3.6 1.1 ± 4.1 4.6 ± 1.0
180 (ao * °'98 ± ^^
130
84 3.6 ± 1.5b 2.4 ± 1.6b
QQ (a = 0.46 ± 0.47)b
Bo O
70
81
aLLS = linear least squares; NLLS = nonlinear least squares; see Appendix B, Section B.I.4 for ex-
planation of regression.
Recalculated excluding the data reported in the second line of this table.
-------
14.4.5 Biosorption
Because of the very low solubility of mirex in water, the concentra-
tion used in biosorptions was reduced to 50 pg ml"* and the microbial mixture
of four bacteria was used at an optical density of 0.01. This concentration
of bacteria corresponded to a dry weight of 6 pg ml-1. To facilitate analyses,
sorptions were conducted in 100-ml Corex bottles, and contents from two bot-
tles were combined for extractions. The correction factors applied for sorp-
tions of mirex and bacterial cells containing mirex onto the glassware were
18% and 14% for the 1-hour sorptions and 3-hour desorptions, respectively.
The results in Table 14.6 clearly indicate the affinity of bacterial cells for
mirex. The sorption coefficients are of the same order as those obtained with
benzo[a]pyrene,
TABLE 14.6. MIREX SORPTION AND DESORPTION ON A MIXED BACTERIAL POPULATION3
\.
Amount in Amount in Sorption
supernatant (%) cells (%) coefficient (x 10~5)
Sorption
Desorption
22 ± 1
23 ± 4
63 ± 7
45 ± 7
4.8 ± 0.5
3.3 ± 0.5
Dry weight of bacterial cells was equivalent to 6 yg ml"1; sorptions were ini-
tiated with 50 pg mirex ml"1.
14.5 CHEMICAL TRANSFORMATION
14.5.1 Photolysis Rate
While the UV absorption of mirex above 290 nm was found to be weak
(e < 0.1 M-1 cm"1), the reported photolyses of mirex (Alley et al., 1974a,b) and
the apparent slowness of other processes justified some photochemical studies.
Solutions of 33 ng ml'1 (6.0 x 10~8 M) mirex in 1% acetonitrile in pure water
were placed outdoors (Menlo Park, California) from November 1, 1976, to May 9,
1977. These solutions were analyzed for mirex at reactions times of 22, 63,
100, and 190 days by GC using an electrolytic conductivity detector operating
in the halogen mode.
A progressive loss of mirex was found in each analysis, with a first-
order rate constant of (3.7 ± 0.8) x 10"3 days-1 calculated from these data.
This rate constant corresponds to a half-life of about 1 year.
A solution of 33 ng ml"1 mirex in 1% acetonitrile in water from Coyote
Creek was also exposed to sunlight over the same six-month period, with analy-
ses for mirex carried out at 83 and 190 days. Including the initial concen-
tration data point, a first-order rate constant for photolysis of (4.2 ± 0.2)
x 10~J days"1 was calculated.
307
-------
Extraction of the glass in the head space and in contact with the
solution showed that no significant loss of mirex had occurred through
volatilization or adsorption in these experiments. Analyses of the mirex
solutions showed that four products were formed in the photolyses; the gc re-
tention times of two of these products correspond to those of Kepone and a
monohydromirex (see Section 14.5.4 for further discussion of products).
Measurement of a quantum yield for mirex photolysis above 290 nm was
not practical since the absorption coefficients in water (or other solvents)
could not be reliably measured. To determine whether the 1-year half-life
measured for mirex photolysis was plausible, we calculated the photolysis
half-life according to the procedure of Zepp and Cline (1977) , using hypothet-
ical values of the quantum yield and the absorption coefficients. As de-
scribed in Part I, Section 6.1, the photolysis rate constant in sunlight is a
function of the solar irradiance Z\ the average absorption coefficients of the
chemical EX> anc^ tne quantum yield, :
2 303
1 "^"—^
k =
p 6.02 x 10-
•^
Using the Z^ data for the summer season in the wavelength region 297.5 to
310 nm and assuming that the absorption coefficients over this region are
equal to a value of 0.01 and that the quantum yield is 1.0, we calculated a
minimum half-life of about 130 days for the direct photolysis of mirex. The
calculated minimum half-lives for the winter and spring season are 850 and
190 days, respectively.
Any other combination of quantum yield and absorption coefficient that
gives a product of 0.01 over the same wavelength region would give the same
calculated half-lives. While reliable data for both $ and e^ are still
needed, these calculations indicate that for a compound like mirex, which has
very weak absorption in the solar spectrum, photolysis may still occur with a
half-life of about a year if the quantum yield is about unity.
Although the above calculation indicates that the measured half-life
may be completely accounted for by the photolysis of mirex in solution, other
processes could also contribute to the transformation rate of mirex in our ex-
periments. The formation of products does show, however, that transforma-
tions are taking place. Since microbiological studies failed to develop any
biodegrading cultures (see Section 14.6), such processes are not likely to oc-
cur in our filter-sterilized solutions, even after a 6-month reaction time.
A photoprocess that could have occurred in our photolysis experiments
is photolysis of the mirex that was adsorbed on the glass of the reaction
tube. Although it is not known whether it would catalyze the photolysis over
that in solution, the sunlight photolysis of mirex on silica gel plates has
been reported (Ivie et al., 1974), and one of the products was reported to be
308
-------
a monohydromirex. On the basis of our experiments we cannot say whether
photolysis of mirex sorbed on glass is important. To rule out such an arti-
fact in the data applied to the environmental assessment, further research is
clearly required. Since the calculation performed according to the procedure
of Zepp and Cline (1977) indicates that the solution phase photolysis may en-
tirely account for the observed half-life, a priority in such additional re-
search should be to determine the quantum yield for direct photolysis.
Experiments near 280 nm (e for mirex is 0.15 at 280 nm) may be acceptable for
this purpose since the absorption band that has a emax at ^ 228 nm is the one
that tails into the solar spectrum region.
14.5.2 Oxidation Rate
Mirex was expected to show very low reactivity toward free radical
oxidation since an initial process of chlorine atom abstraction from mirex by
alkylperoxyl or alkyoxyr* radical is a highly endothermic reaction.* To
demonstrate this low reactivity and to ensure that no other radical reaction
pathways would lead to transformation of mirex, we conducted a free radical
oxidation experiment as described in Part I, Section 5.4.2, and Appendix B,
Section B.2.
A solution of 33 ng ml-1 (6.0 x 1Q-8M) mirex in 1% acetonitrile in
water containing 1.0 x 10~A M AA was heated at 52°C for 164 hours. At the end
of this time all the mirex was found to be unreacted within the ±5% experi-
mental error of the analysis, and no product peaks were found in the gc trace.
Under the conditions of our free radical oxidation experiment, the slowest re-
action that can be quantitatively evaluated has a rate constant of 30 M"1 sec"1
at 25°C. With the assumption that lR02' 1= 10~9 M in the environment, this rate
constant corresponds to a half-life of 0.7 year. The half-life for oxidation
of mirex must then be greater than 0.7 year, and in fact is probably several
orders of magnitude greater.
14.5.3 Hydrolysis Rate
Experiments to determine the susceptibility of mirex to chemical
hydrolysis were conducted in sealed tubes at 100°C using 33 ng ml"1 mirex in 1%
acetonitrile in water. In tubes where an air space was present above the re-
action solution, some volatilization of mirex was found with condensation and
sorption onto the glass in the head space. Without an air space in the sealed
tube, all the mirex was found to be present in solution, within the analytical
experimental error of ±5%, after one month at 100°C.
"The bond dissociation energy for a C-C1 bond is about 80 kcal mole"1 (Benson,
1968), whereas that for 0-C1 is about 48 kcal mole-1 (based on the 0-C1 bond
of t-butyl hypochlorite; Walling and Papaionnou, 1968), making the reaction at
least 30 kcal mole"1 endothermic.
309
-------
No products were found in the GC analyses of any hydrolysis experiment.
Kepone, which we considered to be the most likely hydrolysis product, was re-
peatedly sought in all analyses, but none was found. All reactions were
carried out at a pH of about 7. The literature indicates that hydrolysis of
halides by hydroxyl ion substitution would not be important below pH 10 and
that there is no significant acid-catalyzed reaction (Mabey and Mill, 1977).
Although no evidence for hydrolysis at 100°C was found after one month,
if we assume that the experimental error of ±5% is a loss due to hydrolysis of
mirex, we can calculate an upper limit for a first-order rate constant of
2 x 10~8 sec"1. Allowing for a factor of two in rate for each 10°C interval (see
Part I, Section 6.4 ) we estimate an upper limit for the hydrolysis rate
constant of 1 x 10~10 sec"1 at 25°C. This rate constant corresponds to a half-
life of over 250 years, confirming that hydrolysis of mirex will be a very slow
process in the aquatic environment.
14.5.4 Products from Chemical Transformation
No loss of mirex or appearance of new GC peaks was observed in the hy-
drolysis or oxidation studies. Analysis of the sunlight photolysis reactions
showed the loss of mirex and the appearance of four products. These products
appeared in the GC analysis as one major peak and three minor peaks. It was
calculated, however, that if these products were assumed to have GC response
factors similar to that of mirex, they represented less than 40% of the total
mirex consumed in the photolysis. One of the minor mirex photoproducts showed
the same retention time as that of an authentic sample of Kepone (I)•
By comparing the mass spectrum of the major product peak (^ 25% yield) with the
data reported by Carlson et al. (1976), we identified the product as a monohy-
dromirex. Insufficient material was available to identify any of the other mi-
rex products by GC-MS or to confirm the presence of Kepone by other than its GC
retention time.
Photolyses of 33 ng ml~l (6.0 x 10~8 M) mirex in solutions of 1%
acetonitrile in pure water were also carried out in sealed Pyrex tubes irradi-
ated with high intensity mercury lamps. Analysis of solutions irradiated for
24 and 48 hours using a 1000-watt lamp gave GC product peaks similar in reten-
tion time and in product proportions to those of the sunlight photolysis ex-
periments. Photolysis of the 33 ng ml"1 mirex solution with a 2500-watt lamp
for 100 hours led to complete loss of mirex but without detection of products.
310
-------
Although extensive product studies could not be pursued in this work,
some discussion of products is relevant for the environmental assessment.
Studies of Alley et al. (1973, 1974a) found the monohydromirex (II) to be a
major product of mirex photolysis in cyclohexane or isooctane.
II
This monohydro product was also the major product when mirex was photolyzed on
silica gel plates (Ivie et al., 1974). The monohydro product III is reported
to be the major product from photolysis of mirex in triethylamine solvent
(Alley et al., 1974b).
Ill
The change in products was attributed to photolysis of a charge transfer com-
plex formed between the amine and mirex. In both solution photolyses, forma-
tion of the monohydro products can be explained by initial photodissociation
of a C-C1 bond followed by hydrogen abstraction from solvent; it is difficult
to explain, however, the formation of II on silica gel, which has no obvious
hydrogen donor capability.
The formation of a monohydromirex product in our studies can be ex-
plained by photodissociation of C-C1 followed by hydrogen transfer from
acetonitrile, which was necessarily present in all reactions as a cosolvent.
Based on the work of Alley and coworkers, it is reasonable to assign II as our
monohydromirex product from solution photolysis. Kepone, which was tentative-
ly identified in this work, may be derived from mirex by the following photo-
oxidative process:
•f^-, ""K-,
-Cl
The failure to obtain a good product balance in our studies may be due to
other photooxidative reactions that compete successfully with hydrogen trans
fer and loss of chlorine radical.
311
-------
Cl
further
degradation
Cli
Oxidation products such as Kepone and other carbonyl-containing products may
be as photolabile as mirex itself and further degraded to acidic, water-
soluble compounds, which would not be extracted into hexane for subsequent
analysis.
Experiments to study and identify the potentially numerous and compli-
cated products in reaction mixtures from such photooxidations are beyond the
resources of this project, but evidence for such extensive degradation is pro-
vided in the experiment with the 2500-watt lamp where neither mirex nor
products were found after 100 hours of photolysis, but where some gas,
possibly C02, was generated. Kepone itself has been reported to be ex-
haustively photolyzed in aqueous solution to give carbon dioxide and HC1 as
products (Knoevenagel and Himmelreich, 1973). In an aqueous environment,
solution-phase photolysis of mirex could then lead to extensive photodegrada-
tion of the mirex structure. The monohydromirex or further reduced mirex
structures may also be formed if sufficient concentrations of hydrogen species
are available.
14.6 BIODEGRADATION
14.6.1 Development of Enrichment Cultures
In attempts to develop mirex-biodegrading systems, the following
studies were conducted:
• Aerobic incubations of mirex with aeration effluents from sewage
plants of the city of South San Francisco and the Shell Oil Company
Refinery (Martinez, California).
• Anaerobic aqueous incubations of subsurface soils exposed to mirex
in Florida and Mississippi with ll+C-mirex.
• Anaerobic digestion of mirex with an inoculum obtained from an an-
aerobic sewage digester in the South San Francisco Sewage Plant.
312
-------
Because mirex has such low solubility in water, 40 ug of mirex dis-
solved in methylene chloride was distributed on 1 g of 0.2-mm glass beads. The
solvent was evaporated, and 200 ml of 4:5 diluted aeration effluent or an-
aerobic digester fluid was added to 500- or 250-ml Erlenmeyer flasks, re-
spectively. It was anticipated that, with the greater relatively inert surface
of the glass beads, solution of mirex would take place more readily if decompo-
sition occurred. Appropriate additional sewage plant liquors (50 ml) were
added at three weeks to the flasks containing South San Francisco sewage
samples.
The results of the six-week aerobically incubated flasks are presented
in Table 14.7. No significant degradation of mirex (as detected by GC) was
observed in these flasks, which contained an equivalent of 200 ug mirex liter" .
TABLE 14.7. AEROBIC MIREX ENRICHMENT STUDIES WITH AERATION EFFLUENTS
% Recovery
Source and treatment of aeration effluent of mirexb
South San Francisco sewage plant 92 ± 2
South San Francisco sewage plant plus 50 ug ml~l 100 ± 3
c
Johnson's baby oil
South San Francisco sewage plant, autoclaved 93 ± 2
Shell Oil Refinery wastewater treatment plant 97 ± 17
Shell Oil Refinery wastewater treatment plant, 100 ± 1
autoclaved
SA11 effluents diluted with 0.25 vol of pH 7 phosphate-buffered (NHA)2SO<.
solution.
Results are an average of GC analyses of duplicate flasks incubated in a
rotary shaker at 25°C for 6 weeks.
°To provide n-paraffins, which are frequently good substrates and provide a
good supply of H donors for substitution reactions.
The soil samples that had previously been exposed to mirex were from an
experimental plot (#60202$ in Gulfport, Mississippi, and a sample (#602038) of
the muck from a pond at the site of an airplane crash near Sebring, Florida.
Mirex was introduced into these samples in 1962 and 1968, respectively, and
their mirex contents were reported as 0.096 and 2.5 ug g-1, respectively
(samples , analytical data, and analytical procedures were kindly provided by
313
-------
Dr. J. H. Ford, USDA, Gulfport, Mississippi). Because these samples also con-
tained Kepone and related products, ll*C-mirex (California Bionuclear Corpora-
tion) was used to determine any possible change of mirex in suspension of 20 g
soil, 30 ml H20, and 36 nCi (sp. act. 5.76 mCi/mmole) of 1IfC-mirex (added in
10 yl acetone). These suspensions were prepared in 50-ml (25 x 150 mm) Teflon-
lined screw-capped tubes. Tubes were thoroughly sparged with nitrogen before
they were sealed and they were shaken daily.
Extracts from individual tubes after thin layer chromatography (TLC)
on silica gel (Andrade and Wheeler, 1974) when examined by x-ray film radioau-
tography did not indicate any other radioactive component other than ll*C-mirex.
The x-ray radioautography would have detected less than 1% of conversion
product to a 10-carbon compound if it had been separated by the TLC system.
Band widths on the TLC plates corresponded to Rf values of 0.63 to 0.68, and
there was no discernible difference in intensity in the band of the radioauto-
graphs to suggest more than one component.
The anaerobic incubations with 200 ml of 80% anaerobic digester
suspension from anaerobic digesters at the South San Francisco sewage plant
were conducted at 37°C for up to 12 weeks, in standard tapered ground joint
250-ml Erlenmeyer flasks on which were 150 mm standard tapered extension tubes
with aluminum-foil-covered rubber stoppers at the distal ends. These rubber
stoppers were fitted with small-diameter Bunsen pressure relief valves. Half
of the flasks had the equivalent of 50 pg succinic acid ml"1 added weekly, and
in each case the contents were sparged with nitrogen gas. Control recoveries
obtained with the anaerobic digester sample added to the 11+C-mirex deposited on
glass beads were 97 ± 7% in four flasks. Recoveries of mirex after 12 weeks of
anaerobic digestion averaged 93% and 92% (GC analyses) in duplicate flasks with
and without added succinic acid, respectively.
14.6.2 Discussion
The above screening tests under aerobic conditions with aeration
effluents from two sewage plants, the anaerobic studies with subsurface soil
samples that were exposed to mirex for a long time, and the studies with an-
aerobic digester contents from a sewage plant clearly support the position that
mirex, when added to natural waters, could resist biodegradation for long
periods.
Our high recoveries of "mirex" may not in fact be absolutely indicative
of lack of biodegradation. As has been reported by others, some
undecachloromonohydromirex analogs may have similar GC retention times (Carlson
et al., 1976) and may have Rf values in TLC very close to those of mirex
(Andrade and Wheeler, 1974). In this latter reference, the ratio of
mirex:metabolite from anaerobic sludge digestion was 88:1 with Rf values of 0.74
and 0.65, respectively. No explanation was presented for the differences in
the mirex:metabolite ratios with the anaerobic sludge control and the mirex
stock solution. In a subsequent paper relating to this metabolite (Andrade et
al., 1975), the same Rf by TLC was observed with the acetone:n-hexane (1:9 v/v)
developing solvent they used previously, but with a double development with
n-heptane, as we have used, the Rf was 0.69. By comparison, our R,- for mirex
314
-------
with a double development with n-heptane on silica gel plates was 0.65 (band
width 0.63 to 0.68).
The difference in results obtained in the two laboratories may be at-
tributed to experimental conditions. It is also questionable that we would
have detected a second zone with an Rf difference of 0.09 if we had a
mirexrmetabolite ratio greater than 100.
In the report on the characterization of a mirex metabolite from rhesus
monkeys (Stein et al., 1976), a metabolite was observed at about two and five
weeks after one single oral dose of 1I+C-mirex. The production of this
metabolite after this prolonged period was attributed to bacterial action in
the lower gut or in the feces. It was believed to be a monohydromirex. Jones
and Hodges (1974) tested the activity of a variety of soil fungi, some
bacteria, nine aerobic soils, and four anaerobic lake sediments and did not
observe degradation of 11(C-mirex.
It is apparent that, if microbial degradation of mirex can take place
in a location, it will be very slow and it would form other polychlorinated
monohydro analogs that would be expected to be toxic and biomagnified.
315
-------
15. REFERENCES
Albersmeyer, W. 1958. Quantitative Determination of Aromatic Hydrocarbons in
Aqueous Solutions. Gas-U. Wasserfach 99:269.
Alley, E. G. 1972. The Use of Mirex in Control of the Imported Fire Ant. J.
Environ. Quality 2(1):52-61.
Alley, E.G., D. A. Dollar, B. R. Layton, and J. P. Minyard, Jr. 1973. Photo-
chemistry of Mirex. J. Agr. Food Chem. 21(1):138-139.
Alley, E.G., B. R. Layton, and J. P. Minyard, Jr. 1974a. Identification of the
Photoproducts of the Insecticides Mirex and Kepone. J. Agr. Food Chem.
22(3) -.442-445.
Alley, E.G., B. R. Layton, and J. P. Minyard, Jr. 1974b. Photoreduction of
Mirex in Aliphatic Amines. J. Agr. Food Chem. 22(4) :727-729.
Anashkina, N. P., S. M. Maksimov, and V. M. Dzyuba. 1972. Spectrophotometric
Study of the Oxidation of Benzothiophene and 3-Methylbenzothiophene. Khim.
Seraorg. Soedin., Soderzh. Neft. Nefteprod. 9:160-162; Chemical Abstracts
80:59814y (1974).
Andelman, J. B., and J. E. Snodgrass. 1974. Incidence and Significance of
Polynuclear Aromatic Hydrocarbons in the Water Environment. CRC Critical
Reviews in Environmental Control 4 :69-83.
Andelman, J. B., and M. J. Suess. 1970. Polynuclear Aromatic Hydrocarbons in
the Water Environment. Bull. World Health 43:479-508.
Andelman, J. B., and M. J. Suess. 1971. Photodecomposition of 3,4-Benzpyrene
Sorbed on Calcium Carbonate. In: Organic Compounds in Aquatic Environments,
Rudolfs Research Conference, Fifth 1969. pp. 439-468.
Andrade, Jr., P.S.L., and W. B. Wheeler. 1974. Biodegradation of Mirex by
Sewage Sludge Organisms. Bull. Environ. Contamin. Toxicol. 11:415-416.
Andrade, Jr., P., W. B. Wheeler, and D. A. Carlson. 1975. Identification of a
Mirex Metabolite. Bull. Environ. Contamin. Toxicol. 14:473-479.
American Pesticide Control Officials, Inc. 1959. Pesticide Chemicals Official
Compendium. University of Maryland, College Park, Maryland.
Atkinson, B., and H. W. Fowler. 1974. The Significance of Microbial Film in
Fermenters. Adv. Biochem. Eng. 3:224-227.
316
-------
Aubry, M., M. N. Mayoral, and P. Villardry (1975), Bull. Soc. Chim. Fr.,
(3-4, Pt. 1) , 500-2.
Baker, R. D., and H. G. Applegate. 1970. Effect of Temperature and Ultraviolet
Radiation on the Persistence of Methyl Parathion and DDT in Soils. Agron.
J. 62:509-512.
Ball, J. S. 1962. Nitrogen Compounds in Petroleum. Proc. Amer. Pet. Institute
42(8) :27-30.
Bayly, R.C. , S. Dagley, and D. T. Gibson. 1966. The Metabolism of Cresols by
Species of Pseudomonas. Biochem. J. 101:293-301.
Ben'kovskii, V. G., A. Ya. Baikova, N. S. Lyubopytova, Y. N. Popov, E. A.
Kruglov, A. I. Stekhun, and F. N. Latypova. 1974. Nitrogenous Bases of
Diesel Fuel Produced by the Hydrogen Cracking of Arlam Petroleum Tar.
Neftekhimiya 14:891-898; Chemical Abstracts 82:127279w (1975).
Benson, S. W. 1968. Thermochemical Kinetics. John Wiley and Sons, Inc. New York.
Braids, 0. P., F. L. Hines, and G. W. Volk. 1967. The Occurrence of Carbazole
in a Peat Soil. Soil Sci. Amer. Proc. 31:435-436.
Brocco, D., et al. 1973. Determination of Aza-heterocyclic Compounds in
Atmospheric Dust by Combinations of Thin-layer and Gas Chromatography.
J. Chromat. 84:361-367.
Brodeskii, E. S., V. S. Stopskii, N. V. Rasumov, and E. V. Gusinskaya. 1975.
Refinement of the Structure-Group Characteristics of Nitrogenous Bases
from Sakhalia Petroleums. Neftekhimiya 15:464-469: Chemical Abstracts
83:16666n (1975).
Calvert, J. G., and J. N. Pitts, Jr. 1966. Photochemistry, John Wiley & Sons,
Inc., New York, pp. 780-782, 795-798.
Campbell, A. N. 1930. An Apparatus for the Determination of Solubility. J. Chem.
Soc. London, Part 1:179-180.
Carlson, D. A., K. D. Konyha, W. B. Wheeler, G. P. Marshall, and R. G. Zaylskie.
1976. Mirex in the Environment: Its Degradation to Kepone and Related
Compounds. Science 194:939-941.
Castellano, A., J. P. Catteau, and A. Lablache-Combier. 1975. Mechanism of the
Photochemical Substitution of 6-Membered Ring Monoazoaromatic Compounds by
Methanol and Neutral and HC1 Acidified Medium. Tetrahedron 31:2255-2261.
Chem. Eng. News 1976. Page 9. September 13, 1976.
Chi, C. T., J. A. Howell, and V. Pawlowsky. 1974. The Regions of Multiple
Stable Steady States of a Biological Reactor with Wall Growth, Utilizing
Inhibitory Substrates. Chem. Eng. Sci. 29:207-211.
317
-------
Christensen, H. E., and T. T. Luginbyhl (eds.). 1975. Registry of Toxic
Effects of Chemical Substances. U.S. Dept. of Health, Education and
Welfare, National Institute for Occupational Safety and Health, Rockville,
Maryland.
Christensen, H. E., E. J. Fairchild, B. S. Carroll, and R. J. Lewis, Sr. 1976.
Registry of Toxic Effects of Chemical Substances, 1976 Edition. National
Institute for Occupational Safety and Health, Rockville, Maryland, p. 404.
Glaus, D. 1964. The Decomposition of Toluene by Soil Bacteria. J. Gen.
Microbiol. 36:107-122.
Clemo, G. R. ,1973. Some Aromatic Basic Constituents of Coal Soot.
Tetrahedron 29:3987-3990.
Clugston, D. M., A. E. George, D. S. Montgomery, G. T. Smiley, and H. Sawatzky.
1976. Sulfur Compounds in Oils from the Western Canada Tar Belt. In:
Shale Oil, Tar Sands, and Related Fuel Sources, T. F. Yen, ed. Advances
in Chemistry Series 151, American Chemical Society, Washington, B.C. pp. 11-27,
Cobb, H. E., R. Atherton, and W. Olive. 1974. An Ecological Approach to the
Problem of Biodegradation of Phenolic Wastes. Second Year Progress Report
on AFOSR Grant 72-2356. University of Texas, San Antonio, Department of
Biology, July 1974, 81 p.
Cook, J. W., and N. D. Pugh. 1957. A Quantitative Study of Cholinesterase-
inhibiting Decomposition Products of Parathion Formed by Ultraviolet Light.
J. Assoc. Offie. Agr. Chem. 40:227.
Cowan, C. T., and D. White. 1962. "Adsorption by Organo-Clay Complexes," Clays
and Clay Minerals, Vol. 9, Proceedings of the Ninth Conferences on Clays
and Clay Minerals, Pergamon Press, pp. 459-467.
Cowart, R. P., F. L. Bonner, and E. A. Epps, Jr. 1971. Rate of Hydrolysis of
Seven Organophosphate Pesticides. Bull. Environ. Cont. Tox. 6:231-234.
CRC Handbook. 1974. 54th ed., R. C. Weast, Ed. in Chief. CRC Press,
Cleveland, Ohio.
Gripe, C. R., and R. J. Livingston. 1977. Dynamics of Mirex and Its Principal
Photoproducts in a Simulated Marsh System. Arch. Environ. Contain. Toxicol.
5 :295-303.
Dagley, S. , and M. D. Patel. 1957. Oxidation of _p_-Cresol and Related Compounds
by a Pseudomonas. Biochem. J. 66:227-233.
Daughton, C. G., and D.P.H. Hsieh. 1976. Fate of the Dialkyl Thiophosphoryl
Moiety of Parathion in a Bacterial Chemostat. Div. of Microb. and Biochem.
Technol., 172nd National Meeting, American Chemical Society. Abstract 5.
318
-------
Davis, W. W., and T. V. Parke, Jr. 1942. Determination of Solubilities of
Extremely Low Order. J. Amer. Chem. Soc. 64:101-7.
Doehler, R. W., and W. A. Young. 1960. Some Conditions Affecting the
Absorption of Quinoline by Clay Minerals in Aqueous Suspensions. Natl.
Conf. on Clays and Clay Minerals. 9:468-483.
Dreisback, R. R. 1955. Physical Properties of Chemical Compounds. In:
Advances in Chemistry, Series 15. American Chemical Society.
Eichelberger, J. W., and J. L. Lichtenberg. 1971. Persistence of Pesticides
in River Water. Environ. Sci. Tech. 5:541-544.
Erickson, F. B. 1954. 2-Sulfobenzoic Acid. U.S. Patent 2,642,458, June 16,
1953; Chemical Abstracts 48:5219.
Fedoseeva, G. E., A. Y. Khesina, M. N. Poglazova, L. M. Shabad, and M. N. Meisel.
1968. Oxidation of Various Aromatic Polycyclic Hydrocarbons by Micro-
organisms. Dokl. Akad. Nauk SSSR 183:208-211.
Filatova, A. S., A. L. Koz'minykh, and V. B. Kapitul'skii. 1973. Determination
of Some Aromatic Hydrocarbons in the Air of Coal Tar Processing Shops.
Gig. Tr. Prof. Zabol. 17:49-52; Biol. Abstracts 58:5208 (1974).
Ford, J. F., and V. 0. Young. 1965. Effect of Structure on the Rate of
Reaction of Organosulfur Compounds with Peracetic Acid and tert-Butyl
Hydroperoxide. Amer. Chemical Soc., Div. Petrol. Chem. Preprints
10(2) :C111-122.
Friello, D. A., J. R. Mylroie, and A. M. Chakrabarty. 1975. Use of Genetically
Engineered Multi-Plasmid Microorganisms for Rapid Degradation of Fuel
Hydrocarbons. In: Proceeding of the Third International Biodegradation
Symposium. J. Miles Sharpley and Arthur M. Kaplan, eds. Applied Science
Publishers Ltd., London, pp. 205-214.
Frost, A. A., and R. G. Pearson. 1961. Kinetics and Mechanisms, 2nd ed. John
Wiley and Sons, Inc., New York.
Funchess, M. J. 1917. The Nitrification of Pyridine, Quinoline, Quanidine
Carbonate, etc., in Soils. Alabama Agric. Expt. Sta., Bull. 196; Tech.
Bull. 3:65-82; Chemical Abstracts 11:3362 (1917).
Gaudy, A. F. Jr., M. Ramanathan, and B. S. Rao. 1967. Kinetic Behavior of
Heterogeneous Populations in Completely Mixed Reactors. Biotechnol.
Bioeng. 9:387-411.
Gibson, D. T. 1975. The Microbial Degradation of Aromatic Petroleum Products,
Annual Report No. 2, prepared for Office of Naval Research Under Contract
No. N00014-67-A-0126-0020, University of Texas at Austin, Austin, Texas
(30 June 1975).
319
-------
Gibson, D. T. 1975. Microbial Degradation of Polycyclic Aromatic Hydrocarbons.
In : Proceedings of the Third International Biodegradation Symposium,
J. M. Shapley and A. M. Kaplan, eds. Applied Science Publishers, Ltd.,
London. pp. 57-66.
Gibson, D. T. 1976. Microbial Metabolism of Polycyclic Aromatic Hydrocarbons.
Preprints of the Division of Petroleum Chemistry, Inc., 172nd Annual
Meeting of the Americal Chemical Society 21(3):409.
Gibson, D.T., V. Mahadevan, D. M. Jerina, H. Yagi, and H.J.C. Yeh. 1975.
Oxidation of the Carcinogens Benzo[a]pyrene and Benzo[a]anthracene to
Dihydrodiols by a Bacterium. Science 189:295-297.
Gibson, D. T., R. E. Roberts, M. C. Wells, and V. M. Kobal. 1973. Oxidation
of a Biphenyl by a Beijerinckia species. Biochem. Biophys. Res. Comm.
50:211-219.
Comma, H. M., and S. D. Faust. 1972. Chemical Hydrolysis and Oxidation of
Parathion and Paraoxon in Aquatic Environments in "Fate of Organic Pesti-
cides in Aquatic Environments." Amer. Chem. Soc. Publ. 111:189-209.
Graetz, D., G. Chesters, T. C. Daniel, L. W. Newland, and G. B. Lee. 1970.
Parathion Degradation in Lake Sediments. J. Water Pollut. Contr. Fed.
42:R76-R94.
Gregory, W. W., J. K. Reed, and L. E. Priester, Jr. 1969. Accumulation of
Paration and DDT by Some Algae and Protozoa. J. Protozool. 16:69-71.
Grunwell, R. R., and R. H. Erickson. 1973. Photolysis of Parathion. New
Products. J. Agr. Food Chem. 21:929-931.
Gunther, F. A., D. E. Ott, and F. E. Hearth. 1968. The Oxidation of Parathion
to Paraoxon in Aqueous Media by Silver Oxide. Bull. Environ. Cont. Tox.
3:49-57.
Gunther, F. A., D. E. Ott, and M. Ittig. 1970. The Oxidation of Parathion to
Paraoxon II by Use of Ozone. Bull. Environ. Cont. Tox. 5:87-94.
Raines, W. E., R. V. Helm, G. L. Cook, and J. S. Ball. 1956. Purification and
Properties of Ten Organic Sulfur Compounds—Second Series. J. Phys. Chem.
60:549-555.
Haque, R., and D. Schmedding. 1975. A Method of Measuring the Water Solubility
of Hydrophobic Chemicals: Solubility of Five Polychlorinated Biphenyls.
Bull. Environ. Contain. Toxicol. 14:13-18.
Harrison, R. M., R. Perry, and R. A. Wellings. 1975. Polynuclear Aromatic
Hydrocarbons in Raw, Potable and Waste Waters. Water Research 9 :331-346.
Hartung, G. K., and D. M. Jewell. 1962. Carbazole, Phenazines and Dibenzofuran
in Petroleum Products: Methods of Isolation and Determination. Anal. Chim.
Acta 26:514-528.
320
-------
Harwood, J. H., and S. J. Pirt. 1972. Quantitative Aspects of Growth of the
Methane Oxidizing Bacterium Methylococcus capsulatus on Methane in Shake
Flask and Continuous Chemostat Culture. J. Appl. Bact. 35:597-607.
Hashimoto, K. 1973. Oxidation of Phenols by Yeast. II. Oxidation of Cresols
by Candida tropicalis. J. Gen. Appl. Microbiol. 19:171-187.
Hayatsu, R., R. G. Scott, L. P. Moore, and M. H. Studier. 1975. Aromatic Units
in Coal. Nature 257:378-380.
Haynes, C. G., A. H. Turner and W. A. Waters. 1956. The Oxidation of Mono-
hydric Phenols by Alkaline Ferricyanide. J. Chem. Soc., 2823.
Hesse, P. R. 1971. A Textbook of Soil Chemical Analysis. Chemical Publishing
Company, Inc., New York. Chapter 11.
Hilgetag, G., and H. Teichman. 1965. The Alkylating Properties of Alkyl Thio-
phosphates. Angew Chem. 4:914-922.
Hill IV, J., et al. 1976. Dynamic Behavior of Vinyl Chloride in Aquatic Eco-
systems. U.S. Environmental Protection Agency, EPA-600/3-76-011.
Hirakoso, S. 1968. Inactivation of Some Insecticides by Bacteria in Mosquito
Breeding Polluted Water. Japan. J. Exp. Med. 38:327-334.
Hoffmann, D., G. Rathkamp, and S. Nesnow. 1969. Quantitative Determination
of 9-Methylcarbazoles in Cigarette Smoke. Anal. Chem. 41:1256-1259.
Hoffmann, D., G. Rathkamp, and H. Wuziwodzki. 1968. Studies on Tobacco Smoke.
VI. Determination of Carbazoles in Cigaret Smoke. Beitr. Tabakforsch.
4:253-263; Chem. Abstracts 71:196432 (1969).
Holden, C. 1976. Mirex: Persistent Pesticide on Its Way Out. Science 194:301-303.
Hollister, T. A., G. E. Walsh, and J. Forester. 1975. Mirex and Marine Uni-
cellular Algae: Accumulation, Population Growth and Oxygen Evolution.
Bull. Environ. Contamin. Toxicol. 14:753-759.
Holm-Hansen, 0., C. R. Goldman, R. Richards, P. M. Williams. 1976. Chemical
and Biological Characteristics of a Water Column in Lake Tahoe. Limn.
Ocean. 21(4) :548-562.
Hou, C. T., and A. I. Laskin. 1975. Microbial Conversion of Dibenzothiophene.
Developments Indust. Microbiol. 17:351-362.
Howard, J. A. 1972. Absolute Rate Constants for Reactions of Oxyl Radicals.
Adv. in Free Radical Chemistry 4:49-173.
Howell, J. A., C. T. Chi, and U. Pawlowsky. 1972. Effect of Wall Growth on
Scale-Up Problems and Dynamic Operating Characteristics of the Biological
Reactor. Biotechnol. Bioeng. 14:253-265.
321
-------
Hsieh, D.P.H., and M. Munnecke. 1972. Accelerated Microbial Degradation of
Concentrated Parathion. Proc, IV. Ferm. Symp. Ferment. Technol. 551-554.
Hudson, R. F. 1965. Structure and Mechanism in Organic Phosphorous Chemistry.
Academic Press, New York.
International Agency for Research on Cancer. 1973. Monograph on the Evaluation
of Carcinogenic Risk of the Chemical to Man: Certain Polycyclic Aromatic
Hydrocarbons and Heterocyclic Compounds. World Health Organization,
Geneva, Switzerland. Vol. 3.
Ivie, G. W., H. W. Borough, and E. G. Alley. 1974. Photodecomposition of
Mirex on Silica Gel Chromatoplates Exposed to Natural and Artificial
Light. J. Agr. Food Chem. 22(6):933-935.
Jaglan, P. S., and F. A. Gunther. 1970. Comparison of Hydrolysis Rates of
Methyl Parathion and Methyl Paraoxon by Gas Liquid Chromatography and
Spectrometry. J. Chrom. Sci. 8:483-485.
Jones, A. S., and C. S. Hodges. 1974. Persistence of Mirex and Its Effects on
Soil Microorganisms. J. Agr. Food Chem. 22:435-439.
Joussot-Dubien, J., and J. Houdard. 1967. Reversible Photolysis of Pyridine
in Aqueous Solution. Tetrahedron Letters 44:4389-4391.
Kaplin, V. T., L. V. Semenchenko, and E. G. Ivanov. 1968. Decomposition of a
Phenol Mixture in Natural Waters Miniature-scale Operation. Gidrokhim.
Mater. 45:199-202; Chemical Abstracts 69:69508h.
Karr, C. and L-C. Ta-chuang. 1958. Spectrophotometric Analysis of the
Distillable Low-Temperature Tar Bases. J. Inst. Eng. 31:522-527.
Karr, C., Jr., P. A. Estep, T.-C. L. Chang, and J. R. Comberiati. 1967.
Identification of Distillable Paraffins, Olefins, Aromatic Hydrocarbons,
and Neutral Heterocyclics from a Low-Temperature Bituminous Coal Tar.
Bureau of Mines Bulletin 637. pp. 59-61.
Kataoka, N., A. Imanura, Y. Kawazoe, G. Chihara, and C. Nagata. 1966. Electron
Spin Resonance Study on the Photo-induced Radicals from Related Compounds
to Carcinogenic 4-Nitroquinoline 1-Oxide. Chem. Pharm. Bull. (Tokyo) 14:1171.
Kaywer, W. V., K.C.D. Hickman, W. W. Bond, M. S. Favero, and L. A. Carson. 1975.
Bacteriological Evaluation of an Ultra-Pure Watertilling System. Appl.
Microbiol. 30:704-706.
Ketelaar, J. A., H. R. Gersmann, and M. M. Beck. 1956. Metal-Catalysed
Hydrolysis of Thiophosphoric Esters. Nature. 177:392.
Khesina, A. Y., N. P. Shcherbak, and L. M. Shabad. 1969. Destruction of
Benzo[a]pyrene by Soil Microflora. Byull. Eksp. Biol. Med. 68:70-73.
Chem. Abstracts 72 :44240V, 1970.
322
-------
King, P. H. , and P. L. McCarty. 1968. A Chromatographic Model for Predicting
Pesticide Migration in Soils. Soil Sci. 106:248-261.
Kirby, A. J. , and S. G. Warren. 1965. The Organic Chemistry of Phosphorous.
Elsevier, New York.
Kirschenbaum, I. 1961. Bacteriological Desulfurization of Petroleum. U.S.
Patent 2,975,103.
4
Kirso, U., and M. Gubergrits. 1972. Kinetics and Macromechanism of Oxidation
of Phenols of Various Structures by Molecular Oxygen in Aqueous Alkali.
J. Appl. Chem. (USSR) 45:847-851.
Klevens, H. B. 1950. Solubilization of Polycyclic Hydrocarbons. J. Phys.
Colloid. Chem. 54:283-298.
Kliger, L., and B. YaronJ 1973. Parathion Recovery from Soils after a Short
Contact Period. Bull. Environ. Cont. Tox. 13:714-719.
Knecht, Jr., A. T. (1962), Microbial Oxidation of Dibenzothiophene and Its
Possible Application in the Desulfurization of Coal and Petroleum,
Dissertation Absts. 22:3350.
Knoevenagel, K. and R. Himmelreich.. 1973. Photolysis of Kepone and Kelevan.
Z. Pflanzenkr. Pflanzenschutz 80(3-4):155-159; Chem Abstr. 80:70044t.
Knutson, H., A. M. Kadoum, T. L. Hopkins, G. F. Swoyer, and T. L. Harvey. 1971.
Insecticide Usage and Residues in a Newly Developed Great Plains Irrigation
District. Pesticides Monitor. J. 5:17-27.
Kobylinski, G. J., and R. J. Livingston. 1975. Movement of Mirex from Sediment
and Uptake by the Hogchoker, Trinectes maculatus. Bull. Environ. Contamin.
Toxicol. 14:692-698.
Kodama, K., S. Nakatani, K. Umehara, K. Shimizu, Y. Minoda, and K. Yamada. 1970.
Microbial Conversion of Petro-sulfur Compounds, Part III, Isolation and
Identification of Products from Dibenzothiophene. Agr. Biol. Chem.
34:1320-1324.
Kodama, K., K. Umehara, K. Shimizu, S. Nakatani, Y. Minoda, and K. Yamada.
1973. Identification of Microbial Products from Dibenzothiophene and
Its Proposed Oxidation Pathway. Agr. Biol. Chem. 37:45-50.
Kortus, J., P. Macuch, J. Mayer, K. Durcek, and V. Kremery. 1971. Uptake of
32p-Parathion and 32p Imidan. J. Hyg. Epidemiol. Microbiol. Immunol.
(Prague) 15 :101-103.
Kostenbader, P. D., and J. W. Flecksteiner. 1968. Biological Oxidation of
Coke Plant Weak Ammonia Liquor. Blast Furnace Steel Plant 56:475-480.
323
-------
Kurita, S., T. Endo, H. Nakamura, T. Yaai, and N. Tamiya (1971), Decomposition
of Some Organic Sulfur Compounds in Petroleum by Anaerobic Bacteria, J.
Gen. Appl. Microbiol. 17:185-198.
Laborde, A., and D. T. Gibson (1975), Bacterial Oxidation of Dibenzothiophene,
Abstracts of the Annual Meeting of the Amer. Soc. Microb. Q-45
Land, E. J., and G. Porter. 1963. Primary Photochemical Processis in Aromatic
Molecules. Trans. Far. Soc. 59:2016-2026.
Landa, S., V. Solin, K. Burianek, and B. Pisecka. 1953. Biological Degradation
of Phenols. I. Biological Oxidation of Monobasic Phenols. Chem. Listy.
47:622-629.
Lange's Handbook of Chemistry. 1973. John A. Dean, Ed. McGraw-Hill Book Company.
Langley, R. 1971. Practical Statistics Simply Explained. Dover Publications,
New York 199-211.
Larsen, D. H. , and R. L. Dimmick. 1964. Attachment and Growth of Bacteria on
Surfaces of Continuous-Culture Vessels. J. Bact. 88:1380-1387.
Leadbetter, E. R., and J. W. Foster. 1959. Oxidation Products Formed from
Gaseous Alkanes by the Bacterium Pseudomonas methanica. Archiv. Biochem.
Biophys. 82:491-492.
Leaver, I. H. 1971. Semiquinone Radical Intermediates in the Eosin-Sensitized
Photooxidation of Phenols. Austral. J. Chem. 24:891.
Lichtenstein, E. P., and K. R. Schulz. 1964. The Effects of Moisture and
Microorganisms on the Persistence and Metabolism of Some Organophosphorous
Insecticides in Soils with Special Emphasis on Parathion. J. Econ. Entomol.
57:618-627.
Lohrbacher, H., H. D. Puels, and H. W. Schlipkoeter. 1971. Studies on the
Storage and Metabolism of 3,4-Benzpyrene (Benzo[a]pyrene) in Microorganisms.
Zentralbl. Bakteriol. Parasitenkd. Infectionskr. Hyg. Erste Abt. Orig.
Reihe B. Hyg. Praev. Med. 155:168-174. (Biol. Abs. 56:436, 1973).
Mabey, W., and T. Mill. 1978. Critical Review of Hydrolysis of Organic Com-
pounds in Water Under Environmental Conditions. To be published in J.
Phys. Chem. Ref. Data 7:000.
Mackay, D., and A. W. Wolkoff. 1973. Rate of Evaporation of Low Solubility
Contaminants from Water Bodies to the Atmosphere. Environ. Sci. Tech.
7:611-614.
Maczynski, A., and Z. Maczynska. 1965. Vapor-Liquid Equilibrium in Binary
System Quinoline-Water. Bull. Acad. Pol. Sci. 13:299-302.
324
-------
Malaney, G. W., P. A. Lutin, J. J. Cibulka, and L. H. Hickerson. 1967.
Resistance of Carcinogenic Organic Compounds to Oxidation by Activated
Sludge. J. Water Pollution Control Adm. 39:2020-2029.
Maleszewska, J. 1974. Degradation of Methyl Parathion by Microorganisms
Occurring in Surface Water and Sewage. Pol. Arch. Hydrobiol. 21:163-171.
Maliyantz, A. A. 1935. Desulfurizing Crude Oil by Microbiological Methods.
Azerbaidzhanskoe Neftyanoe Khozyaistvo 6:89-93.
McEachern, D. M., et al. 1975. Enthalpies of Combustion and Sublimation and
Vapor Pressures of Three Benzoquinolines. J. Chem. Eng. Data 20:226-228.
McGinnes, P. R., and V. L. Snoeyink. 1974. Determination of the Fate of
Polynuclear Aromatic Hydrocarbons in Natural Water Systems. WRC Report
No. 80, PB 232 168. \
McKee, J. E., and H. W. Wolf (eds.). 1963. Water Quality Criteria, 2nd edition.
The Resources Agency of California, State Water Resources Control Board,
Sacramento , California.
McLafferty, F. W. 1973. Interpretation of Mass Spectra. 2nd ed. W. A.
Benjamin, Inc., Reading, Massachusetts, p. 118.
Medvedev, V. A., and V. D. Davydov. 1973. Transformation of Individual Coal-
Tar-Chemical- Industry Organic Products in a Chermozen Soil. Poehvodenie
pp. 22-28; Chem. Abstracts 78:706185 (1973).
Merck Index, The, 7th Edition. 1960. Merck and Co., Inc. Rahway, New Jersey.
Merkel, P. B., and D. R. Kearns. 1972. Remarkable Solvent Effects on the
Lifetime of 1-A Oxygen. J. Amer. Chem. Soc. 94:1029-1030.
6
Midwest Research Institute and RvR Consultants. 1975. Initial Scientific and
Minieconomic Review of Parathion. EPA-540/1-75-001 (PB-241 819/2WP), U.S.
Environmental Protection Agency, Office of Pesticide Programs.
Mill, T..v D. G. Hendry, W. R. Mabey, H. Richardson, B. Y. Lan, and A. Baraze.
1977. Oxidation of Organic Compounds in Dilute Aqueous Solution. Abstract.
173rd Meeting of the American Chemical Society, New Orleans, March 20-25.
PHYS. 222.
Monod, J. 1949. The Growth of Bacterial Cultures. Am. Rev. Microbiol.
3:371-394.
Moriconi, E. J., B. Rakoczy, and W. F. O'Connor. 1961. Ozonolysis of Poly-
cyclic Aromatics. VIII. Benzo[a]pyrene. J. Amer. Chem. Soc. 83:4618-4623.
Munnecke, D. M. 1976. Enzymatic Hydrolysis of Organophosphate Insecticides,
A Possible Pesticide Disposal Method. Appl. Environ. Microbiol. 32:7-13.
325
-------
Munnecke, D. M., and D.P.H. Hsieh. 1974. Microbial Decontamination of Parathion
and £-Nitrophenol in Aqueous Media. Appl. Microbiol. 28(2):212-217.
Munnecke, D. M., and D.P.H. Hseih. 1975. Microbial Metabolism of a Parathion-
Xylene Pesticide Formulation. Appl. Microbiol. 30(4):575-580.
Munnecke, D. M., and D.P.H. Hsieh. 1976. Pathways of Microbial Metabolism of
Parathion. Appl. Environ. Microbiol. 31(l):63-69.
Musso, H. 1967. "Phenol Coupling" in Oxidative Coupling of Phenols, W. I.
Taylor and A. R. Battersby (eds.). Marcel-Dekker, Inc., New York, pp. 1-94.
Nakatani, S., T. Akasaki, K. Kodama, Y. Minoda, and K. Yamada. 1968. Microbial
Conversion of Petro—sulfur Compounds Part II, Culture Conditions of Di-
benzothiophene—Utilizing Bacteria. Agr. Biol. Chem. 32:1205-1211.
National Academy of Science. 1972. Report of the Mirex Advisory Committee to
William D. Ruckelshaus, Administrator, Environmental Protection Agency.
National Academy of Science, Washington, D.C.
National Academy of Sciences; 1972. Particulate Polycyclic Organic Matter.
National Academy of Sciences, Washington, D.C.
Naucke, W., H. V. Laaser, and F. Tarkmann. 1972. Water-Soluble Organic Sub-
stances in Slightly Decomposed Sphagnum Peats. Fourth Proceedings of
Intern. Peat Congress 4:45-60.
Naumann, K. 1967. Decomposition of Parathion by Soil Bacteria. Phytopathol.
Z. 60:343-357.
Niaussat, P., and J. Ottenwalder. 1969. Apparation of Benzo (3,4) pyrene in
Cultures of Marine Phyto-plankton; Incidence of Associate Bacteria. Rev.
Hyg. Med. Soc. 17:487-496. (Biol. Abstracts 53:26642, 1972).
Oka, S. 1962. Transfer of Antiseptics of Microorganisms and Their Toxic Effects.
VI. Toxic Effect and Adsorption of Phenols and Esters of Acid Antiseptics
on Bacterial Cells. Agr. Biol. Chem. (Tokyo) 20:515-525.
Pawlowsky, V., and J. A. Howell. 1973a. Mixed Culture Bioxidation of Phenol.
I. Determination of Kinetic Parameters. Biotechnol. and Bioeng.
15(5)-.889-896.
Pawlowsky, V., and J. A. Howell. 1973b. Mixed Culture Biooxidation of Phenol.
II. Steady State Experiments in Continuous Culture. Biotechnol. Bioeng.
15(5):897-903.
Lawlowsky, V., J. A. Jowell, and C. T. Chi. 1973. Mixed Culture Biooxidation
of Phenol. III. Existence of Multiple Steady States in Continuous
Culture with Wall Growth. Biotechnol Bioeng. 15(5):905-916.
Perrin, D. D. 1965. Dissociation Constants of Organic Bases in Aqueous
Solution. Butterworth, London.
326
-------
Pfoertner, K., and D. Boese. 1970. Die Photosensibilisierte Oxydation
einwertiger Phenols zu Chinonen. Helv. Chim. Acta 53:1553-1566.
Phillips, L. S., and C. Hinshelwood. 1953. Adaption of Aerobacter aerogenes
to Resist Phenols and Various Phenols. J. Chem. Soc., 3679-3683.
Pitts, J. N., Jr., J.K.S. Wan, and E. A. Schuck. 1964. Photochemical Studies
in an Alkali Halide Matrix. I. An o-Nitrobenzaldehyde Actinometer and Its
Application to a Kinetic Study of the Photoreduction of Benzophenone by
Benzhydrol in a Pressed Potassium Bromide Disk. J. Amer. Chem. Soc.
86:3606-3610.
Poglazova, M. N., and M. N. Meisel. 1971. Localization of Benzo[a]pyrene
in Bacterial Cells. I Mikrobiologiya 40:1050-1053.
Poglazova, M. N., G. E. Fedoseeva, A. Y. Khesina, M. N. Meisel, and L. M. Shabad.
1966. Modification of Benzpyrene by Soil Microorganisms. Dokl. Akad.
Nauk. SSSR 169 :1174-1177.
Poglazova, M. N., G. E. Fedoseeva, A. J. Khesina, M. N. Meissei, and L. M.
Shabad. 1967a. Destruction of Benzo[a]pyrene by Soil Bacteria. Life
Sciences 6:1053-1062.
Poglazova, M. N., G. E. Fedoseeva, A. J. Khesina, M. N. Meissei, and'L. M.
Shabad. 1967b. Further Investigations of the Decomposition of Benzo[a]
pyrene by Soil Bacteria. Dokl. Biol. Sci. 176:1165-1167.
Poglazova, M. N., G. E. Fedoseeva, A. J. Khesina, M. N. Meissei, and L. M.
Shabad. 1968. The Oxidation of Benzo[a]pyrene by Microorganisms in
Relation to Its Concentration in the Medium. Dokl. Biol. Sci. 179:1460-
1462.
Poglazova, M. N., G. E. Fedoseeva, A. Y. Khesina, M. N. Meisel, and L. M. Shabad.
1971. Benzofa]pyrene Metabolism by Microflora of Various Soils and
Separate Species of Microorganisms. Dokl. Akad. Nauk SSSR 198:1211-1213.
Poglazova, M. N., A. Y. Khesina, G. E. Fedoseeva, M. N. Meisel, and L. M. Shabad.
1972. Destruction of Benzo[a]pyrene in Waste Waters by Microorganisms.
Dokl. Akad. Nauk SSSR 204:222-225.
Ponec, R., and M. Prochazka. 1974. Effect of Substitution on Oxidation of
Sulfides. Coll. Czech. Chem. Commun. 39:2088-2098. Chem. Abstr.
82:111295n (1975).
Poulson, R. E., C. M. Frost, and H. B. Jensen. 1974. Characteristics of
Synthetic Crude from Crude Shale Oil Producted by in situ Combustion
Retorting. Amer. Chem. Soc., Div. Fuel Chem., Prepr. 19(2):175-182.
Present, D. 1958. Kinetic Theory of Gases. McGraw-Hill Book Company, New
York. 280pp.
327
-------
Pupp, C., R. C. Lao, J. J. Murray, R. F. Pottie. 1974. Equilibrium Vapor
Concentrations of Some Polycyclic Aromatic Hydrocarbons, Arsenic Trioxide
(As406) and Selenium Dioxide, and the Collection Efficiencies of these
Air Pollutants. Atmos. Environ. 8:915-25.
Radding, S. B., T. Mill, C. W. Gould, D. H. Liu, H. L. Johnson, D. C. Bomberger,
and C. V. Fojo. 1976. The Environmental Fate of Selected Polynuclear
Aromatic Hydrocarbons. EPA 560/2-75-009, Environmental Protection Agency,
Washington, D.C., February 1976. 122pp.
Ray, S., and R. W. Frei. 1972. Separation of Polynuclear Aza-heterocyclics
by High Speed Liquid Chromatography on a Chemically Bonded Stationary Phase.
J. Chromatography. 71:451-457.
Reid, R. C., and T. K. Sherwood. 1966. The Properties of Gases and Liquids.
2nd ed. McGraw-Hill Book Company, New York.
Robbins, W. J. 1916. The Cause of the Disappearance of Coumarin, Vanillin,
Pyridine and Quinoline in the Soil. Science 44:894-895.
Robbins, W. J. 1917. The Cause of the Disappearance of Coumarin, Pyridine, and
Quinoline in the Soil. Alabama Agric. Expt. Station, Bull. 195; Tech. Bull.
2:49-64; Chemical Abstracts 11:3362 (1917),
Robbins, W. R. 1971. Student Projects on the Oxidation by Marine Bacteria of
Aromatic Compounds Found in Oil. Massachusetts Institute of Technology,
Report No. MITSG 71-10, Cambridge, Mass., June 28, 1971.
Ruzicka, J. H., J. H. Thomson, and B. B. Wheals. 1967. The Gas Chromatographic
Determination of Organophosphorous Pesticides. II. A Comparative Study of
Hydrolysis Rates. J. Chromatog. 31:37-47.
Sagardia, F., J. J. Rigau, A. Martinez-Lahoz, F. Fuentes, C. Lopez, and W.
Flores. 1975. Degradation of Benzothiophene and Related Compounds by a
Soil Pseudomonas in an Oil Aqueous Environment, Applied Microbiol. 29:722-725.
Sawicki, E. et al. 1965. Use of Gas-liquid and Thin-layer Chromatography in
Characterizing Air Pollutants by Fluorimetry. Talania. 13:619-629.
Chemical Abstracts.
Saltzman, S., U. Mingelgrin, and B. Yaron. 1976. Role of Water in the Hydrolysis
of Parathion and Methyl Parathion on Kaolinite. J. Agric. Food Chem.
24(4):739-743.
Saltzman, S., and B. Yaron. 1972. Parathion Adsorption from Aqueous Solutions
as Influenced by Soil Components. Pestic. Chem., Proc. Int. Congr. Pestic.
Chem. 2nd. 6:87-100.
Sandi, E. 1959. Beitrage zur Analyse einiger insecticider Thiophosphorsaureester.
Beit. Anal. Chem. 167:241-245.
328
-------
Schmidt, C. E., A. F. Sharkey, Jr., and R. A. Friedel. 1974. Mass Spectro-
metric Analysis of Product Water from Coal Gasification. Bureau of Mines
Advancing Energy Utilization Technology Program, Technical Progress Report
86, December 1974.
Sethunathan, N. 1973. Parathion Degradation in Submerged Rice Soils in the
Philippines. J. Agr. Food Chem. 21:504-506.
Sethunathan, N., and T. Yoshida. 1973. Parathion Degradation in Submerged
Rice Soils in the Philippines. J. Agr. Food Chem. 21:504-506.
Shabad, L. M. 1968. The Distribution and the Fate of the Carcinogenic
Hydrocarbon Benzo[a]pyrene. Z. Krebsforsch. 70:204-210.
Shabad, L. M., Y. L..Cohan,'A. P. Ilnitsky, A. Y. Khesina, N. P. Shcherbak, and
G. A. Smirnov. 1971. .The Carcinogenic Hydrocarbon Benzo[a]pyrene in the
Soil. J. Natl. Cancer Inst. 47:1179-1192.
Sidderamappa, R., K. P. Rajaram, and N. Sethunathan. 1973. Degradation of
Parathion by Bacteria Isolated from Flooded Soil. Appl. Microbiol.
26(6):846-849.
Silvey, J.K.G., and A. W. Roach. 1975. The Taste and Odor Producing Aquatic
Actinomycetes. CRC Critical Reviews in Environmental Control 5:233-273.
Sisler, F. D., and C. Zobell 1947. Microbial Utilization of Carcinogenic
Hydrocarbons. Science 106:521-522.
Slater, J. H., A. T. Bull, and E. Senior. 1976. Microbial Interaction-
Competition. Proceed. Soc. Gen. Microbiol. 3:103.
Smith, L. D. 1976. Inhibition of Clostridium betulinum by Strains of
Clostridium perfringens Isolated from Soil. Appl. Microbiol. 30:319-323.
Snyder, L. R. 1969. Nitrogen and Oxygen Compound Types in Petroleum. Total
Analyses of a 400-700°F Distillate from a California Crude Oil. Anal.
Chem. 41:314-324.
Stein, V. B., K. A. Pittman, and M. W. Kennedy. 1976. Characterization of a
Mirex Metabolite from Monkeys. Bull. Environ. Contam. Toxicol. 15:140-146.
Stevens, B., and B. E. Alger. 1968. The Photoperoxidation of Unsaturated
Organic Molecules; II. The Autoperoxidation of Aromatic Hydrocarbons.
J. Phys. Chem. 72:3468-3474.
Stevens, B., S. R. Perez, and J. A. Ors. 1974. Photoperoxidation of Unsaturated
Organic Molecules: XIV. 02 ^Ag Acceptor Properties and Reactivity. J.
- Amer. Chem. Soc. 96:6846-6850.
Strawinski, R. J. 1950. Method of Desulfurizing Crude Oil, U.S. Patent
2,521.701.
329
-------
Strawinski, R. J. 1951. Purification of Substances by Microbial Action.
U.S. Patent 2,574,070.
Suess, M. J. 1972. Aqueous Solutions of 3,4-Benzpyrene. Water Research
6:981-985.
Suess, M. J. 1976. The Environmental Load and Cycles of Polycyclic Aromatic
Hydrocarbons. In: International Conference on Environmental Screening
and Assessment, Las Vegas, Nevada, September 14-19, 1975. Institute of
Electrical and Electronic Engineers, Inc.
Suta, B. E. 1977. Human Population Exposures to Mirex and Kepone. Discussion
Draft under EPA ORD Contract 68-01-4314.
Tagatz, M. E., P. W. Bothwick, J. M. Ivey, and J. Knight. 1976. Effects of
Leached Mirex on Experimental Communities of Estaurine Animals. Arch.
Environ. Contam. Toxicol. 4(4) :435-442.
Topiwala, H. H., and G. Hamer. 1971. Effect of Wall Growth in Steady State
Continuous Cultures. Biotechnol. Bioengin. 13(6):919-922.
Tsivoglou, E. C., et al. 1965. Tracer Measurements of Atmospheric Reaeration-I.
Laboratory Studies. J. Water Pollution Control Federation 37:1343-1362.
Utkina, L. F. 1965. Identification of Some Aromatic Hydrocarbons in Coal Tar on
the Basis of Quasilinear Luminescence and Absorption Spectra at Low
Temperatures. Tr. Mezhvuz. Konf. Peddgog. Inst. Radiofiz. Specktrosk.
1st Moscow. 53-8; Chem. Abstracts 66:67743a (1967).
f
Vernon, C. A. 1957. Phosphoric Esters and Related Compounds. The Chem. Soc.,
Spec. Publ. 8, London.
von Rumker, R., E. W. Lawless, A. F. Meiners, K. A. Lawrence, G. L. Kelar, and
F. Horay. 1974. Production, Distribution, Use and Environmental Impact
Potential of Selected Pesticides. EPA 540/1-74-001, U.S. Environmental
Protection Agency, Office of Pesticide Programs.
Vymetal, J. 1974a. Quinoline Bases in Black Coal Tar. Chem. Listy 68:1234-
1253; Chemical Abstracts 82:75217c (1975).
Vymetal, J. 1974b. Separation by Rectification of Pyridine Bases from Products
of High Temperature Carbonization of Coal. Erdoel Kohle, Erdgas, Petrochem.
Brennst.-Chem. 27:8-12; Chemical Abstracts 81:138391g (1974).
Walker, J. D., and R. R. Colwell. 1974. Microbial Petroleum Degradation:
Use of Mixed Hydrocarbon Substrates. Applied Microbiol. 27:1053-1060.
Walker, J. D., and R. R. Colwell. 1975. Effects of Petroleum on Estuarine and
Marine Microorganisms. Can. J. Microbiol. 21:305-13.
330
-------
Walker, J. D., R. R. Colwell and L. Petrakis (1975), Microbial Petroleum
Degradation: Application of Computerized Mass Spectrotnetry, Can. J.
Microbiol. 21:1760-1767.
Walker, J. D., R. R. Colwell, and L. Petrakis, 1976. Biodegradation of
Petroleum by Chesapeake Bay Sediment Bacteria. Can. J. Microbiol.
22 :423-428.
Walling, C., and C. G. Papaioannou. 1968. Bond Dissociation Energies of
t-Butyl Hypohalites. J. Phys. Chem. 72:2260-2262.
Waters, W. A. 1964. Mechanism of Oxidation of Organic Compounds. Methuen and
Co., Ltd., London, 152 pages.
Weber, K. 1976. Degradation of Parathion in Seawater. Water Research 105:
237-241. \
Wiberg, K. B. 1965. Oxidation in Organic Chemistry. Academic Press, New York,
443 pages.
Wilkinson, T. G., H. H. Topiwala, and G. Hamer. 1974. Interactions in a Mixed
Bacterial Population Growing on Methane in Continuous Culture. Biotechnol.
and Bioengin. 16:41-59.
Williamson, K., and P. L. McCarty. 1976a. A Model of Substrate Utilization by
Bacterial Films. J. Water Pollut. Control Fed. 48(1):9-24.
Williamson, K., and P. L. McCarty. 1976b. Verification Studies of the Biofilm
Model for Bacterials Substrate Utilization. J. Water Pollut. Control Fed.
48(2) :281-296.
Wise, J. 1975. The Lake Tahoe Study, as Requested by the 92nd Congress in
Section 114 of The Federal Water Pollution Control Act Amendments of
1972. U.S. Environmental Protection Agency.
Wodzinski, R. S., and D. Bartolini. 1972. Physical State in Which Naphthalene
and Bibenzyl are Utilized by Bacteria. Appl. Microbiol. 23:1077-1081.
Wodzinski, R. S., and J. E. Coyle. 1074. Physical State of Phenanthrene for
Utilization by Bacteria. Appl. Microbiol. 27:1081-1084.
Wodzinski, R. S., and M. J. Johnson. 1968. Yields of Bacterial Cells from
Hydrocarbons. Appl. Microbiol. 16:1886-1891.
Wodzinski, R. S., and D. Larocca. 1977. Bacterial Growth Kinetics on Diphenyl-
methane and Naphtharene-Heptamethylnonane Mixtures. Appl. Microbiol.
33:660-665.
Wolfe, N. L., R. G. Zepp, G. L. Baughman, R. C. Fincher, and J. A. Gordon.
1976. Chemical and Photochemical Transformation in Selected Pesticides in
Aquatic Systems. U.S. Environmental Protection Agency, EPA-600/3-76-067.
331
-------
Wu, W.R.K., and H. H. Storch. 1967. Hydrogenation of Coal and Tar. Bulletin
633, U.S. Dept. of the Interior, Bureau of Mines, pp. 154.
Yamada, K. , Y. Minoda, K. Kodama, S. Nakatani, and T. Akasaki. 1968. Microbial
Conversion of Petroleum Compounds Part 1, Isolation and Identification of
Dibenzothiophene Utilizing Bacteria. Agr. Biol. Chem. 32:840=845.
Yang, R. D., and A. E. Humphrey. 1975. Dynamic and Steady State Studies of
Phenol Biodegradation in Pure and Mixed Cultures. Biotechnol. Bioeng.
17:1211-1235.
Yasuno, M., S. Hirakoso, M. Sasa, and M. Uchida. 1965. Inactivation of Some
Organophosphorous Insecticides by Bacteria in Polluted Water. Jap. J.
Exp. Med. 35(6) :545-563.
Zepp, R. G., and D.M. Cline. 1977. Rates of Direct Photolysis in Aquatic
Environments. Environ. Sci. Technol. 11:359-366.
Zobell, E. 1943. The Effect of Solid Surfaces upon Bacterial Activity.
J. Bact. 46:39-56.
Zobell, C. E. 1953. Process of Removing Sulfur from Petroleum Hydrocarbons
and Apparatus. U.S. Patent 2,641,564.
Zherebtsov, I. P., and V. P. Lopatinskii. 1970. Izv. Tomsk. Politekh. Inst.
163:12-17; Chem. Abstracts 75:117862x (1971).
Zuckermann, B. M., K. Deubert, M. Mackiewicz, and H. Gunner. 1970. Studies on
Biodegradation of Parathion. Plant Soil. 33(2):273-281.
332
-------
Appendix A
INPUT DATA FOR NINE-COMPARTMENT MODEL
The nine—compartment computer model was used to simulate the transport and
transformation of eleven compounds in a pond, a river, a eutrophic lake, and an
oligotrophic lake. The sorption partition coefficients and the rate constants
used as input to the computer model are given in Tables A.I through A.43. All
rate and equilibrium values are at 25°C.
TABLE A.I. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR 2-CRESOL:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient 10 10
Pseudo-first-order rate constants
Volatilization (hr"1) 0 0
Photolysis (hr-1) 2 x 10~'t 0
Oxidation (hr-1) 0 0
Hydrolysis (hr-1) 0 0
Kinetic constants for biodegradation
Bacterial population 106 107
(cells ml-1)
Active bacterial population 105 105
(cells ml-1)
Maximum growth rate (hr"1) 0.56 0.28
Half-saturation constant 0.99 1.98
(yg ml-1)
Yield factor (cells ug-1) 106 106
333
-------
U)
to
TABLE A. 2. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR £-CRESOL:
RIVER SIMULATION
Water compartments
Sorption partition coefficient
Pseudo-first-order rate
constants
Volatilization (hr"*1)
Photolysis (hr"1)
Oxidation (hr-1)
Hydrolysis (hr-1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-*)
Active bacterial population
(cells ml-1)
Maximum growth rate (hr'1)
Half -saturation constant
1
10
0
5 x 10-"*
0
0
106
105
0.56
0.99
2
10
0
5 x I0~k
0
0
106
10s
0.56
0.99
3
10
0
5 x 10-1*
0
0
106
105
0.56
0.99
Sediment compartments
7
10
0
0
0
0
107
105
0.28
1.98
8
10
0
0
0
0
10?
10s
0.28
1.98
9
10
0
0
0
0
107
10s
0.28
1.98
(Pg ml-1)
Yield factor (cells ug""1)
106
106
106
106
10 6
10 6
-------
U)
ui
Ul
TABLE A. 3. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR £-CRESOL:
EUTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr"1)
Oxidation (hr-1)
Hydrolysis (hr"1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-1)
Active bacterial popula-
tion (cells ml"1)
Maximum growth rate
(hr-1)
Half-saturation constant
(ug ml-1)
Yield factor (cells ug"1)
1
10
0
2 x 10"1*
0
0
106
10s
0.56
0.99
106
2
10
0
2 x 10'1*
0
0
106
1Q5
0.56
0.99
106
3
10
0
2 x 10"1*
0
0
106
105
0.56
0.99
106
5
10
0
2 x 10~5
0
0
10s
105
0.38
1.5
106
Sediment compartments
7
10
. . ... _- .
0
0
0
0
10?
105
0.28
1.98
106
8
10
0
0
0
0
10?
105
0.28
1.98
106
9
10
0
0
0
0
10?
105
0.28
1.98
10 6
-------
TABLE A.A. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR 2.-CRESOL:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr-1)
Photolysis (hr-1)
Oxidation (hr-1)
Hydrolysis (hr-1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-1)
Active bacterial popu-
lation (cells ml-^)
Maximum growth rate
(hr-1)
Half-saturation constant
Gig ml'1)
Yield factor (cells ug-1)
1
10
0
1 x 10~2
0
0
102
10
0.56
0.99
10s
2
10
0
1 x 10~3
0
0
102
10
0.56
0.99
106
3
10
0
1 x 10~3
0
0
102
10
0.56
0.99
106
5
10
0
1 x lO"4
0
0
102
10
0.38
1.5
106
Sediment compartments
7
10
0
0
0
0
107
10
0.28
1.98
106
8
10
0
0
0
0
107
10
0.28
1.98
106
9
10
0
0
0
0
107
10
0.28
1.98
106
-------
TABLE A. 5. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZ[a]ANTHRACENE:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient 2.5 x 101* 2.5 x I0k
Pseudo-first-order rate constants
Volatilization (hr-1) 7.0 x 10~6 0
Photolysis (hr"1) j ' 1.4 x 10~z 0
Oxidation (hr"1) 0 0
Hydrolysis (hr"1) 0 0
Biodegradation (hr-1) 0 0
337
-------
U)
U)
oo
TABLE A. 6. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZ[a]ANTHRACENE:
RIVER SIMULATION
Water compartments Sediment compartments
2378
Sorption partition coefficient 2.5 x lO1* 2.5 x 101* 2.5 x 104 2.5 x 101* 2.5 x 101* 2.5 x 101*
Pseudo- first-order rate
constants
Volatilization (hr'1) 7.0 x 10~6 7.0 x 10~6 7.0 x 10"6 000
Photolysis (hr""1) 3.4 x 10~2 3.4 x 10~2 3.4 x 10~2 000
Oxidation (hr'1) 000 000
Hydrolysis (hr"1) 000 000
Biodegradation (hr'1) 0 0 0 0 00
-------
TABLE A.7. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZ[a]ANTHRACENE:
EUTROPHIC LAKE SIMULATION
Water compartments
Sediment compartments
8
Ul
u>
VO
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr-1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
2.5 x 104 2.5 x 101* 2.5 x 101* 2,5 x 101* 2.5 x 10** 2.5 x 101* 2.5
7.0 x 10~6 7.0 x 10~6 7.0 x 10"6 0
0
1.4 x 10~2 1.4 x 10"2 1.4 x 10"2 1.4 x 10~3 0
0 0 0 00
00000
00000
0
0
0
0
0
0
0
0
0
0
-------
U)
£»
O
TABLE A. 8. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZ[a]ANTHRACENE:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sorptlon partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr'1)
Oxidation (hr"1)
Hydrolysis (hr'1)
Biodegradation (hr"1)
1
2.5 x 101*
7.0 x 10~6
6.9 x 10~2
0
0
0
2
2.5 x 1011
7.0 x 10~6
6.9 x 10~2
0
0
0
3
2.5 x I0k
7.0 x 10~6
6.9 x 10~2
0
0
0
5
2.5 x 101*
0
6.9 x 10~3
0
0
0
Sediment compartments
7
2.5 x 101*
0
0
0
0
0
8
2.5 x 101*
0
0
0
0
0
9
2.5 x 101*
0
0
0
0
0
-------
TABLE A. 9. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[a]PYRENE:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient 5 x 101* 5 x 101*
Pseudo-first-order rate constants
Volatilization (hr-1) 2.0 x 10"5 0
Photolysis (hr-1) ; 9.2 x 10" 2 0
Oxidation (hr-1) 0 0
Hydrolysis (hr"1) 0 0
Biodegradation (hr"1) 0 0
341
-------
TABLE A.10. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[a]PYRENE:
RIVER SIMULATION
Water compartments
Sediment compartments
8
to
Sorption partition coefficient 5.0 x 101* 5.0 x 101* 5.0 x 101* 5.0 x 101* 5.0 x
Pseudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr-1)
Oxidation (hr'1)
Hydrolysis (hr'1)
Biodegradation (hr'1)
5.0 x 10~3 5.0 x 10~3 5.0 x 10~3 0
2.3 x 10"1 2.3 x 10""1 2.3 x 10-1 0
000 0
000 0
000 0
0
0
0
0
0
5.0 x
0
0
0
0
0
-------
TABLE A.11. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[a]PYRENE:
EUTROPHIC LAKE SIMULATION
Water compartments
Sediment compartments
8
ui
*»
W
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr-1)
Oxidation (hr~J)
Hydrolysis (hr"1)
Biodegradation (hr"1)
5.0 x 101* 5.0 x 10** 5.0 x 104 5.0 x 101* 5.0 x 101* 5.0 x 104 5.0 x 104
1.0 x 10"3 1.0 x 10~3 1.0 x 10~3 0 0
9.2 x 10"2 9.2 x 10"2 9.2 x 10~2 9.2 x 10~3 0
00000
00000
00000
0
0
0
0
0
0
0
0
0
0
-------
Ul
TABLE A.12. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[a]PYRENE:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr-1)
Photolysis (hr-1)
Oxidation (hr"1)
Hydrolysis (hr'1)
Blodegradation (hr"1)
1
5.0 x ID4
1.0 x 10"3
4.6 x 10"1
0
0
0
2
5.0 x 1011
1.0 x 10~3
4.6 x 10"1
0
0
0
3
5.0 x 101*
1.0 x 10"3
4.6 x 10'1
0
0
0
5
5.0 x 104
0
4.6 x 10~2
0
0
0
Sediment compartments
7
5.0 x 10k
0
0
0
0
0
8
5.0 x 101*
0
0
0
0
0
9
5.0 x 104
0
0
0
0
0
-------
TABLE A.13. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR QUINOLINE:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient 0 0
Pseudo-first-order rate constants
Volatilization (hr-1) 0 0
Photolysis (hr-1) ; 2 x 10~A °
Oxidation (hr"1) 0 0
Hydrolysis (hr-1) 0 0
Kinetic constants for biodegradation
Bacterial population 106 107
(cells ml-1)
Active bacterial population 105 105
(cells ml-1)
Maximum growth rate (hr-1) 0.74 0.37
Half-saturation constant 0.16 0.32
Gig ml-1)
Yield factor (cells pg"1) 3.3 x 106 3.3 x 106
345
-------
U)
TABLE A.14. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR QUINOLINE:
RIVER SIMULATION
Water compartments
Sorption partition coefficient
Paeudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr"1)
Oxidation (hr-1)
Hydrolysis (hr-1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-1)
Active bacterial population
(cells ml-1)
Maximum growth rate (hr"1)
Half -saturation constant
1
0
0
6 x 10"1*
0
0
106
105
0.74
0.16
2
0
0
6 x 10-1*
0
0
106
10*
0.74
0.16
3
0
0
6 x 10-1*
0
0
10 6
10s
0.74
0.16
Sediment compartments
7
0
0
0
0
0
107
10*
0.37
0.32
8
0
0
0
0
0
107
105
0.37
0.32
9
0
0
0
0
0
107
105
0.37
0.32
(Wg ml-1)
Yield factor (cells
3.3 x 10"6 3.3 x 10"6 3.3 x 10~6 3.3 x 10~6 3.3 x 10"6 3.3 x 10~6
-------
TABLE A.15. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR QUINOLINE:
EUTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr-1)
Oxidation (hr"1)
Hydrolysis (hr'1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-1)
Active bacterial popula-
tion (cells ml-1)
Maximum growth rate
(hr-1)
Half-saturation constant
1
0
0
2 x 10-1*
0
0
106
105
0.74
0.16
2
0
0
2 x 10-"
0
0
106
105
0.74
0.16
3
0
0
2 x 10-"
0
0
106
105
0.74
0.16
5
0
0
2 x 10"5
0
0
106
10 5
0.50
0.24
Sediment compartments
7
0
0
0
0
0
107
10s
0.37
0.32
8
0
0
0
0
0
107
105
0.37
0.32
9
0
0
0
0
0
107
10s
0.37
0.32
(ug ml-1)
Yield factor (cells ug-1) 3.3 x 10~6 3.3 x 10~6 3.3 x 10~6 3.3 x 10~6 3.3 x 1Q-6 3.3 x 10"6 3.3 x 10~6
-------
(jj
TABLE A.16. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR QUINOLINE:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr-1)
Photolysis (hr-1)
Oxidation (hr-1)
Hydrolysis (hr'1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml'1)
Active bacterial popu-
lation (cells ml-1)
Maximum growth rate
(hr-1)
Half-saturation constant
1
0
0
1.2 x ID'3
0
0
102
10
0.74
0.16
2
0
0
1.2 x 10"3
0
0
102
10
0.74
0.16
3
0
0
1.2 x 10~3
0
0
102
10
0.74
0.16
5
0
0
1.2 x 10-1*
0
0
102
10
0.50
0.24
Sediment compartments
7
0
0
0
0
0
103
10
0.37
0.32
8
0
0
0
0
0
103
10
0.37
0.32
9
0
0
0
0
0
103
10
0.37
0.32
(yg ml'1)
Yield factor (cells yg-1) 3.3 x 106 3.3 x 106 3.3 x 106 3.3 x 106 3.3 x 106 3.3 x 106 3.3 x 106
-------
TABLE A.17. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[f]QUINOLINE:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient 1.4 x 103 1.4 x 103
Pseudo-first-order rate constants
Volatilization (hr"1) 1.6 x 10~6 0
Photolysis (hr"1) 0.1 0
Oxidation (hr-1) 1.0 x 10~7 0
Hydrolysis (hr"1) 0 0
Biodegradation (hr"1) 3.6 x 10~3 3.6 x 10~3
349
-------
TABLE A.18. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[f]QUINOLINE:
RIVER SIMULATION
Water compartments
Sediment compartments
8
00
Ul
o
Sorptlon partition coefficient 1.4 x io3 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr-1)
Hydrolysis (hr-1)
Biodegradation (hr""1)
8.0 x 10~6 8.0 x IO"6 8.0 x IO"6
0.25
0.25
0.25
1.0 x 10~7 1.0 x 10~7 1.0 x 10"?
3.6 x 10~3 3.6 x 10~3 3.6 x IO"3
0
0
0
0
0
0
0
0
0
0
0
0
1.8 x 10~3 1.8 x 10~3 1.8 x 10~3
-------
TABLE A.19. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[f]QUINOLINE:
EUTROPHIC LAKE SIMULATION
Water compartments
Sediment compartments
8
U)
in
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103
2.0 x 10~6 2.0 x 10"6 2.0 x 10~6 0
0.10
0.10
0.10
1.0 x 10~2 0
1.0 x 10~? 1.0 x 10~7 1.0 x 10~7 1.0 x 10~7 0
3.6 x 10~3 3.6 x 10~3 3.6 x 10~3 2.5 x 10"3 1.8 x 10~3 1.8 x 10~3 1.8 x 10~3
0
0
0
0
0
0
0
0
-------
TABLE A.20. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[f]QUINOLINE:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
m constants
to
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
1
1.4 x 103
2.0 x 10~6
0.50
1.0 x 10~?
0
3.6 x 10~7
2
1.4 x 103
2.0 x 10~6
0.50
1.0 x 10~7
0
3.6 x 10"7
3
1.4 x 103
2.0 x 10~6
0.50
1.0 x 10"7
0
3.6 x 10~7
5
1.4 x 103
0
5.0 x 10"2
1.0 x 10~7
0
2.5 x 10~7
Sediment compartments
7
1.4 x 103
0
0
0
0
1.8 x 10~7
8
1.4 x 103
0
0
0
0
1.8 x 10~7
1,4
0
0
0
0
1.8
9
x 103
x 10" ?
-------
TABLE A.21. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR CARBAZOLE:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient
Pseudo-first-order rate constants
Volatilization (hr"1)
Photolysis (hr-1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
200
0
-2
4.6 x 10
2.9 x 10~3
0
2.3 x 10~2
200
0
0
0
0
2.3 x 10
-3
353
-------
TABLE A.22. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR CARBAZOLE:
RIVER SIMULATION
00
Ul
Water compartments
Sorption partition coefficient
Pseudo-first-order rate
constants
Volatilization (hr-1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr'1)
1
200
0
1.1 x 10'1
2.9 x 10~3
0
2.3 x 10~2
2
200
0
1.1 x 10~!
2.9 x 10"3
0
2.3 x 10"2
3
200
0
1.1 x 10'1
2.9 x 10~3
0
2.3 x 10~2
Sediment compartments
7
200
0
0
0
0
2.3 x 10~3
8
200
0
0
0
0
2.3 x 10~3
9
200
0
0
0
0
2.3 x 10~3
-------
TABLE A.23. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR CARBAZOLE:
EUTROPHIC LAKE SIMULATION
Water compartments
Sediment compartments
8
u>
tn
ui
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr-1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
200
200
200
200
200
4.6 x 10'
.-2
2.9 x 10
0
,"3
9.2 x 10'
2.9 x 10'
0
,-2
-3
4.6 x 10'
2.9 x 10
0
-2
-3
4.6 x 10~3 0
2.9 x 10~3 0
200
0
0
0
0
200
0
0
0
0
2.3 x 10~2 2.3 x 10~2 2.3 x 10~2 1.6 x 10~2 2.3 x 10~3 2.3 x 10~3 2.3 x 10~3
-------
TABLE A.24. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR CARBAZOLE:
OLIGOTROPHIC LAKE SIMULATION
ui
Water compartments Sediment compartments
12357
Sorption partition
coefficient 200 200 200 200 200
Pseudo-first-order rate
constants
Volatilization (hr""1) 000 00
Photolysis (hr"1) 4.6 x 10~2 9.2 x 10~2 4.6 x 10"2 4.6 x 10~3 0
Oxidation (hr-1) 2.9 x 10~3 2.9 x 10~3 2.9 x 10"3 2.9 x 10~3 0
Hydrolysis (hr'1) 000 00
8
9
200 200
0 0
0 0
0 0
0 0
Biodegradation (hr"1)
2.3 x 10~6 2.3 x 10~6 2.3 x 10"6 1.6 x 10"6 2.3 x 10"7 2.3 x 10~7 2.3 x 10~7
-------
TABLE A.25. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR 7H-DIBENZO[c,g]CARBAZOLE.
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient 0 0
Pseudo-first-order rate constants
Volatilization (hr-1) j 0 0
r
Photolysis (hr-1) | 0.462 0
r
Oxidation (hr"1) 0.001 0.001
Hydrolysis (hr-1) 0 0
Biodegradation (hr-1) 0 0
357
-------
TABLE A.26. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR 7H-DIBENZO[c,g]CARBAZOLE:
RIVER SIMULATION
Ul
00
Water compartments
Sorption partition coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr-1)
Biodegradation (hr-1)
1
0
0
0.693
0.001
0
0
2
0
0
0.693
0.001
0
0
3
0
0
0.693
0.001
0
0
Sediment compartments
7
0
0
0
0.001
0
0
8
0
0
0
0.001
0
0
9
0
0
0
0.001
0
0
-------
to
in
. TABLE A.27. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR 7H-DIBENZO[c,g]CARBAZOLE:
EUTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr"1)
Oxidation (hr-1)
Hydrolysis (hr"1)
Biodegradation (hr~:)
1
0
0
0.462
0.001
0
0
2
0
0
0.462
0.001
0
0
3
0
0
0.462
0.001
0
0
5
0
0
0.046
0.001
0
0
Sediment compartments
7
— -
0
0
0
0.001
0
0
8
0
0
0
0.001
0
0
9
0
0
0
0.001
0
0
-------
Ul
CTI
o
TABLE A.28. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR 7H-DIBENZO[c,g]CARBAZOLE:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr""1)
Oxidation (hr"1)
Hydrolysis (hr'1)
Biodegradation (hr"1)
1
0
0
Io4
0.001
0
0
2
0
0
1.4
0.001
0
0
3
0
0
1.4
0.001
0
0
5
0
0
1.4
0.001
0
0
Sediment compartments
7
0
0
0
0.001
0
0
8
0
0
0
0.001
0
0
9
0
0
0
0.001
0
0
-------
TABLE A.29.
NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[b]THIOPHENE;
POND SIMULATION
Water
(compartment 1)
Sediment
(compartment 7)
Sorption partition coefficient
Pseudo-first-order rate constants
Volatilization (hr-1)
Photolysis (hr-1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
50
3 x 10~3
2.4 x lO"4
0
0
3.5 x 10~2
50
0
0
0
0
1.7 x 10
-2
361
-------
W
CT*
to
TABLE A.30. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[b]THIOPHENE:
RIVER SIMULATION
Sorption partition coefficient
Pseudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr-1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Water
1
50 50
1.5 x 10~2 1.5
6.0 x lO"4 6.0
0 0
0 0
compartments
2 3
50
x 10~2 1.5 x 10~2
x 10~4 6.0 x 10"1*
0
0
7
50
0
0
0
0
Sediment
50
0
0
0
0
compartments
8 9
50
0
0
0
0
Biodegradation (hr-1) 3.5 x 10 3.5 x 10"2 3.5 x 10~2 8.5 x 10~3 8.5 x 10~3 8.5 x 10~3
-------
TABLE A.31. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[b]THIOPHENE:
EUTROPHIC LAKE SIMULATION
Water compartments Sediment compartments
23578
Sorption partition
coefficient 50 50 50 50 50 50 50
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1) 3.5 x 10~2 3.5 x 10~2 3.5 x 10~2 1.3 x 10"2 4.0 x 10~3 4.0 x 10~3 4.0 x 10~3
3.8 x 10
2.4 x 10""*
0
0
4.5 x 10
2.4 x 10"1*
0
0
3.8 x 10
2.4 x 10""*
0
0
0
2.4 x 10~5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
-------
U)
en
TABLE A.32. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR BENZO[b]THIOPHENE:
OLIGOTROPHIC LAKE SIMULATION
Water compartments Sediment compartments
12357
Sorption partition
coefficient 50 50 50 50 50
Pseudo-first- order rate
constants
Volatilization (hr"1) 3.8 x 10~3 4.5 x 10"3 3.8 x 10~3 0 0
Photolysis (hr"1) 1.2 x 10~3 1.2 x 10"3 1.2 x 10"3 1.2 x 10"1* 0
Oxidation (hr"1) 000 00
Hydrolysis (hr"1) 000 00
8 9
50 50
0 0
0 0
0 0
0 0
Biodegradation (hr"1)
3.5 x 10~6 3.5 x ID"6 3.5 x 10"6
1.3 x lO"6 4.0 x lO"7 4.0xlO"7 4.0 x 10"7
-------
TABLE A.33. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR DIBENZOTHIOPHENE:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient
Pseudo-first-order rate constants
Volatilization (hr"1) j|
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
1.4 x 103
9.6 x 10
7.3 x 10'
0
0
5.3 x 10"
-3
_2
1.4 x 103
0
0
0
0
5.3 x 10
-2
365
-------
en
TABLE A.34. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR DIBENZOTHIOPHENE:
RIVER SIMULATION
1
Water
compartments
2
3
7
Sediment
compartments
8
9
Sorption partition coefficient 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103
Pseudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr-1)
Oxidation (hr-1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
4.8 x 10~2 4.8 x 10~2 4.8 x 10"2
108 x 10~3 1.8 x 10~3 1«8 x 10~3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5.3 x 10~2 5.3 x 10~2 5.3 x 10"2
5.3 x 10~2 5.3 x 10~2 5.3 x 10~2
-------
TABLE A.35. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR DIBENZOTHIOPHENE:
EUTROPHIC LAKE SIMULATION
Water compartments
Sediment compartments
8
u>
o>
-j
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103 1.4 x 103
1.2 x 10
7.3 x 10""1*
0
0
1
7
0
0
.4
.3
x 10 '
x Hf1*
1.
7.
0
0
2 x 10
3 x 10"1*
0
7.3 x 10~5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5.3 x 10~2 5.3 x 10~2 5.3 x 10~2 3.5 x 10~2 2.6 x 10~2 2.6 x 10~2 2.6 x 10~2
-------
TABLE A.36. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR DIBENZOTHIOPHENE:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sediment compartments
8
00
-------
TABLE A.37. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR METHYL PARATHION:
POND SIMULATION
Water
(compartment 1)
Sediment
(compartment 7)
Sorption partition coefficient
Pseudo-first-order rate constants
Volatilization (hr"1)
Photolysis (hr-1)
Oxidation (hr-1)
Hydrolysis (hr"1)
50
0
8 x 1Q-1*
0
4 x IQ-1*
50
0
0
0
4 x
Kinetic constants for biodegradation
Bacterial population
(cells ml-1)
Active bacterial population
(cells ml"1)
Maximum growth rate (hr"1)
Half-saturation constant
(yg ml-1)
Yield factor (cells yg"1)
106
105
0.61
2.6
106
107
105
0.2
5.2
106
369
-------
TABLE A.38. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR METHYL PARATHION:
RIVER SIMULATION
Water compartments
Sorption partition coefficient
Pseudo-first-order rate
constants
Volatilization (hr'1)
Photolysis (hr"1)
Oxidation (hr-1)
Hydrolysis (hr-1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-1)
Active bacterial population
(cells ml'1)
Maximum growth rate (hr"1)
Half-saturation constant
l ..
1
50
0
2 x 10-3
0
4 x ID'1*
10 6
IO5
0.61
2.6
2
50
0
2 x 10~3
0
4 x 10"1*
10 6
IO5
0.61
2.6
3
50
0
2 x ID"3
0
4 x 10-1*
IO6
10*
0.61
2.6
Sediment compartments
7
50
0
0
0
4 x 10-1*
IO7
10s
0.2
5.2
8
50
0
0
0
4 x 10"1*
IO7
IO5
0.2
5.2
9
50
0
0
0
4 x 10-1*
IO7
IO5
0.2
5.2
(Ug ml-1)
Yield factor (cells pg"
106
io6
106
IO6
10 6
IO6
-------
to
TABLE A.39. NINE-COMPARTMENT COMPUTER MODEL INPUTS
' FOR METHYL PARATHION:
EUTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr-1)
Photolysis (hr'1)
Oxidation (hr"1)
Hydrolysis (hr~l)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-1)
Active bacterial popula-
tion (cells ml'1)
Maximum growth rate
(hr-1)
Half-saturation constant
(Hg ml'1)
Yield factor (cells pg'1)
1
50
0
8 x 10~4
0
4 x 10~4
106
10s
0.61
2.6
106
2
50
0
1 x ID"3
0
4 x 10~4
106
105
0.61
2.6
106
3
50
0
8 x ID"4
0
4 x ID"4
106
105
0.61
2.6
106
5
50
0
8 x 10~5
0
4 x 10'4
10 6
105
0.4
0.4
106
Sediment compartments
7
50
— rr~rTT
0
0
0
4 x ID'4
107
105
0.2
0.52
106
8
50
0
0
0
4 x 10~4
107
105
0.2
0.52
106
9
50
0
0
0
4 x 10-4
107
105
0.2
0.52
106
-------
^J
to
TABLE A.40. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR METHYL PARATHION:
OLIGOTROPHIC LAKE SIMULATION
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr-1)
Oxidation (hr-1)'
Hydrolysis (hr-1)
Kinetic constants for
biodegradation
Bacterial population
(cells ml-*)
Active bacterial popu-
lation (cells ml-*)
Maximum growth rate
(hr-1)
Half-saturation constant
(Hg ml-1)
Yield factor (cells yg-1)
1
50
0
4 x 10~3
0
4 x lO'1*
102
10
0.61
0.26
106
2
50
0
5 x 10"3
0
4 x ID""
102
10
0.61
0.26
106
3
50
0
4 x ID"3
0
4 x 1Q-1*
102
10
0.61
0.26
10 6
5
50
0
4 x 10"1*
0
4 x 1Q-1*
102
10
0.4
0.4
106
Sediment compartments
7
50
0
0
0
4 x 10'1*
103
10
0.2
0.52
106
8
50
0
0
0
4 x 10~**
103
10
0.2
0.52
10 6
9
50
0
0
0
4 x 10"1*
103
10
0.2
0.52
106
-------
TABLE A.41. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR MIREX:
POND SIMULATION
Water Sediment
(compartment 1) (compartment 7)
Sorption partition coefficient
Pseudo-first-order rate constants
Volatilization (hr"1)
Photolysis (hr-1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
3 x 105
1 x 10"11
0
0
0
0
3 x 105
0
0
0
0
0
373
-------
TABLE A.42. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR MIREX:
RIVER SIMULATION
Water compartments
Sorption partition coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr-1)
Hydrolysis (hr-1)
Biodegradation (hr~l)
1
3 x 105
1.4 x 10~3
0
0
0
0
2
3 x 105
1.4 x 10~3
0
0
0
0
3
3 x 105
1.4 x 10~3
0
0
0
0
Sediment compartments
7
3 x 105
0
0
0
0
0
8
3 x 105
0
0
0
0
0
3
0
0
0
0
0
9
x 105
-------
TABLE A.43. NINE-COMPARTMENT COMPUTER MODEL INPUTS
FOR MIREX: EUTROPHIC AND
OLIGOTROPHIC LAKE SIMULATIONS
Ol
Water compartments
Sorption partition
coefficient
Pseudo-first-order rate
constants
Volatilization (hr"1)
Photolysis (hr"1)
Oxidation (hr"1)
Hydrolysis (hr"1)
Biodegradation (hr"1)
1
3 x 105
3.5 x 10" "
0
0
0
0
2
3 x 105
3.5 x 10"1*
0
0
0
0
3
3 x 105
3.5 x Kf1*
0
0
0
0
5
•=•
3 x 105
0
0
0
0
0
Sediment compartments
7
3 x 105
0
0
0
0
0
8
3 x 105
0
0
0
0
0
3
0
0
0
0
0
9
x 105
-------
Appendix B
EXPERIMENTAL PROCEDURES
The first four sections of this appendix supplement the descriptions of
the general laboratory procedures that were given in Part I of this report.
The remaining sections give the specific experimental conditions for each
chemical studied. These sections are conveniently numbered to coincide with the
section number for that chemical in the body of the report. For example, Section
B.4 gives the specific experimental conditions for p-cresol, which is discussed
in Section 4 of Part II.
B.I PHYSICAL TRANSPORT
B.I.I Solubility
The general procedure described by Campbell (1930) was used to measure
the solubilities of solids in the yg ml"-'- range. A small amount of the solid
substrate is placed in an all-glass apparatus, which is immersed in a water bath
and shaken gently. This apparatus has two compartments separated by a glass
frit. When the flask is inverted, the aqueous solution is filtered through the
frit to remove solid substrate. The filtration step can be carried out without
removing the apparatus from the water bath. After equilibration in the water
for several days, the sample is filtered in the apparatus and the filtrate is
analyzed for the substrate.
Generally, at least three measurements are made. Also, measurements are
made on samples that have been heated to 35° to 40°C, allowed to equilibrate,
and then cooled in the water bath. Since the concentrations are fairly high,
potential problems that are encountered with low-solubility materials, such as
adsorption onto the frit during filtration and the possibility that finely di-
vided particulate substrate is not removed during filtration, are not likely to
be significant.
Solutions of compounds having a solubility in the ng ml~^ range were
prepared by the procedure described by Haque and Schmedding (1975). The sub-
strate is dissolved in an organic solvent and put on the walls of a 5-gallon
(18.8 liter) carboy. The carboy is rotated slowly on its side while the solvent
evaporates, so that a thin film of substrate coats the wall of the carboy. A
large Teflon-coated magnetic stirring bar is added, and the carboy is filled
with the purest water available. Care must be taken to prevent the substrate
from coating the bottom of the carboy so that it is not dislodged by the stir-
ring bar. The solution is allowed to stir gently for at least a week to assure
that equilibration has taken place. Samples of water are withdrawn with a
glass siphon and analyzed for the substrate.
In several cases, we found that even with these precautions particulate
matter, presumably substrate, could be observed in the carboy, and the substrate
376
-------
concentration was reduced by centrifugation at 10,000 rpm. The solubility
measurements of low-solubility (less than 0.1 ug ndr1) compounds should be made
on centrifuged samples.
B.I.2 Absorption Spectra
Pure water was used as the solvent for measurement of absorption spectra
whenever possible. However, it was often necessary to increase the solubility
by adding a water-soluble cosolvent, such as acetonitrile. The general
procedure was to prepare a substrate solution of a minimum of about 10~5 M in
water, using acetonitrile to dissolve any solid, undissolved substrate. Some
trial and error was required to minimize the amount of acetonitrile. The ab-
sorption spectra were measured in 1-cm and 10-cm cells on a Gary model 14
spectrophotometer. To obtain satisfactory spectra with 10-cm cells, we always
ran solvent versus solvent tjo obtain a baseline. Then, the cell containing sol-
vent in the sample beam was refilled with the substrate solution, and the ab-
sorption spectrum was measured. Suitable standards were prepared to assure that
the same absorbance was obtained in the 10-cm cell with a solution that was
one-tenth the concentration used in the 1-cm cell.
The molar extinction coefficients e, are obtained from Beer's law.
Absorbance = - log — = e. &S (B-l)
J. A
o
where IQ is the incident light flux and I is the transmitted light flux in the
spectrophotometer. If £, the cell path length, is in centimeters and if S, the
substrate concentrations, is in M, then the molar extinction coefficient e, is
in units of cm~l M"-*-. The average molar extinction coefficient for each
nominal wavelength is calculated from the average of the molar extinction co-
efficients at the lower and upper limits of the wavelength interval (Table B.I).
TABLE B.I. NOMINAL WAVELENGTHS AND WAVELENGTH INTERVALS FOR UV AND VISIBLE
ABSORPTION SPECTRA
Nominal Wavelength
wavelength interval
(nm) (nm)
297.5 ± 1.25
300.0 to 320.0 ± 1.25
323.1 ± 1.9, - 1.85
330.0 ± 5.0
340.0 and higher ±5.0
*The actual precision of a measured wavelength is about ± 0.5 nm.
377
-------
B.I. 3 Volatilization Rates
Volatilization rates were measured using the method described by Hill
et al. (1976). A solution of the substrate in pure water was prepared at a
concentration that is below its saturation value. About 1 liter of this solu-
tion was placed in a 2-liter beaker equipped with a stirring bar. The solution
was purged with nitrogen to remove most of the dissolved 02- At the start of
the experiment (t = 0), the concentration of substrate was measured and the 02
concentration was measured with a Delta model 2010 02-analyzer. Successive
substrate and 02 measurements were made at regular time intervals.
The substrate concentration versus time data were then fit to an expo-
nential decay curve of the form
-kSt
[St] = [SQ]e V (B.2)
which is the integrated form of the first-order rate expression
-=K] (B.3)
The oxygen concentration data are fit to the integrated form of equation (B.4)
- k° ([02]gat - [02]t) (B.4)
dt
which is
-k°t
= [°2]a ~ [°2] ~ ^ G (B'5)
'sat
where [02]sat is the saturation concentration of 02 in water at the temperature
of the measurement and is a constant because the concentration of 02 in the air
is constant.
The linear least squares routine supplied with the Hewlett-Packard Model
65 calculator was used to calculate values of k^ and k^. This program gives a
linear least squares fit to InS versus t, plus the variance of the parameter
estimates. The value of k^/kv was calculated from these values.
B.I.4 Sorptiqn
Preparation of Ca-Montmorillonite and Natural Sediments—The montmoril-
lonite clay used in these studies was a Wyoming montmorillonite obtained from
Dr. William John, Department of Geology, University of Missouri, Columbia,
Missouri. Clay suspensions (about 1% by weight) were prepared by soaking a
measured weight of clay in distilled-deionized water for at least one week.
The clay suspension was then passed through an ion exchange column that had
been presaturated with calcium ions to convert the clay from the sodium form
to the calcium form. This was necessary because the Na-montmorillonite sus-
pension could not be centrifugedbut the Ca-montmorillonite could. The parti-
cle size was less than 1 urn.
378
-------
The procedures used to prepare and store the natural sediments were
designed to preserve the sediments in their natural state as well as possible.
Natural sediments were screened to remove large rocks, twigs, and other debris.
The mesh sizes of the screens used were 4, 16, and 28 per 2.54 cm. Following
the screening, the sediment was mixed, using a Humbolt splitter to be sure the
sediment was uniform. A small volume of each sediment was mixed with two volumes
of 0.1 M calcium chloride and the pH was recorded. The remainder of the screened
and split sediment was stored in 100-ml Nalgene bottles at 4°C until use. The
sediments were never allowed to dry out, and no attempts were made to kill or
remove the bacteria in the sediments. Before use, the sediment was resuspended
in water and allowed to settle for 30 seconds. The sediment was then trans-
ferred as a slurry using a pipet.
The characteristics of the Ca-montmorillonite clay and natural sediments
used in these studies are gijven in Table B.2. The organic carbon (OC) values,
expressed as percent carbon by weight, were determined using the Walkley and
Black procedure, which involves oxidation of the organic material by chromate
followed by back-titration with ferrous ammonium sulfate (Hesse, 1971).
TABLE B.2. SOURCES AND CHARACTERISTICS OF
Ca-MONTMORILLONITE CLAY AND NATURAL SEDIMENTS
Source
Sediment OCC
Location and description
Cation
exchange
capacity
(meq/100 g)
Wyoming
Navarro River
Des Moines
River
Oconee River
Coyote Creek
Searsville Pond
Ca-montmorillonite clay
Mendocino County, California
An unpolluted river that
drains redwood forests,
orchards, and pasture
Iowa
Georgia
San Jose, California
A eutrophic, polluted
stream
Woodside, California
A small eutrophic but
unpolluted pond
6.7
7.1
0.05
0.5
0.8
6.2 0.8
6.5 1.9
6.7 5.0
69.0
4.5
10.5
8.5
13.5
34.5
Organic carbon, Walkley and Black values, corrected for recovery by multiplying
experimental value by 1.33.
379
-------
Sorption isotherm measurements—In most cases, the clay and sediment
isotherm measurements were made at two sediment loadings and two substrate con-
centrations. Replicate flasks were used at each level, and at least three
analyses of each flask were made. Suitable blanks of both sorbent and sub-
strate were carried through the experimental steps and analyzed. Contact times
of 1 to 16 hours were used. There were no experimental problems in the sedi-
ment sorption studies with the 16-hour time, except for j^-cresol, which biode-
graded rapidly during the experiment. In that case, a 1-hour exposure was used.
For the biosorption studies, the partitioning time was about 1 hour. At longer
times, sorption by the glassware was a problem with low-solubility substrates.
As the experiments progressed, the experimental plan for the isotherm
measurements evolved into the experimental plan described in Table B.3. We
consider this to be the minimum number of data points that will permit a sound
statistical analysis of the data.
TABLE B.3. RECOMMENDED EXPERIMENTAL PLAN FOR ISOTHERM MEASUREMENTS
Number of flasks3
Substrate
concentration
None
Low
High
No
sediment
1
2
2
Low
sediment
2
2
2
High
sediment
2
2
2
£3
Four replicate measurements of the substrate concentration at
equilibrium in each flask should be made.
The mechanics of the isotherm measurements were generally the same
whether the sorbent was Ca-montmorillonite clay, a natural sediment, or our
bacterial mixture. A solution* of the substrate in water was prepared. An
aliquot of the substrate solution and an aliquot of a suspension of the sorbent
were mixed and allowed to shake for a specified period of time. The mass of
dry sorbent used was determined, normally by a gravimetric procedure, in a
separate experiment. A portion of the mixture was centrifuged to separate the
substrate remaining in solution and the sorbent. The supernatant and usually
the sorbent were analyzed separately to measure the substrate concentration.
Statistical Analysis of Isotherm Data—The data were fit to the
Freundiich isotherm equation with n = 1,
S = K S (B.6)
s p w
The importance of using true solutions below the substrate solubility limit
cannot be overemphasized, especially with low-solubility substrates.
380
-------
The value of Sw was always measured. If Ss was not measured independently,
then it was claculated from
S = (S - S )V /m
s o wy w s
(B.7)
where SQ is the initial concentration of substrate used (determined from the
concentration in flasks without sorbent), Vw is the volume of solution (in ml),
and ms is the mass of sediment (in grams) added to the flask. Preliminary
estimates of Kp were obtained by two methods: when only Sg was measured, every
concentration measurement was used, and when both Ss and Sw were measured,
average Ss and Sw were calculated for each flask. In both cases the data were
then fit to an equation of the form
y =
(B.8)
(notice the similarity to equation B.6, using a linear least squares regression
method. The regression equations are:
b = —
~x.y.
yx
ra =
Ex.2
Ex..2
(n - I)'
-1
(E.9)
(B.10)
(B.ll)
95% confidence interval = (tn_ija ) Syx(Zxi / (B.12)
where b = Kp is the slope, x. and y^ are the individual or average measurements
from each flask of Sw and S^ respectively, Syx is the standard error, r2 is the
correlation coefficient, and tn_i>a is the t-value from Student's t-test for n
measurements at a - 0.05 confidence. These expressions were programed and used
on an Hewlett-Packard model 65 calculator.
To test that the linear Freundlich isotherm (n = 1, equation B.6) does pass
through the origin, the data were fit to a linear equation of the form
S = K S + a
s p w o
(B.13)
where ao is the intercept. The HP-65 Stat-Pac routine was used to estimate the
95% confidence intervals about a0. If these confidence intervals include the
origin, then equation (B.6) was used.
381
-------
The limitations of the linear method can be removed by stating the
problem as:
When sorbent is present
S. - a
(B.14)
s p
When sorbent is not present
S = S. ; i = h or £ (B.15)
w i
The hat on a variable indicates that its value is estimated by the regression
procedure. §. is equivalent to SQ and represents the original amount of sub-
strate present in each flask, §^ being the concentration in flasks with the high
amount of substrate and S^ being the concentration in flasks with the low amount
of substrate. Ms^represents the amount (grams) of sorbent present. In this
procedure, S, or S. was estimated using all the flasks as in the linear method.
Since each flask was handled separately and at a different time, the
above procedure was modified to include "flask effects" as an estimated param-
eter. Flask effects include such things as biases due to instrument drift and
systematic errors on the part of the analyst. The resulting problem formulation
is suitable for input to a nonlinear regression program that estimates values
for S, , S., a , K^, and the flask effects. The actual results of the nonlinear
regressions are comparable to the estimates obtained from the linear least squares
regressions, except that the nonlinear approach gives smaller confidence limits
for the parameter estimates.
The Biomedical Statistical Series nonlinear regression program BMD07R
performed the regression indicated above by equations (B.14) and (B.15). BMD07R
executes a user-supplied FORTRAN subroutine to compute the estimate of the
dependent variable and its first derivatives with respect to the regression
parameters for each observation. The data available to the subroutine consisted
of S and Mg/Vw augmented by dichotomous dummy variables indicating the presence
or absence of sorbent, the amount (high or low) of substrate, and the flask
measured (one of two flasks). These dummy variables controlled the computation
of the dependent variable estimate for each observation by one of the two
regression equations (B.14) or (B.15).
When both Sg and Sw were measured, the linear least squares procedures
using average values for Ss and Sw are not statistically correct because again
response variables appear on both sides of the regression equation. In addition,
the averaging procedure throws away valuable information about experimental
variance. To deal more correctly with this situation, eight simultaneous non-
linear regression equations were used. Four regressions use only the substrate
concentrations measured on the sediment from the various flasks.
S = K S. i = 1, 2, 3, 4 (B.16)
s p i
and four regressions use only the concentrations measured in the supernatant, Sw,
382
-------
s = S 1=1, 2, 3, 4 (B.17)
w
where S..^ is the estimated value of substrate concentration in a particular flask.
With this formulation, the response variables appear only on the left-hand side
of the regression equation. The subscript "i" on the right-hand side is the
independent variable, since it indicates the conditions used to set up the flask.
For example, i = 1 indicates the two duplicate flasks that contained high sedi-
ment and high substrate concentrations. The common parameters K- and S^ tie
the simultaneous regressions together and assure that the resulting estimate
for Kp is conditioned on both the sediment and supernatant concentrations
measured.
The method is superficially similar to the simple linear least squares
procedure used to estimate Kp. The regression equations (B.17) use the super-
natant concentrations to estimate a single concentration that best represents
the concentration in the fljasks for each different substrate and sediment level.
This representative concentration is then used with the concentration of sub-
strate on the sediment in equations (B.16) to estimate 1C. An important dif-
ference between the two approaches, however, is that, with the nonlinear approach
the supernatant concentration, S^, that represents a particular substrate and
sediment level is not necessarily the average supernatant concentration. The
values of S^ determined by the method are almost always very close to the
average except when concentration measurements are highly scattered.
BMD07R also performed the regression indicated by equations (B.16) and
(B.17) above. In this case the data available to the dependent variable esti-
mation subroutine was augmented by dichotomous dummy variables indicating
substrate or supernate measurement of S^, sediment levels (low or high), and
a flask indicator (one of two flasks) . These variables controlled the regres-
sion equation (B.16) or (B.17) used to estimate the dependent variable for each
observation.
An important feature of nonlinear approach is that it does not require
that the same number of observations of both Sg and Sw be made in each flask.
The formulation also does not require that individual measured values of Sg
and,Sw from a particular flask be paired (see equation B.6). This is an
important feature because any pairing of data points is artificial and could
bias the estimate of K^.
Biosorption and Desorption — Biosorption and desorption studies were con-
ducted with mixtures of four species of gram-positive and gram-negative aquatic-
origin bacteria. The mixtures contained equal optical densities of Azotobacter
beijerinckii ATCC 19366, Bacillus cereus ATC 11778, Escherichia coli ATC 9637,
and Serratia marcescens ATCC 13880. In the early stages of this program,
Flavobacterium capsulatum ATCC 14666 was used, but because this organism was
difficult to centrifuge to a compact pellet and clear supernatant, it was re-
placed with the above indicated Serratia marcescens.
383
-------
The test organisms were transferred several times in Trypticase-Soy broth
at 25°C before they were used for sorption studies. Sixteen-hour cultures were
either in the late logarithmic or early stationary growth phases. At this stage,
each culture was harvested by centrifuging, washed with 0.05% potassium phos-
phate buffer (pH 7.0), resuspended and diluted with this buffer until the
suspension had an optical density of 2 to 4.
Appropriate aliquots of suspensions of each of the four organisms were
combined and diluted with buffer to form a mixture containing equal optical
densities of each organism and a mixture that, when mixed with the solution of
the substrate, resulted in the desired concentration and organisms. With sub-
strates having a low solubility, the density of the bacterial mixture was lower
than with more soluble substrates.
In some instances, biosorption studies were also conducted with heat-
killed cells. Consequently an aliquot of the above mixture of organisms was
heated at 100°C for 15 minutes, cooled, and centrifuged. The resulting pellet
was resuspended in fresh buffer to the original volume, and an appropriate
volume of this suspension was diluted as above with a solution of the substrate
under study to yield the corresponding cell densities and substrate concentrations.
Biosorption studies were conducted by incubating the viable and heat-
killed cell mixtures in Corex centrifuge tubes or bottles for 1 hour at 25°C.
Cells were maintained in suspension by placing the containers in roller drums
or on a rotary shaker. The tubes or bottles were centrifuged for 10 minutes at
12,000 or 16,000 G, respectively, and the supernatants were carefully decanted.
The supernatants were extracted with an organic solvent (usually ethyl acetate).
The solvent extract was dried and then assayed directly or concentrated before
assay. To assay sorbed substrate in the pellets, water was added and this
suspension was solvent extracted as above. With some substrates that were
tenaciously retained by the cell pellets, the water-suspended cells in the
presence of some solvent were slowly frozen and thawed three times before
extraction.
Desorptions were conducted only if the sorption partition coefficients
were 10,000 or more. The cell pellets from replicate sorption studies were
suspended in volumes of buffer or buffer and solvent equivalent to those used
for sorptions, incubated with shaking at 25°C for 3 hours, and centrifuged.
Both supernatants and pellets were analyzed by the procedures used for sorp-
tion determinations.
In some cases, corrections were made for sorption on glassware of sub-
strate and cultures containing sorbed substrate. Separate controls consisted
of extraction of tubes from which the incubated suspensions in test substrate
solutions were decanted in lieu of separation of cells by centrifuging.
Dry weights of cells used in sorption studies were determined by
weighing the pellet obtained after cells from aliquots of mixed viable or
heat-killed bacterial suspensions were centrifuged, washed with distilled
water, and dried for 16 hours at 90-95°C.
384
-------
The biosorption partition coefficients of chemicals between bacteria
and buffer were determined as
K = pg substrate per g dry wt of cells (B.18)
p yg substrate per ml in supernatant
Purification of Humic Acid—The humic acid (Fluka AG, Buchs SG, distrib-
buted by Tridom Chemical, Inc.) was dissolved in O.l.NNaOH and centrifuged 45
minutes at 10,000 rpm. The supernatant was decanted off; the solids were re-
suspended in O.lNNaOH and centrifuged again for 45 minutes at 10,000 rpm. The
supernatant was added to the previous batch and the combined solution filtered
twice through glass fiber filters, then through a 0.45-ym Millipore HA filter at
40 psig until a flow rate of 1 ml min~l was obtained. The filtrate was returned
to the filter and refiltered at 350 psig. The filtrate was then placed in a
presoaked 0.98-inch-diameter dialysis tube (Union Carbide) and put into running
deionized water for 24 hours. The solution was then passed through an OH-
saturated amberlite IRA-400 exchange column, then through an H-saturated amber-
lite IR-120 exchange column to neutralize the humic acid. This solution was
stored in a refrigerator and diluted just prior to use in the photochemical
studies (Section B.2.2 of this Appendix).
B.2 CHEMICAL TRANSFORMATION
B.2.1 Preparation of Reaction Solutions
Solutions of chemicals were prepared in pure water and natural waters.
All glassware used was routinely baked out overnight at 580°C. Pure water was
obtained from a Millipore water purification system. The first stage of this
system purifies tap water by reverse osmosis (Milli-RO-4); the second stage
uses a Milli-Q system in which the water is passed through a carbon filter, two
ion exchange filters, and finally a 0.22-um cellulose ester filter. The filter-
sterilized water obtained typically shows greater than 18 Mohm-cm resistivity
and less than 1 ug ml"1 total organic carbon (below detectable limit of instru-
ment).
Natural waters were obtained from Coyote Creek, Searsville Pond, and
Lake Tahoe^. These waters were filtered through 0.22-um filters before prep-
aration of the solutions to sterilize the waters and to remove any heterogeneous
components from solution. The latter was a precaution to ensure that no
adventitious processes due to heterogeneity occurred in the solution kinetics
processes being measured. Some water quality parameters for these filtered
waters are given in Table B.4.
Reaction solutions were usually prepared from a stock solution of the
chemical in acetonitrile. An aliquot of this stock solution was diluted with
an amount of water to give a 1% or 0.1% acetonitrile in water solution.
Preparation of the stock solution and subsequent dilutions were made to obtain
1 yg ml-1 or less concentration of chemicals. As discussed in Part I of this
report, 1% acetonitrile does not have any significant effect on the kinetic
processes under study; the presence of 1% acetonitrile was sufficient to
385
-------
produce usable concentrations of chemicals for our kinetics experiments. When
the solubility of the chemical in water was high enough, water was used as the
sole solvent. When the solubility of a chemical in a solution was uncertain,
the absence of light scattering from a beam of green light or a CC>2 laser passed
through the solution was interpreted as indicating a homogeneous solution (i.e.,
free of particulates).
TABLE B.4. SOME WATER QUALITY PARAMETERS3 FOR 0.22-ym FILTERED WATER
FROM SEARSVILLE POND, COYOTE CREEK, AND LAKE TAHOE
Searsville Lake
Coyote Creek
South Lake Tahoe
Total
organic
carbon
(mg ml"1)
10
7
2
Total
inorganic
carbon
(mg ml-1)
< 10
< 7
< 2
Iron
(mg ml"1)
0.010
0.030
0.020
Copper
(mg ml-1)
0.024
0.014
< 0.005
Total
Hardness
(mg ml-1)
700
280
28
o
Analyses were made by LFE Environmental Analysis Laboratories, Richmond, CA.
Except for the experiments with quinoline and _p—cresol, all analyses were
performed by high pressure liquid chromatography (HPLC) on a Waters Associates
chromatograph apparatus (Model 6000A pump, MV6K injector, and a M440 absorbance
detector). Separations were made with a 30 cm x 4 mm pBondapak C^g column
(reverse phase). The solvent eluent composition, UV detector wavelength, and
internal standard, if any, are given below in the sections for the specific
chemical. Peak areas in the HPLC traces were determined on a SpectraPhysics
Autolab Minigrator.
Data for concentration as a function of time were fit to a first-order
kinetic rate law using a computer least squares program. The first-order kinetic
rate constants and the standard deviations are presented in the respective
sections on the chemicals in the body of the report.
B.2.2 Photochemical Studies
Laboratory Photolyses—Reaction mixtures of the chemical (4 ml) were
placed in 10-mm-O.D. borosilicate tubes (Pyrex 7740) and photolyzed on a
merry-go-round reactor (Ace Glass). The irradiation source was a Hanovia 450-
watt, medium-pressure Hg lamp contained in a borosilicate immersion well. The
distance between the irradiation source and tubes was about 10 cm. The reaction
temperature was at the ambient operating temperature of the system (^28°C).
The photolyses were carried out using several different filter systems,
which were placed between the Hg lamp and reaction mixtures. In all laboratory
photolyses, the borosilicate glass immersion well (8-mm total glass path length)
served as a filter to screen out all light below 290 nm. The filter systems
used in the photolyses are described below.
386
-------
313 nm filter system: Corning CS 7-54 glass filter with a 0.001 M
potassium chromate solution in 3% aqueous sodium carbonate circulated
in the immersion well. This system transmitted primarily the
313.2 and 312.6 nm Hg lines, which represented greater than 95%
of the light incident on the reaction solutions (a small Hg line
at 302.2 nm was also present).*
366 nm filter system: Corning glass CS 0-52 and CS 7-60 glass
filters. This pair of glass filters transmitted only the 365.1,
365.6, and 366.4 nm lines of the Hg lamp, with no other lines
observed (less than 1% of light outside the 366 nm band).*
Borosilicate glass only : This system used only the immersion
well as filter and was employed only for £-cresol and benzo[b]-
thiophene experiments I Examination of the transmitted light
showed that, in the' region where these compounds absorb (less
than 330 nm) , the same two Hg lines were present as described
above for the 313 nm system, but the filters did affect the
relative intensities of the lines.
Sample tubes were removed from the merry-go-round reactor at appropriate
times and immediately analyzed.
The quantum yield () was calculated from the photolysis rate constant
k , obtained from the slope of the first-order plot of photolysis data using
the following equation:
k = 2.303 e, I,£ = d(ln C /C)/dt (B.19)
p A A O
where e is the absorption coefficient of chemical at wavelength A, & is the
pathlength, and 1^ is measured as above. The value of I, is constant for the
system since the same tubes were used in both the actinometry and in the chemi-
cal photolyses. Any differences produced by reflection effects in using the
tubes should be negligible.
The photochemical system was calibrated at 313 nm and at 366 nm wavelengths
using the TD-nitrobenzaldehyde actinometer system (Pitts et al., 1968; Calvert
and Pitts, 1966). In these calibration experiments, a solution of approximately
1 x lO"^ M jD-nitrobenzaldehyde in acetonitrile was photolyzed in the borosili-
cate tubes for reaction times of up to 10 minutes at either wavelength in the
merry-go-round photochemical reactor. The solutions were analyzed for starting
aldehyde using reverse phase HPLC with 50% acetonitrile in water as eluent.
The light incident on the tubes was then calculated according to the method of
Calvert and Pitts (1966), using a quantum yield for £-nitrobenzaldehyde
photolysis of 0.50.
The light flux (1^) incident on the samples in the 10-mm-O.D. tubes at
313 nm was about 1 xlO1^ photons sec"1 liter'l, and at 366 nm about 6 x
photons sec"1 liter"1. It was necessary to check 1^ frequently using the
JU
Light transmitted by the filters was determined with a grating monochromator.
387
-------
actinometer because the walls of the immersion well became dirty and diminished
the light from the Hg lamp. Precautions were especially necessary when using
the 313 nm filter system since the chromate solutions decomposed after ir-
radiation times of several days.
The natural waters used in laboratory photolysis studies were not
characterized with regard to the natural organic substances in them. Photolyses
were carried out for each compound in a "synthetic" natural water made with pure
water and the humic acid prepared as described in Section B.I. 4. Several dif-
ferent humic acid solutions were used in these studies over the two years work.
The absorbancies of these solutions at the particular wavelengths of interest
(either 313 or 366 nm) are given in the sections on the individual chemicals in
the body of the report.
As discussed in Part I of this report, the presence of light-absorbing
natural materials in waters has a screening effect that makes the photolysis
rate slower than in pure water solutions. To determine whether the natural
substances exerted any other effect in the photolyses (such as photosensiti-
zation or excited state quenching), a screening factor was calculated using
the expression (R.G. Zepp, private communication, 1977):
1 - 10"a*
S. = 9 -n, , (B.20)
A 2.303 a£
where a£ is the absorbance when a 1-cm cell is used and S, is the ratio of the
rate when ai > 0.02 to the rate when aH <0.02 (taken as pure water in this
study) . The product of S^ and the rate constant in pure water then gives the
rate constant that should be observed if only a screening effect is operative,
and with which the rate constant in natural waters should be compared. UV
spectral data for the 0.22-ym filtered natural waters are given in Table B.5.
TABLE B.5. UV ABSORBANCE OF 0.22-ym FILTERED NATURAL WATERS _
Absorbance in 1-cm cell, at wavelength X(nm)
Source _ 280 300 313 325 350 366 375 400
Coyote Creek 0.11 0.08 0.06 0.05 0.04 0.03 0.03 0.01
(Collected July 1976)
Lake Tahoe -« <0.01 *~
(Collected April 1976)
Searsville Pond 0.14 0.09 0.07 0.05 0.02 0.02 0.01 0.01
Aucilla River3 0.95 0.76 0.65 0.56 0.41 -- 0.29 0.21
water from the Aucilla River in Florida was provided by Dr. R. G. Zepp
and was used only in methyl parathion studies.
388
-------
Sunlight Photolyses—Outdoor photolyses using sunlight were carried out
with pure water solutions of each chemical to validate the computer calculation
of half-life in sunlight based on measured values of e, and 4>. Sunlight
photolyses require attention to placement of apparatus. Photolysed solutions
were placed in a location free of excessive reflections from walls and windows
and without morning and afternoon shadows. We used 11-mm-O.D. borosilicate
tubes held in a rack at a 60° angle to the horizon; the tubes were made from
the same glass stock used in the laboratory photolyses. These tubes are much
more readily sealed for long exposure and, judging from the good agreement
between computed and measured values for the half-life (t^/2) in sunlight for
most chemicals, are quite satisfactory for this purpose.
B.2.3 Free Radical Oxidaticjn
To screen the susceptibility of chemicals to free radical oxidation, the
chemical was oxidized using 4,4'-azobis(4-cyanovaleric acid) (AA) as the source
of free radicals. Solutions of chemicals of concentrations 1 yg ml"-*- or less
were prepared in pure water that contained 1.00 x 10~^ M AA. The solutions,
saturated with air by shaking, were placed in a 50°C water bath for 100 hours.
Solutions of the chemical at the same concentration without AA were run simul-
taneously as controls under identical reaction conditions. The 100-hour re-
action time corresponds to one half-life for AA. The solutions were then analyzed
for starting chemical and products.
For chemicals that showed significant oxidation at 100 hours, the same
experiment was repeated at shorter reaction times and to lower conversions.
Data for the disappearance of the chemical as a function of time were then
fit to a first-order rate law using a computer least squares program. This
rate constant was used to estimate rate constants for free radical oxidation of
chemicals in aquatic environments (see Section 6.3 in Part I of this report).
B.2.4 Hydrolysis
Screening reactions for hydrolysis were carried out for mirex, as
discussed in Section 14. The detailed hydrolysis studies of methyl parathion
are described in B.13 and Section 13.
B.3 BIODEGRADATION
The basic procedures used in evaluating biodegradability are described in
Part I of this report and the minor variations are designated in the appropriate
sections dealing with biodegradability studies on specific compounds.
The principal natural aquatic reservoirs used as representative sources
for cultures were:
• A eutrophic pond near Searsville Lake in Woodside, California.
• Coyote Creek, a eutrophic stream in San Jose, California.
389
-------
* Aeration effleunt from the Palo Alto, California, sewage treatment
facility. The organic matter treated is primarily of domestic origin.
* Aeration effluent from the treatment plant in the Shell Oil Refinery,
Martinez, California. This plant treats wastes that could have a
great similarity to those that might be expected from a coal
liquefaction plant.
* Aeration effluent from the sewage plant of the Monsanto Chemical
Company installation in Anniston, Alabama, where parathion and
methyl parathion are produced.
* Lake Tahoe, California, a large, deep, cold oligotrophic lake
between California and Nevada.
The water quality from the two eutrophic sources varied considerably with
the season but there is no means of avoiding this problem. In fact, this can
be a positive factor in the results of our investigations when the results
are compared with those obtained with the more regulated sewage or wastewater
treatment plants. Arbitrarily, incubation temperatures of 25°C were selected
for all incubations associated with waters other than Lake Tahoe. Samples from
Lake Tahoe were all incubated at 15°C. The selection of the media for enrich-
ment studies was also arbitrary. The high phosphate level (0.2%) was necessary
to maintain a constant pH with many water sample that were incubated without any
added substrate.
All enrichment processes were conducted under aerobic conditions with the
initial water samples diluted 4:5 to provide a Kf^PO^-^HPO^ mixture (equivalent
to 0.2%) for buffering at pH 7.0 and (NH^SO^ or NI^NC^ (equivalent to 0.01%)
nitrogen supplement. The concept was to use samples as large as possible.
Enrichments were generally initiated with 5-liters of diluted water samples in
9-liter bottle fermentors, but where preliminary observations indicated a rapid
development of biodegrading organisms, 1.2 liter of diluted samples was used
in 2.6-liter Fernbach flasks incubated in rotatory action shakers with tempera-
ture control.
Rapid estimations of pollutant degradation were made by extracting with
two volumes of solvent at acidic or neutral pH and then determining the UV
absorption characteristics of the solvent extracts. The details of the pro-
cedures are indicated in each section dealing with a specific chemical in the
body of the report. In some cases, this method was not applicable to the first
incubations with water samples because of interfering UV absorbing compounds in
the water sample.
The periods between transfers differed with the development of biodegrading
cultures. Generally, subtransfers were made to basal salts/added compounds
media in which test compounds were present at the original concentration and at
an increased concentration.
390
-------
B.4 £-CRESOL
B.4.1 Source
£-Cresol (99+% Aldrlch "Gold Label") was used as received
B.4.2 Physical Transport Experiments
Analysis—The 1-ml aqueous samples were extracted with 1 ml diethyl ether
and analyzed directly by GC. Recovery was > 99%.
Instrument: Perkin-Elmer Model 3920, FID
Column: 3 m x 3 im O.D. glass, packed with 10% FFAP
on 80/100 mesh acid-washed Chromosorb W DCMS.
Column temperature:; 190°C
Carrier gas: N2 at 60 ml min"-'-
Retention time: 4.75 min
Internal standard: cv-cresol
Sensitivity: 1 ng
Volatilization rate—Solutions of _p-cresol containing 5 yg ml~^ were
stirred in 2000-ml beakers. Aliquots were removed at suitable intervals over
121 hours and analyzed.
B.4.3 Chemical Transformation Experiments
Analysis—Aqueous samples were extracted three times with equal volumes
of ethyl acetate and analyzed directly by GC.
Instrument: Hewlett-Packard 700 GC, FID
Column: 3 m x 3 mm O.D. stainless steel packed with 10% FFAP
on acid-washed Chromosorb W
Column temperature: 165°C
Carrier gas: N2
Internal standard: p_-nitrotoluene
B.4.4 Biodegradation Experiments
Analysis—Aqueous samples and cells were extracted with two equal volumes
of ethyl acetate, dried over Na2SOtt, concentrated, and analyzed by GC.
Instrument: Hewlett-Packard 5711 GC, FID
Column: 1.8 m x 2 mm 10% DECS on 80/100 mesh Chromosorb W(HP)
Column temperature: 175°C
Carrier gas: N2 at 35 ml min~l
Retention time: 6.36 min
Internal standard: 1,3,4,5,6,7,8-octahydroanthracene
391
-------
B.5 BENZ[a]ANTHRACENE
B.5.1 Source
Benz[a]anthracene (Eastman) was used as received.
B.5.2 Physical Transport Experiments
Analysis—Aqueous samples were analyzed by direct injection onto the
HPLC column. The sample was first diluted by 25% with acetonitrile to minimize
sorption onto the sample loop. The sample loop size was 3.5 ml.
Instrument: Du Pont Model 848; Aminco fluorescence detector
Column: 22 cm x 4.6 mm O.D. Du Pont Zorbax ODS
Mobile phase: 90% acetonitrile, 10% H20
Pressure: 2000 psig
Internal standard: phenanthrene
Sediments were extracted with ethyl acetate. A 20-ul sample was analyzed
by HPLC using the above method, except that the mobile phase was 70% acetonitrile
30% water at a pressure of 3000 psig.
Volatilization rate—A solution of benz[a]anthracene containing 4 ng ml
was stirred in a 2000-ml beaker. Samples were removed at suitable intervals
over a 24 hour period.
B.5.3 Chemical Transformation Experiments
Analysis—Aqueous samples were analyzed by direct injection onto the
HPLC column.
Instrument: Waters Associates Model 6000A; Model M440 UV
detector at 280 nm
Column: 30 cm x 4 mm pBondapak C-^g
Mobile phase: 80% acetonitrile, 20% water
B.5.4 Biodegradation and Biosorption
Analysis—Aqueous samples and cells were extracted twice with equal
volumes of ethyl acetate.and analyzed by HPLC.
Instrument: Spectrophysics Model 3500 HPLC; fluorescence detector
(Aminco) at 360 nm
Column: 30 cm x 4 mm yBondapak C-^g
Mobile phase: 75% methanol, 25% water
Flow rate: 1.6 ml min"-*-
Retention time : 280 seconds
Internal standard: benzo[a]pyrene
Detection limit: 9 ng
392
-------
B.6 BENZO[a]PYRENE
B.6.1 Source
Benzo[a]pyrene (99+% Aldrich Gold Label) was used as obtained.
B.6.2 Physical Transport Experiments
Analysis—For solubility studies, 1000 ml samples were extracted
twice with 50 ml hexane and concentrated to 1 ml; 10 yl was analyzed by HPLC.
Instrument: Waters Associates Model 6000A; M440 UV detector at
254 hm •
Column: 30 cm x 4 mm uBondapak CJ^Q (Waters Associates)
Mobile phase: 80% acetonitrile, 20% water
Flow rate: 3.6 ml min~l
Internal standard: triphenylbenzene
Calcium montmorillonite clay isotherm supernatant solutions were
extracted in the same manner and analyzed by HPLC using the following conditions:
Instrument: Du Pont Model 848; UV detector at 265 nm
Mobile phase: 99.5% hexane, 0.5% isopropanol
Pressure: 2000 psig
Column: 22 cm x 2 mm Zorbax SIL
Internal standard: triphenylbenzene
Sensitivity: 2 ng
Samples (100-ml) of the remaining sorption isotherm supernatant and
volatilization solutions were extracted with 10 ml benzene and analyzed by
HPLC;
Instrument: Du Pont Model 848; Aminco fluorometer detector
Column: 22 cm x 4.6 mm Zorbax ODS
Mobile phase : 100% methanol
Pressure: 800 psig
Internal standard: pyrene
Sensitivity: 500 pg
Volatilization rate—A solution containing 0.95 ng ml"1 benzo[a]pyrene
was stirred in a 2000-ml beaker for 72 hours; samples were removed periodically
for analysis.
B.6.3 Chemical Transformation Experiments
Analysis—Aqueous samples were analyzed by direct injection onto the
HPLC column.
393
-------
Instrument: Waters Associates Model 6000A; Model M440 UV detector
at 254 nm
Column: 30 cm x 4 mm uBondapak C±Q (Waters Associates)
Mobile phase: 70% methanol, 30% water
Internal standard: 1,3,5-tris(4-methoxyphenyl)benzene
B.6.4 Biojiegradation and Biosorption Experiments
Analysis—Aqueous samples and cells were extracted with two equal volumes
of ethyl acetate and analyzed by HPLC. Some samples were dried and concentrated
before analysis.
Instrument: Spectraphysics Model 3500; Aminco fluorometer detector
at 360 nm
Column: 30 cm x 4 mm uBondapak C-^g (Waters Associates)
Mobile phase: 85% methanol, 15% water
Flow rate : 1.6 ml min"-'-
Retention time: 400 sec
Internal standard: chrysene
Sensitivity: 100 pg
B.7 QUINOLINE
B.7.1 Source
Quinoline (96% Aldrich) was purified by preparative gas chromatography
before use.
B.7.2 Physical Transport Experiments
Analysis—Aqueous 1-ml samples were extracted with 1 ml CS2; 1-ul samples
were analyzed by GC.
Instrument: Perkin-Elmer Model 3920, FID
Column: 6 ft x 1/8 in^ 5% Carbowax 20M on 80/100 mesh Chromosorb
W DCMS
Column temperature: 170°C
Carrier gas: N£ at 60 ml min~
Retention time: 165 seconds
Internal standard: naphthalene (RT = 90 sec)
Volatilization rate—A solution containing 8 yg ml"-'- quinoline was
stirred in a 2000-ml beaker. Samples were removed at suitable intervals over
100 hours.
B.7.3 Chemical Transformation Experiments
Analysis—The pH of the aqueous samples was adjusted to be greater than
12; three extractions were made with equal volumes of ethyl acetate and analyzed
by GC.
394
-------
Instrument: Hewlett-Packard 5700; FID
Column: 300 cm x 0.32 cm 10% FFAP on A/W Chromosorb W
Column temperature: 165°C
Carrier gas: N2
Internal standard: biphenyl
B.7.4 Bipdegradation and Biosorption Experiments
Analysis—Aqueous samples and cells were extracted with two equal
volumes of ethyl acetate and analyzed by HPLC.
Instrument: SpectraPhysics Model 3500; Schoeffel Variable UV
detector at 220 nm
Mobile phase: 60% methanol, 40% water
Column: 30 cm x J4 mm pBondapak C-.Q (Waters Associates)
Flow rate: 1.6 ml min~l
Retention time: 320 seconds
Sensitivity: 100 pg
B.8 BENZO[f]QUINOLINE
B.8.1 Source
Benzo[f]quinoline (Eastman) was used as received.
B.8.2 Physical Transport Experiments
Analysis—The aqueous samples were diluted until they contained 57%
acetonitrile; 0.75 ml was injected directly onto the HPLC column.
Instrument: Du Pont Model 848; UV detector at 254 nm
Column: 4.6 mm x 22 cm Zorbax ODS
Mobile phase: 90% methanol, 10% water
Pressure: 2000 psig
-Internal standard: anthracene
Sediments were extracted with 15 ml ethyl acetate and concentrated to
1 ml. The extracts were analyzed under the conditions listed above.
Volatilization rate—A solution containing 12 pg ml" benzo[fJquinoline
was stirred in a 2000-ml beaker; samples were removed at suitable intervals
over 104 hours.
B.8.3 Chemical Transformation Experiments
Analysis—Aqueous samples were analyzed by direct injection onto the
HPLC column.
395
-------
Instrument: Waters Associates Model 660A; Model M440 UV detector
at 254 nm
Column: 30 cm x 4 mm yBondapak C-^g
Mobile phase: 75% methanol, 25% water
B.8.4 Biodegradation and Biosorption Experiments
Analysis—Aqueous samples and cells were extracted twice with equal
volumes of ethyl acetate and analyzed by HPLC.
Instrument: SpectraPhysics 3500; Schoeffel variable fluorescence
detector at 360 nm
Column : 24 cm x 6 mm Spherisorb ODS (SpectraPhysics)
Mobile phase: 55% methanol, 45% water
Flow rate: 1.6 ml min"-'-
Retention time: 190 seconds
Internal standard: anthracene
Sensitivity: 1 ng
B.9 9H-CARBAZOLE
B.9.1 Source
9H-Carbazole (99+% Aldrich) was recrystallized from methanol/ethyl
acetate before use.
B.9.2 Physical Transport Experiments
Analysis—In the solubility experiments, 5 ml of solution was extracted
twice with 2 ml, and then 1 ml of ethyl acetate, and concentrated to 0.5 ml;
3-yl samples were analyzed by GC.
Instrument: Hewlett-Packard 5750; FID
Column: 10 ft x 1/8 in. 5% SE-30
Column temperature: programmed from 155° to 188°C at 8° min~
Carrier gas: helium at 45 ml min
Retention time : 4 minutes
Internal standard: fluoranthene (RT = 6 minutes)
Sensitivity: 5 ng
Direct injection of 20-yl aqueous samples onto a HPLC column was used
in the sorption isotherm and volatilization analyses.
Instrument: Du Pont Model 848; UV detector at 245 nm
Column: 4.6 mm x 25 cm Zorbax ODS
Mobile phase: 75% acetonitrile, 25% water
Pressure: 2000 psig (3.1 ml min~l)
396
-------
Volatilization rate—A solution of carbazole containing 0.6 ygml"1 was
stirred in a 2000-ml beaker; samples were removed at suitable intervals over
an 80-hour period.
B.9.3 Chemical Transformation Experiments
Analysis—Aqueous samples were injected directly onto the HPLC column.
Instrument: Waters Associates Model 660A; Model M440 UV detector
at 254 nm
Column: 30 cm x 4 mm yBondapak C-ig (Waters Associates)
Mobile phase: 55% acetonitrile, 45% water
B.9.4 Biodegradation and Biosorption Experiments
i
Analysis—Aqueous samples and cells were extracted twice with equal
volumes of ethyl acetate and analyzed by HPLC.
Instrument: SpectraPhysics Model 3500; SpectraPhysics UV detector
at 254 nm and Aminco fluorometer at 254 nm
Column: 24 cm x 6 mm Spherisorb ODS (SpectraPhysics)
Mobile phase: 65% methanol, 35% water
Flow rate: 1.6 ml min~^-
Retention time: 360 seconds
Internal standard: anthracene
Sensitivity: 70 ng (UV),6ng (fluorometer)
B.10 7H-DIBENZO[g,g]CARBAZOLE
B.10.1 Source
7H-Dibenzocarbazole (98% Aldri'ch) was used as received.
B.10.2 Physical Transport Experiments
Analysis—Solubility measurements were carried out by direct aqueous
injection of 1 ml onto the HPLC column.
Instrument: Du Pont Model 848; Aminco fluorescence detector
Column: 24 cm x 6 mm Spherisorb ODS
Mobile phase: 95% methanol, 5% water
Volatilization and isotherm supernatant solutions were diluted until
they contained 25% acetonitrile. They were analyzed by direct aqueous injection
of 100 yl onto the HPLC column.
Instrument: Du Pont Model 848; Aminco fluorometer
Column: 4.6 mm x 22 cm Zorbax ODS
397
-------
Mobile phase: 85% acetonitrile, 15% water
Pressure: 2000 psig
Sediments were extracted with 10 ml ethyl acetate and analyzed under
the same conditions described above.
Volatilization rate—A solution of DEC containing 16 ng ml~^ was stirred
in a 2000-ml beaker; samples were removed at suitable intervals over a 5-day
period.
B.10.3 Chemical Transformation Experiments
Analysis—Samples were analyzed by direct aqueous injection onto the
HPLC column.
Instrument: Waters Associates Model 660A; M440 UV detector at 254 nm
Column: 30 cm x 4 mm VBondapak Cjg (Waters Associates)
Mobile phase: 85% methanol, 15% water
Internal standard: triphenylbenzene
B.10.4 Biodegradation and Biosorption Experiments
Analysis—Aqueous samples and cells were extracted with two equal volumes
of ethyl acetate and analyzed by HPLC.
Instrument: SpectraPhysics Model 3500; Aminco fluorescence detector
at 360 nm
Column: 30 cm x 4 mm pBondapak C^g (Waters Associates)
Mobile phase: 80% methanol, 20% water
Flow rate: 1.6 ml min"-*-
Retention time: 340 seconds
Internal standard: benzo[a]pyrene
Sensitivity: 50 pg
B.ll BENZO[b]THIOPHENE
B.ll.l Source
Benzo[b]thiophene (thianaphthene, Aldrich 97%) was recrystallized from
methanol before use.
B.ll.2 Physical Transport Experiments
Analysis—Aqueous 2-ml samples were extracted twice with 0.9 ml CS« and
the volume adjusted to 2 ml with CS£; 4-yl injections were made onto the dC.
398
-------
Instrument: Hewlett-Packard 5750; FID
Column: 6 ft x 1/8 in. 5% Carbowax 20M
Column temperature: 124°C
Carrier gas: He at 45.5 ml min"1
Internal standard: naphthalene
Volatilization rate—Solutions of benzothiophene containing from 18 to
25 pgml"1 were stirred in 2000-ml beakers for 75 to 450 minutes. Samples were
removed and analyzed at suitable intervals.
B.11.3 Chemical Transformation Experiments
Analysis—Aqueous sjamples were injected directly onto the HPLC column.
Instrument: Waters Associates Model 6000A; M440 UV detector at 254 nm
Column: 30 cm x 4 mm yBondapak Cjg (Waters Associates)
Mobile phase: 55% acetonitrile, 45% water
Internal standard: phenanthrene
B.11.4 Biodegradation and Biosorption Experiments
Analysis—In the biodegradation experiments, aqueous samples and cells
were extracted with two equal volumes of ethyl acetate and analyzed
by HPLC. Biosorption analyses were done in the same manner using hexane as
extractant.
Instrument: SpectraPhysics Model 3500; Schoeffel variable UV
detector at 228 nm
Column: 30 cm x 4 mm pBondapak C-^g (Waters Associates)
Mobile phase: 60% methanol, 40% water
Flow rate: 1.6 ml min~^
Retention time: 380 seconds
Sensitivity: 500 pg
B.12 DIBENZOTHIOPHENE
B.12.1 Source
Dibenzothiophene (Aldrich 95%) was recrystallized from methanol before
use.
B.12.2 Physical Transport Experiments
Analysis—After adjustment to contain 25% acetonitrile, the aqueous
samples were analyzed by direct injection onto the HPLC column. Sample size
was 0.75 ml.
399
-------
Instrument: Du Pont Model 848; UV detector at 254 nm
Column: 4.6 mm x 22 cm Zorbax ODS
Mobile phase: 80% acetonitrile, 20% water
Flow rate: 3.0 ml min~l
Sediments were extracted with 15 ml ethyl acetate and concentrated to
1 ml before analysis using the above conditions.
Volatilization rate — Solutions of DBT ranging in concentration from
140 to 200 ng ml""1 were stirred in 2000-ml beakers for 11 to 69 hours. Samples
were removed at suitable intervals for analysis.
B.12.3 Chemical Transformation Experiments
Analysis — Samples were analyzed by direct injection onto the HPLC column.
Instrument: Waters Associates Model 6000A; M440 UV detector at
254 nm
Mobile phase: 60% acetonitrile, 40% water
Column: 30 cm x 4 mm uBondapak C-g
B.12.4 Biodegradation and Biosorption Experiments
Analysis — Aqueous samples and cells were extracted three times with equal
volumes of ethyl acetate and analyzed by HPLC.
Instrument: SpectraPhysics Model 3500; SpectraPhysics 230 UV
detector at 254 nm
Column : 24 cm x 6 mm Spherisorb ODS (SpectraPhysics)
Mobile phase: 55% methanol, 45% water
Flow rate: 1.6 ml min~l
Retention time : 640 seconds
Internal standard: anthracene
Sensitivity : 20 ng
B.I 3 METHYL PARATHION
B.13.1 Source
Methyl parathion (Chem Services) was used as received.
B.13.2 Physical Transport Experiments
Analysis — Aqueous 2-ml samples were extracted twice with 1 ml hexane;
5-yl samples were analyzed directly by GC. Recovery was 99+%.
400
-------
Instrument: Perkin-Elmer Model 3920, FID
Column: 30% OV-1 on 100-120 mesh Gaschrom-Q
Column temperature: 190-195°C
Carrier gas: N2 at 80 ml min~"l
Internal standard: Tributyl phosphate
Volatilization rate—A solution containing 7 yg ml"-*- methyl parathion
was stirred in a 2000-ml beaker. Samples were removed at suitable intervals
for analysis.
B.13.3 Chemical Transformation Experiments
Analysis—Samples wpre analyzed by direct aqueous injection onto the
HPLC column.
Instrumnet: Waters Associates Model 6000A; M440 UV detector at
280 nm
Column: 30 cm x 4 mm yBondapak
Mobile phase: 50% acetonitrile, 50% water
Internal standard: _p_-nitro toluene
Hydrolysis rate—One volume of methyl parathion in acetonitrile was
diluted to 100 volumes with buffered or natural waters to give a 1.00 x 10^ M
(26.0 ppm) methyl parathion solution in 1% acetonitrile. More dilute solutions
were prepared by an analogous procedure.
Buffered solutions were made using deionized-distilled water with buffer
salts at the concentrations listed below. Buffers at 0.067 M in total phosphate
salt were used in early studies but were replaced when catalysis by hydrogen
phosphate dianion became evident. Acetate and borate buffers were used
successfully with no catalysis problems. The following buffer solutions were
used:
pH 3.00 100 ml 0.067 M NaH2P04, 0.50 ml 0.1 M H3P04
pH 5.00 (a) 100 ml 0.067 M NaH2P04, 3.00 ml 0.067 M K2HP04
(b) 0.01 ml sodium acetate, acetic acid to adjust pH
pH 6.00 88 ml 0.067 M NaH2P04, 12.00 ml 0.067 M K2HP04
pH 7.00 (a) 39 ml 0.067 M NaH2P04, 61.0 ml 0.067 M K2HP04
(b) Same solution diluted tenfold
PH 8.00 (a) 5.5 ml 0.067 M NaH2P04, 94.5 ml 0.067 M K2HP04
(b) 0.01 ml Na2B4Oy
pH 9.00 (a) 0.025 M Na2B407, 0.001 ml acetic acid to adjust pH
pH 10.00 0.010 M Na2B40?, 0.001 M NaOH to adjust pH
pH 11.25 0.050 M K2C03
401
-------
All pH solutions were adjusted to the given pH values and were checked
using a Coleman Model 39 pH meter calibrated against standard buffer solutions.
Reaction mixtures in volumetric flasks were immersed in a darkened,
thermostated water bath at 20°, 40°, 50°, and 70°C. The solutions were pre-
pared and the reactions were conducted in the dark, to prevent photolytic
degradation, and were analyzed immediately after removal from the thermostated
bath. Reactions were carried through at least two half-lives of methyl para-
thion to verify pseudo—first—order kinetics in each type of water and pH.
B.13.4 Biodegradation and Biosorption Experiments
Analysis—Aqueous samples and cells were extracted with two equal
volumes of ethyl acetate and analyzed directly by GC.
Instrument: Varian Associates Model 5711; APIS (Varian Mode1 2740)
Column: 5 ft x 1/4 in. 4% SE-30 + 6% OV-210 on 80/160 mesh Gaschrom Q
Column temperature: 200°C
Carrier gas: N2 at 33 ml min"-'-
Retention time: 3.73 minutes
Internal standard: ethyl parathion
Sensitivity: 50 pg
B.14 MIREX
B.14.1 Source
Mirex (Chem Services) was purified by column chromatography through
silica gel followed by sublimation.
B.14.2 Physical Transport Experiments
Analysis—Aqueous 150-ml samples and sediments were extracted twice with
10 ml hexane, dried with sodium sulfate, and concentrated to 0.1 ul; 5-yl
injections were analyzed by GC.
Instrument: Tracor Model 220; Ni63 electron capture detector
Column: 5 ft x 1/8 in. OV-210
Temperature 180°C
Carrier and purge gas: 5% methane, 95% argon
Sensitivity: 10 pg
Special precautions were taken to avoid contamination of the samples:
Glassware was cleaned by baking at 700°C and/or by multiple rinses in Milli-Q
water and distilled-in-glass hexane.
For extraction, the distilled-in-glass hexane was further purified with
fuming sulfuric acid, followed by neutralization with sodium bicarbonate, then
redistillation over sodium sulfate. The sodium sulfate used to dry the
402
-------
extracts was purified by boiling 3 times for 10 minutes in excess distilled-
in-glass hexane and drying at 100°C.
Volatilization rate—A solution of mirex containing 86 pg ml"-'- was
stirred in a 2000-ml beaker for 60 hours; samples were removed at suitable
intervals for analysis.
B.14.3 Chemical Transformation Experiments
Analysis—Aqueous 100-ml samples were extracted twice with 50 ml
distilled-in-glass hexane, dried with Na2SO^, and concentrated to 0.5 ml;
5-yl aliquots were analyzed by GC.
Instrument: Tracer Model 220, Coulson electrolytic conductivity
detector
Column: 3 ft, 5%!SE-30
Column temperature: 170°-220°C at 5° min
Carrier gas: helium
Internal standard: aldrin
B.14.4 Biodegradation and Biosorption Experiments
Analysis—Aqueous samples and cells were extracted twice with equal
volumes of 40% isopropanol in hexane and analyzed by GC.
Instrument: Varian Model 5711, electron capture detector
(Analog Technology, scandium foil)
Column: 5 ft x 2 mm 3% OV-17 on Gas Chrom-Q
Column temperature: 220°C
Carrier gas: N2 at 30 ml min"-1-
Internal standard: Methoxychlor
Sensitivity: 90 pg
403
-------
Appendix C
NATURAL WATER SOURCES
One goal of this research was to determine what comparisons can be made
between experimental data obtained using several actual waters and sediments.
To this end, screening and detailed studies were planned to include natural
waters whenever possible. This appendix describes the sources of the natural
waters and sediments to provide a more complete view of the types of waters our
sources represent.
C.I LAKE TAHOE
Lake Tahoe is located in the Sierra Nevada at the California-Nevada border
at 39° latitude. A recent article by Holm-Hansen et al. (1976) describes the
chemical and biological characteristics of a water column in Lake Tahoe. This
lake is large (499 km^) and very oligotrophic. Surface temperatures vary from
7° to 20°C with season, with the temperature below 50 meters constant at about
6°C. Additional information regarding land use in the Tahoe watershed is avail-
able in an EPA study (Wise, 1975).
Water and sediment samples for our studies were taken about 30 meters off-
shore at Sugar Pine Point on the California side of the lake. Water samples
were taken at several intermediate depths to 30 meters where the sediment sample
was taken for biodegradation screening studies. Some water quality parameters
for a sample of water that was filter-sterilized are given in Table B.4; UV
spectral data for this water are given in Table B.5.
C.2 COYOTE CREEK
Coyote Creek originates near the Blue Ridge and Pine Ridge areas of the
coast mountain range in Santa Clara County, California. Fed by springs and
runoff, Coyote Creek drains first into Coyote Reservoir and then into Anderson
Reservoir. It continues below the reservoir and parallels U.S. Highway 101 for
about 40 km until it enters the South San Francisco Bay delta system. The creek
system runs through agricultural and light urban regions below the second
reservoir until it passes through the City of San Jose near the bay. In the
agricultural region, seasonal canning industries discharge some wastes into the
creek.
Our samples were taken in a park near San Jose State University in San Jose,
where the creek is about 3 meters wide and less than 1 meter deep. The water is
eutrophic, and our water samples had a slight yellow coloration, much of which
was removed by filtration. Water quality parameters for the filter-sterilized
water are given in Table B.4; UV spectral data for this water are given in
Table B.5. Characteristics of a sediment from Coyote Creek at our sampling
point are given in Table 5.3 of Part I of this report.
404
-------
C.3 SEARSVILLE POND
Searsville Pond is an unofficial name that we assigned to our sampling
location in a wetlands area near Searsville Lake. This is a light urban region
about 6 km west of Menlo Park and the Stanford University campus. Most of the
wetlands area was formerly a game refuge and is now part of the Jasper Ridge
Biological Preserve of Stanford University. The wetlands area is fed by
Alembique Creek, an intermittent, highly organic creek originating in the
foothills of the coastal mountain range in San Mateo County.
Our sampling location was on the west side of a causeway over the wetlands
area, where emergent vegetation surrounded about 2000 m^ of open water, which
we used as our sampling pond. After two years of low rainfall, this area dried
up in the spring of 1977 after our laboratory studies had been completed.
While the waters from this area were definitely eutrophic, there were no obvious
anthropogenic discharges tb either the pond or wetlands areas. The uplands
surrounding the wetland are covered by a mosaic of oak-woodland, chaparral, and
annual grassland vegetation.
Some water quality parameters and UV spectral data for a sample of filtered
water are given in Tables B.4 and B.5, respectively. Characteristics of a
sediment from this pond are given in Table 5.3 of Patt I of this report.
405
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-78-074
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
ENVIRONMENTAL PATHWAYS OF SELECTED CHEMICALS IN FRESH-
WATER SYSTEMS Part II: Laboratory Studies
5. REPORT DATE
May 1978 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.H. Smith, W.R. Mabey, N. Bohonos, B.R. Holt, S.S. Lee,
T-W. Chou, D.C. Bomberger, and T. Mill
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
SRI International
333 Ravenswood Avenue
Menlo Park, CA 94025
10. PROGRAM ELEMENT NO.
1NE625
11. CONTRACT/GRANT NO.
68-03-2227
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory—Athens, GA
Office of Research and Development
U.S. Environmental Protection Agency
Athens, GA 30605
13. TYPE OF REPORT AND PERIOD COVERED
Final, 6/30/75 to 4/30/77
14. SPONSORING AGENCY CODE
EPA/600/01
15. SUPPLEMENTARY NOTES
See ENVIRONMENTAL PATHWAYS OF SELECTED CHEMICALS IN FRESHWATER SYSTEMS Part I:
ground and Experimental Procedures (EPA-600/7-77-113)
Back-
16. ABSTRACT
Environmental exposure assessment models and laboratory procedures for
predicting the pathways of potentially harmful chemicals in freshwater environ-
ments were described in Part I of this report (EPA-600/7-77-113). Procedures
were developed for measuring the rates of volatilization, photolysis, oxidation,
hydrolysis, and biotransformations as well as the sorption partition coefficients
on natural sediments and on a mixture of four bacteria. The results were inte-
grated with a simple computer model to predict the pathways of chemicals in
ponds, rivers, and lakes. This second part of the project report describes
the successful application of these procedures to 11 chemicals of environmental
interest. The chemicals were £-cresol, benz[a]anthracene, benzo[a]pyrene,
quinoline, benzo[fJquinoline, 9H-carbazole, 7H-dibenzo[c,g]carbazole, benzo[b]-
thiophene, and dibenzothiophene, which might be found in the effluents of
plants using or processing fossil fuels, and methyl parathion and mirex,
which are agricultural pesticides.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Adsorption, Sorption, Oxidation,
lydrolysis, Photolysis,
Contaminants, Degradation, Mathematical
aodels, Hydrocarbons, Pesticides,
Aromatic polycyclic hydrocarbons.
Environmental Assessment,
Volatilization, Microbial
degradation, Biodegra-
dation, Transformation,
Aquatic systems, Environ-
nental simulations, naturs
waters, Solar photolysis,
R-f OOf.T-1-
57H
99A
48G
68D
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
432
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
406
if U. S. GOVERNMENT PRIMTING OFFICE: 1978 — 757-140/1302
------- |