rxEPA
United States
Environmental Protection
Agency
Environmental Criteria and
Assessment Office
Research Triangle Park NC 27711
600884020A2
July 1984 £t I
External Review Draft
Research and Development
Air Quality
Criteria for
Ozone and Other
Photochemical
Oxidants
Review
Draft
(Do Not
Cite or Quote)
Volume II of V
NOTICE
This document is a preliminary draft. It has not been formally
released by EPA and should not at this stage be construed to
represent Agency policy. It is being circulated for comment on its
technical accuracy and policy implications.
-------
NOTICE
Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
U.S Environmental Protection Agency
-------
ABSTRACT
Scientific information is presented and evaluated relative to the health
and welfare effects associated with exposure to ozone and other photochemical
oxidants. Although it is not intended as a complete and detailed literature
review, the document covers pertinent literature through 1983 and early 1984.
Data on health and welfare effects are emphasized, but additional infor-
mation is provided for understanding the nature of the oxidant pollution pro-
blem and for evaluating the reliability of effects data as well as their
relevance to potential exposures to ozone and other oxidants at concentrations
occurring in ambient air. Separate chapters are presented on the following
exposure-related topics: nature, source, measurement, and concentrations of
precursors to ozone and other photochemical oxidants; the formation of ozone
and other photochemical oxidants and their transport once formed; the proper-
ties, chemistry, and measurement of ozone and other photochemical oxidants;
and the concentrations of ozone and other photochemical oxidants that are
typically found in ambient air.
The specific areas addressed by chapters on health and welfare effects
are the toxicological appraisal of effects of ozone and other oxidants; effects
observed in controlled human exposures; effects observed in field and epidemio-
logical studies; effects on vegetation seen in field and controlled exposures;
effects on natural and agroecosystems; and effects on nonbiological materials
observed in field and chamber studies.
in
0190LG/B May 1984
-------
CONTENTS
Page
VOLUME I
Chapter 1. Summary and Conclusions 1-1
VOLUME II
Chapter 2. Introduction 2-1
Chapter 3. Precursors to Ozone and Other Photochemical
Oxi dants 3-1
Chapter 4. Chemical and Physical Processes in the Formation
and Occurrence of Ozone and Other Photochemical
Oxidants 4-1
Chapter 5. Properties, Chemistry, and Measurement of Ozone
and Other Photochemical Oxidants 5-1
Chapter 6. Concentrations of Ozone and Other Photochemical
Oxidants in Ambient Air 6-2
VOLUME III
Chapter 7. Effects of Ozone and Other Photochemical Oxidants
on Vegetation 7-1
Chapter 8. Effects of Ozone and Other Photochemical Oxidants
on Natural and Agroecosystems 8-1
Chapter 9. Effects of Ozone and Other Photochemical Oxidants
on Nonbiological Materials 9-1
VOLUME IV
Chapter 10. Toxicological Effects of Ozone and Other
Photochemical Oxidants 10-1
VOLUME V
Chapter 11. Controlled Human Studies of the Effects of Ozone
and Other Photochemical Oxidants 11-1
Chapter 12. Field and Epidemiological Studies of the Effects
of Ozone and Other Photochemical Oxidants 12-1
Chapter 13. Evaluation of Integrated Health Effects Data for
Ozone and Other Photochemical Oxidants 13-1
0190LG/B
IV
May 1984
-------
TABLE OF CONTENTS
LIST OF TABLES xi
LIST OF FIGURES xiv
LIST OF ABBREVIATIONS AND SYMBOLS xvi i i
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxi i i
2. INTRODUCTION 2-1
2.1 PURPOSE AND LEGISLATIVE BASIS OF DOCUMENT 2-1
2.2 THE OXIDANT PROBLEM 2-2
2.3 SCOPE AND ORGANIZATION OF DOCUMENT 2-4
2.4 REFERENCES 2-7
3. PRECURSORS TO OZONE AND OTHER PHOTOCHEMICAL OXIDANTS 3-1
3.1 INTRODUCTION 3-1
3.2 DESCRIPTION AND CHARACTERIZATION OF PRECURSORS 3-1
3.2.1 Description and Basic Nomenclature of Nonmethane
Organic Compounds 3-1
3.2.1.1 Hydrocarbons 3-2
3.2.1.2 Aldehydes 3-4
3.2.1.3 Other Organic Compounds 3-5
3.2.2 Pertinent Chemical and Physical Properties of
Nonmethane Organic Compounds 3-5
3.2.3 Description and Properties of Nitrogen Oxides 3-7
3.3 SAMPLING, MEASUREMENT, AND CALIBRATION METHODS FOR
PRECURSORS TO OZONE AND OTHER PHOTOCHEMICAL OXIDANTS 3-8
3.3.1 Nonmethane Organic Compounds 3-9
3.3.1.1 Nonmethane Hydrocarbons 3-9
3.3.1.2 Aldehydes 3-24
3.3.1.3 Other Oxygenated Organic Species 3-29
3.3.2 Nitrogen Oxides 3-30
3.3.2.1 Measurement Methods for N02 and NO 3-30
3.3.2.2 Sampling Requirements 3-35
3.3.2.3 Calibration 3-36
3.4 SOURCES AND EMISSIONS OF PRECURSORS 3-37
3.4.1 Manmade Sources and Emissions 3-37
3.4.1.1 Trends in Emissions of Volatile Organic
Compounds 3-39
3.4.1.2 Trends in Emissions of Nitrogen Oxides 3-39
3.4.1.3 Geographic Distribution of Manmade
Emissions of Volatile Organic Compounds 3-42
3.4.1.4 Geographic Distribution of Manmade
Emissions of Nitrogen Oxides 3-42
3.4.1.5 Profiles of Emissions of Volatile Organic
Compounds 3-42
3.4.1.6 Profiles of Emissions of Nitrogen Oxides .... 3-55
019FFM/A 6/30/84
-------
TABLE OF CONTENTS
(continued)
Paqe
3.4.2 Natural Sources and Emissions 3-62
3.4.2.1 Natural Sources and Emissions of Volatile
Organic Compounds 3-62
3.4.2.2 Natural Sources and Emissions of Nitrogen
Oxides 3-75
3.4.2.3 Local Natural Sources of NO 3-79
3.5 REPRESENTATIVE CONCENTRATIONS OF OZONE PRECURSORS IN
AMBIENT AIR 3-81
3.5.1 Concentrations of Nonmethane Organic Compounds in
Ambient Air 3-81
3.5.1.1 Urban Nonmethane Hydrocarbon Concentrations .. 3-82
3.5.1.2 Nonurban Nonmethane Hydrocarbon
Concentrations 3-83
3.5.1.3 Nonmethane Hydrocarbon Concentrations Aloft .. 3-87
3.5.1.4 Urban Aldehyde Concentrations 3-87
3.5.1.5 Aldehyde Concentrations in Rural
Atmospheres 3-92
3.5.2 Atmospheric Concentrations of Nitrogen Oxides 3-92
3.5.2.1 Urban NO Concentrations 3-92
3.5.2.2 NonurbanxNO Concentrations 3-93
3.6 SUMMARY 3-95
3.6.1 Nature of Precursors to Ozone and Other Photochemical
Oxidants 3-95
3.6.2 Measurement of Precursors to Ozone and Other
Photochemical Oxidants 3-96
3.6.3 Sources and Emissions of Precursors 3-99
3.6.4 Ambient Air Concentrations of Precursors 3-100
3.6.4.1 Hydrocarbons in Urban Areas 3-100
3.6.4.2 Hydrocarbons in Nonurban Areas 3-100
3.6.4.3 Aldehydes in Urban Areas 3-101
3.6.4.4 Aldehydes in Nonurban Areas 3-101
3.6.4.5 Nitrogen Oxides in Urban Areas 3-101
3.7 REFERENCES 3-102
4. CHEMICAL AND PHYSICAL PROCESSES IN THE FORMATION AND
OCCURRENCE OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS 4-1
4.1 INTRODUCTION 4-1
4.2 CHEMICAL PROCESSES 4-1
4.2.1 Formation of Ozone and Oxidants 4-2
4.2.2 Initiation and Termination of Photochemical
Reacti ons 4-4
4.2.3 Limitations to Ozone Accumulation 4-6
4.2.4 Recent Work on Photochemical Smog Reactions 4-6
4.2.5 Relationship of Ozone to Aerosol-Related
Phenomena 4-9
VT
019FFM/A 6/30/84
-------
TABLE OF CONTENTS
(continued)
Page
5.
4. 3 METEOROLOGICAL AND CLIMATOLOGICAL PROCESSES
4.3.1 Atmospheric Mixing
4.3.2 Wind Speed and Mixing
4.3.3 Effects of Sunlight and Temperature
4.3.4 Transport of Ozone and Other Oxidants and
Thei r Precursors
4.3.5 Surface Scavenging in Relation to Transport
4. 3. 6 Stratospheric-Tropospheric Ozone Exchange
4. 3. 7 Stratospheric Ozone at Ground Level
4.4 SUMMARY
4. 5 REFERENCES
PROPERTIES, CHEMISTRY, AND MEASUREMENT OF OZONE AND OTHER
PHOTOCHEMICAL OXIDANTS
5.1 INTRODUCTION
5.2 PROPERTIES OF OZONE, PEROXYACETYL NITRATE, AND HYDROGEN
PEROXIDE
5.2.1 Ozone
5.2.2 Peroxyacetyl Nitrate
5. 2. 3 Hydrogen Peroxide
5.3 ATMOSPHERIC REACTIONS OF OZONE AND OTHER PHOTOCHEMICAL
OXIDANTS
5. 3. 1 Introduction
5.3.2 Atmospheric Reactions of Ozone with Organic
Compounds
5.3.2.1 Alkenes
5. 3. 2. 2 Al kanes and Al kynes
5.3.2.3 Aromatics
5.3.2.4 Oxygen-Containing Organics
5.3.2.5 Nitrogen-Containing Organics
5.3.2.6 Sulfur-Containing Organics
5. 3. 2. 7 Other Reactions
5.3.2.8 Atmospheric Lifetimes
5.3.2.9 Aerosol Formation
5.3.3 Atmospheric Reactions of Ozone with Inorganic
Compounds and wi th Li ght
5,3.4 Reactions of Ozone in Aqueous Droplets
5.3.5 Atmospheric Reactions of Peroxyacetyl Nitrate
(PAN)
5. 3.6 Atmospheric Reactions of Hydrogen Peroxide
5.3.7 Atmospheric Reactions of Formic Acid
5.4 REACTIONS OF OZONE AND PEROXYACETYL NITRATE IN SOLUTION
5.4. 1 Ozone
5.4. 1. 1 Al kenes
5.4.1.2 Amines
5. 4. 1. 3 Sul fur Compounds
5.4.1.4 Aromatics
5.4.1.5 Aldehydes and Ketones
. . 4-13
4-14
4-19
4-26
. . 4-27
4-31
4-31
4-37
. . 4-39
. . 4-43
5-1
5-1
5-1
5-1
5-2
5-3
5-8
5-8
5-9
5-9
5-13
5-14
5-14
5-15
5-17
5-17
5-17
5-17
5-19
5-20
5-22
5-24
5-25
5-25
5-26
5-26
5-28
5-29
5-30
5-31
vn
019FFM/A 6/30/84
-------
TABLE OF CONTENTS
(continued)
5.4.2 Peroxyacetyl Nitrate 5-32
5.4.2.1 Alkenes 5-33
5.4.2.2 Amines 5-33
5.4.2.3 Sulfur Compounds 5-33
5.4.2.4 Aldehydes 5-34
5.5 SAMPLING AND MEASUREMENT OF OZONE AND OTHER PHOTOCHEMICAL
OXIDANTS 5-35
5.5.1 Introduction 5-35
5.5.2 Quality Assurance in Ambient Air Monitoring for
Ozone 5-37
5.5.3 Sampling Factors in Ambient Air Monitoring for
Ozone 5-38
5.5.3.1 Sampling Strategies and Air Monitoring
Needs 5-39
5.5.3.2 Air Monitoring Site Selection 5-39
5.5.4 Measurement Methods for Total Oxidants and Ozone 5-41
5.5.4.1 Total Oxidants 5-41
5.5.4.2 Ozone 5-43
5.5.5 Generation and Calibration Methods for Ozone 5-49
5.5.5.1 Generation 5-49
5.5.5.2 Calibration 5-50
5.5.6 Relationship between Methods for Total Oxidants
and Ozone 5-59
5.5.6.1 Predicted Relationship 5-60
5.5.6.2 Empirical Relationship Determined from
Simultaneous Measurements 5-63
5.5.7 Methods for Sampling and Analysis of Peroxyacetyl
Nitrate and Its Homologues 5-73
5.5.7.1 Introduction 5-73
5.5.7.2 Analytical Methods 5-74
5.5.7.3 Generation and Calibration 5-80
5.5.7.4 Methods of Analysis of Higher Homologues 5-83
5.5.8 Methods for Sampling and Analysis of Hydrogen
Peroxide 5-84
5.5.8.1 Introduction 5-84
5.5.8.2 Sampling 5-84
5.5.8.3 Measurement 5-85
5.5.8.4 Generation and Calibration Methods 5-90
5.6 SUMMARY 5-91
5.6.1 Properties 5-91
5.6.2 Reactions of Ozone and Other Oxidants in Ambient
Air 5-92
5.6.3 Reactions of Ozone and Peroxyacetyl Nitrate
i n Aqueous (Bi ologi cal) Systems 5-93
viii
019FFM/A 6/30/84
-------
TABLE OF CONTENTS
(continued)
Page
5.6.4 Sampling and Measurement of Ozone and Other
Photochemical Oxidants 5-95
5.6.4.1 Quality Assurance and Sampling 5-95
5.6.4.2 Measurement Methods for Total Oxidants
and Ozone 5-96
5.6.4.3 Calibration Methods 5-99
5.6.4.4 Relationships of Total Oxidants and Ozone
Measurements 5-101
5.6.4.5 Methods for Sampling and Analysis of Pero-
xyacetyl Nitrate and Its Homologues 5-103
5.6.4.6 Methods for Sampling and Analysis of
Hydrogen Peroxi de 5-108
5.7 REFERENCES 5-113
6. CONCENTRATIONS OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
IN AMBIENT AIR 6-1
6.1 INTRODUCTION 6-1
6.2 HISTORICAL DATA ON 020NE/OXIDANT CONCENTRATIONS AND
TRENDS IN AMBIENT AIR 6-3
6.2.1 Summary of Urban Oxidant Data, 1964 through 1975 6-3
6.2.2 Summary of Rural and Remote Ozone Data, 1957
through 1975 6-4
6.2.3 Seasonal and Diurnal Variations in Ozone or
Oxidants Prior to 1970 6-11
6.2.4 Trends in Nationwide Ozone and Oxidant
Concentrations 6-16
6.3 OVERVIEW OF OZONE CONCENTRATIONS IN URBAN AREAS 6-18
6.4 OVERVIEW OF OZONE CONCENTRATIONS IN NONURBAN AREAS 6-26
6.4.1 National Air Pollution Background Network (NAPBN) 6-26
6.4.2 Sulfate Regional Experiment Sites (SURE) 6-29
6.5 VARIATIONS IN OZONE CONCENTRATIONS: DATA FROM SELECTED
URBAN AND NONURBAN SITES 6-34
6.5.1 Temporal Variations in Ozone Concentrations 6-34
6.5.1.1 Diurnal Variations in Ozone Concentrations ... 6-34
6.5.1.2 Seasonal Variations in Ozone Concentrations .. 6-50
6.5.1.3 Weekday-Weekend Variations in Ozone
Concentrations 6-54
6.5.2 Spatial Variations in Ozone Concentrations 6-57
6.5.2.1 Urban Versus Nonurban Variations 6-57
6.5.2.2 Intracity Variations 6-61
6.5.2.3 Indoor-Outdoor Ozone Concentration Ratios .... 6-64
6.5.2.4 Macroscale Variations in Ozone
Concentrations: Effects of Altitude and
Latitude 6-69
6.5.2.5 Microscale Variations in Ozone
Concentrations: Effects of Monitor
PI acement 6-70
IX
019FFM/A 6/30/84
-------
TABLE OF CONTENTS
(continued)
6.6 CONCENTRATIONS OF PEROXYACETYL NITRATE (PAN) AND
PEROXYPROPIONYL NITRATE (PPN) IN AMBIENT AIR 6-71
6.6.1 Introduction 6-71
6.6.2 Historical Data 6-74
6.6.3 Ambient Air Concentrations of PAN and Its
Homo!ogues i n Urban Areas 6-76
6.6.4 Ambient Air Concentrations of PAN and Its
Homologues in Nonurban Areas 6-82
6.6.5 Temporal Variations in Ambient Air
Concentrations of Peroxyacetyl Nitrate 6-85
6.6.5.1 Diurnal Patterns 6-85
6.6.5.2 Seasonal Patterns 6-89
6.6.6 Spatial Variations in Ambient Air Concentrations
of Peroxyacetyl Ni trate 6-91
6.6.6.1 Urban-Rural Gradients and Transport of PAN ... 6-91
6.6.6.2 Intracity Variations 6-93
6.6.6.3 Indoor-Outdoor Ratios of PAN Concentrations .. 6-98
6.7 CONCENTRATIONS OF OTHER PHOTOCHEMICAL OXIDANTS IN
AMBIENT AIR 6-98
6.8 SUMMARY 6-99
6.8.1 Ozone Concentrations in Urban Areas 6-102
6.8.2 Trends in Urban and Nationwide Ozone
Concentrations 6-105
6.8.3 Ozone Concentrations in Nonurban Areas 6-106
6.8.4 Patterns in Ozone Concentrations 6-107
6.8.5 Concentrations and Patterns of Other Photochemical
Oxidants 6-109
6.8.5.1 Concentrations 6-109
6.8.5.2 Patterns 6-111
6.8.6 Relationship Between Ozone and Other Photochemical
Oxidants 6-112
6.9 REFERENCES 6-115
019FFM/A 6/30/84
-------
LIST OF TABLES
Table Page
3-1 Physical and chemical properties of nitric oxide and
nitrogen dioxide 3-8
3-2 Percentage difference from known concentrations of
nonmethane hydrocarbons obtained by sixteen users 3-11
3-3 Summary of problems associated with gathering NMOC data
by means of automated analyzers 3-12
3-4 Summary of recommendations to reduce the effects of problems
listed in Table 3-3 3-13
3-5 Identification key for typical heavy hydrocarbon
chromatogram, C4 to C10 (university campus site,
Cincinnati, Ohio, August 24, 1981) 3-18
3-6 Summary of advantages and disadvantages of primary
collection media for NMOC analysis 3-21
3-7 GC/continuous NMOC analyzer comparisons, least-squares
regressions 3-25
3-8 Methods for measuring nitrogen dioxide 3-31
3-9 National estimates of volatile organic compound emissions,
1982 3-50
3-10 Hydrocarbon exhaust emission factors for light-duty,
gasoline-powered vehicles for all areas except California
and high-altitude 3-52
3-11 Predominant hydrocarbons in exhaust emissions from
gasoline-fueled autos 3-53
3-12 Summary of emission chracteristies for autos fueled by
gasoline, diesel, and alcohol-gasoline or ether-gasoline
blends 3-56
3-13 National estimates of emissions of nitrogen oxides, 1982 3-58
3-14 NO/NO ratios in emissions from various types of sources 3-59
3-15 Isoprene emission rates 3-65
3-16 Monoterpene emission rates 3-66
3-17 Forest survey data for isoprene-emitting hardwoods 3-72
3-18 Area-wide biogenic emission fluxes 3-74
3-19 Global estimates of nitrogen transformation 3-78
3-20 Nonmethane hydrocarbon concentrations measured between
6:00 and 9:00 a.m. in various United States cities 3-84
3-21 Hydrocarbon composition typically measured in urban areas
(from sample collected in Milwaukee, 1981 3-85
3-22 Nonmethane hydrocarbon concentrations measured in nonurban
atmospheres 3-86
3-23 Nonmethane hydrocarbon concentrations in samples collected
aloft (1000 to 5000 ft) during morning hours (6:00 to
10: 00 a. m. ) 3-88
3-24 Formaldehyde concentrations in several United States cities 3-90
3-25 Average 6:00 to 9:00 a.m. NO concentrations and HC/NO
ratios i n urban areas * 3-94
XI
019FFM/A 6/30/84
-------
LIST OF TABLES
(continued)
Table page
4-1 Documented episodes of transport of stratospheric
ozone to ground 1 eve! 4-38
5-1 Physical properties of ozone 5-3
5-2 Physical properties of peroxyacetyl nitrate 5-4
5-3 Infrared absorptivities of peroxyacetyl nitrate at
approximate resolution of 1.2 cm 5-5
5-4 Physical properties of hydrogen peroxide 5-7
5-5 Normal electrode potentials of some hydrogen peroxide-
containing oxidation-reduction systems of biological
importance 5-8
5-6 Calculated lifetimes of selected organics resulting from
atmospheric loss by reaction with 03 and with OH and N03
radi cal s 5-18
5-7 Performance specifications for automated methods 5-46
5-8 List of designated reference and equivalent methods 5-47
5-9 Factors for intercomparison of data calibrated by UV
photometry versus KI colorimetry 5-53
5-10 Response of NBKI reagent and Mast meter to various oxidants 5-62
5-11 Comparison of corrected instrument readings to
colorimetric oxidant readings during atmospheric sampling 5-67
5-12 Summary of parameters used in determination of PAN by
GC-ECD 5-75
5-13 PAN infrared absorptivities 5-78
5-14 Summary of ozone monitoring techniques 5-98
5-15 Ozone calibration techniques 5-100
5-16 Summary of parameters used in determination of PAN by
GC-ECD 5-105
5-17 Infrared absorptivities of peroxyacetyl nitrate 5-106
5-18 Measurement methods for hydrogen peroxide 5-110
6-1 Summary of maximum oxidant concentrations recorded in
selected cities, 1964-1967 6-5
6-2 Oxidant concentrations observed in selected urban areas
of the United States, 1974-1975 6-6
6-3 Summary of oxidant concentrations in ambient air at rural
and remote sites, 1957 through 1967 6-7
6-4 Concentrations of tropospheric ozone before 1962 6-9
6-5 Summary of ozone data from Research Triangle Institute
studies, 1973 through 1975 6-10
6-6 Second-highest 1-hour ozone concentrations reported for
80 Standard Metropolitan Statistical Areas having
populations > 0.5 million 6-23
6-7 Annual ozone summary statistics for three NAPBN sites 6-28
6-8 Concentrations of ozone during 6-day period of high
values at NAPBN site in Mark Twain National Forest,
Missouri, 1979 6-30
xn
019FFM/A 6/30/84
-------
LIST OF TABLES
(continued)
Table Page
6-9 Summary of ozone concentrations measured at Sulfate
Regional Experiment (SURE) nonurban stations, August
through December 1977 6-33
6-10 Total days when maximum daily ozone concentration
exceeded or was less than specified concentrations,
April through September, 1979 through 1981, at Pasadena
and Pomona, California, and at Washington, DC and
Dallas, Texas 6-45
6-11 Ozone concentrations at sites in and aound New Haven,
Connecticut, 1976 6-62
6-12 Quarterly maximum 1-hour ozone values at sites in and
around New Haven, Connecticut, 1976 6-63
6-13 Peak ozone concentrations at eight sites in New York
City and adjacent Nassau County, 1980 6-65
6-14 Summary of reported indoor-outdoor ozone ratios 6-68
6-15 Means and standard errors of ozone concentrations measured
over 4 years at two sampling heights at three stations in
the rural, upper-midwestern United States 6-72
6-16 Summary of concentrations of peroxyacetyl nitrate in
ambient air in urban areas of the United States 6-77
6-17 Relationship of ozone and peroxyacetyl nitrate at urban
and suburban sites in the United States 6-81
6-18 Ambient air measurements of peroxypropionyl nitrate
concentrations by electron capture gas chromatography
at urban sites in the United States 6-83
6-19 Concentrations of peroxyacetyl and peroxypropionyl nitrates
in Los Angeles, Oakland, and Phoenix, 1979 6-84
6-20 Concentrations in ambient air of peroxyacetyl and
peroxypropionyl nitrates and ozone at nonurban remote sites
in the United States 6-86
6-21 PAN and ozone concentrations in ambient air, New
Brunswick, NJ, for September 25, 1978, to May 10, 1980 6-93
6-22 Concentrations of hydrogen peroxide in ambient air at
urban and nonurban sites 6-101
6-23 Second-highest 1-hr ozone concentrations in 1982 in Standard
Metropolitan Statistical Areas with populations > million,
given by Census divisions and regions 6-103
xm
019FFM/A 6/30/84
-------
LIST OF FIGURES
Page
2-1 Chemical changes occurrinr during photo-irradiation of
hydrocarbon-nitrogen oxide-air systems 2-3
3-1 Light hydrocarbon chroraatocjram, C2 to C5 (frora the university
campus site, Cincinnati, Crno. August 24, 1981) 3-16
3-2 Heavy hydrocarbon chroraatogram, C4 to C10 (from University
campus site, Cincinnati, Ohio, August 24, 1981) 3-17
3-3 National trend in estimated emissions of volatile organic
compounds, 1970 through 1982 3-40
3-4 National trena in estimated emissions of nitrogen oxides,
1970 through 19S2 ....... 3-41
3-5 Comparative trends in mobile source emissions of nitrogen
oxides (NO } and volatile organic compounds (VOC) versus
vehicl e mi ies traveled, 1970 through 1981 3-43
3-6 Total volatile organic compound emissions by county in
the cotermi nous tJni ted States, 1978 3-44
3-7 Area source volatile organic compound emissions by county
i n the cotermi nous Uni ted States, 1978 3-45
3-8 Total NO emissions By county in the coterminous United
States, 1978 * . . , 3-46
3-9 Area source NO emissions t/y county in the coterminous
United States,xl&78 .......'.......'........ 3-47
3-10 Total nonmethnne huii-ocarbcn emission rates as a function
of temperature for isoprene-emitting hardwoods 3-68
3-11 Estimated diurnal cycie of isoprene and monoterpene
emission rates .................... 3-69
3-12 Comparison of four aV'ometnc relationships for
detenu i nati on of 1 ea f bi oriAts 3-73
3-13 The nitrogen cycle 3-76
4-1 Reaction scheme for the HO'-initiated oxidation of
2-butene-NO system 4-7
4-2 Isopleths (m x I02) of m&a.r, sjramer corning mixing heights 4-18
4-3 Isopleths (m x 1C2) of mean summer afternoon mixing heights 4-18
4-4 Percentage of summer 2315 GMT (6:15 p.ru EST, 3:15 p.m.
PST) soundings with an elevated inversion base between 1
and 500 m above ground 1 eve's . 4-20
4-5 Mean resultant surface wind pattern for the United
States for July 4-22
4-6 Percentage of summer 1115 GMT (6:1.5 a.m. EST, 3:15 a.m.
PST) soundings with an inversion base at the surface
and wind speeds at the surface _< 2.5 m/sec 4-23
4-7 Isopleths (m/sec) ot mean summer wind speed averaged
through the morni ng mixi ng 1 ayer 4-25
4-8 Isopleths (m/sec) of mean summer wind speed averaged
through the afternoon rrixl r,g 1 ay&r 4-25
xi v
019FFM/A 6/30/84
-------
LIST OF FIGURES
(continued)
Page
4-9 Schematic cross section, looking downwind along the jet
stream, of a tropopause folding event as modeled by
Daniel sen (1968) 4-33
4-10 Measured vertical cross sections of 03, dewpoint, and
the 500 mb chart and the flight track for October 5,
1978 4-34
4-11 Hypothesized models of the process that mixes tropopause
folding events into the troposphere 4-36
5-1 Top: IR gas spectrum of pure PAN (optical path length
10 cm); (A) 1.5 torr, (B) 10.0 torr. Bottom: Raman
spectrum of PAN at -40°C (liquid); excitation light,
514.5 nm (100 mW) 5-6
5-2 Rate of aqueous-phase oxidation of S(IV) by 03 (30 ppb)
and H202 (1 ppb), as a function of solution pH 5-21
5-3 Ozone and oxidant concentration in the Pasadena area,
August 1955 5-65
5-4 Ozone and oxidant concentration in the Los Angeles area 5-65
5-5 Measurements for ozone and oxidants in Los Angeles 5-69
5-6 Measurements for ozone and oxidants in St. Louis 5-70
5-7 Measurement of ozone and oxidants, Houston Ship Channel,
August 11, 1973 5-72
6-1 Long-term monthly ozone variations at Quillayute,
Washington 6-12
6-2 Long-term monthly ozone variations at Mauna Loa, Hawaii 6-12
6-3 Monthly variation of mean hourly oxidant concentrations for
Los Angeles and Denver , 6-14
6-4 Average monthly ozone concentrations recorded at Whiteface
Mountain in New York 6-14
6-5 Diurnal variation of hourly oxidant concentrations in
Philadelphia and Denver 6-15
6-6 National trend in the composite average of the second-
highest daily 1-hour concentration, 1975 through 1981 6-17
6-7 Average daylight (6:00 a.m. to 8:00 p.m.) concentrations
of ozone in the second and third quarters (April through
September) , 1981 6-19
6-8 Average daylight (6:00 a.m. to 8:00 p.m.) concentrations
of ozone in the first and fourth quarters (January through
March and October through December), 1981 6-20
6-9 Collective distributions of the three highest 1-hour 03
concentrations at valid sites for 1979, 1980, and 1981
(906 station-years) 6-22
xv
019FFM/A 6/30/84
-------
LIST OF FIGURES
(continued)
Figure page
6-10 Locations of the eight national forest (NF) stations
comprising the National Air Pollution Background Network
(NAPBN) 6-27
6-11 Trajectory analysis plots at the NAPBN site at Mark Twain
National Forest, MO, July 21, 1979 6-31
6-12 Location of SURE monitoring stations 6-32
6-13 Diurnal pattern of 1-hour ozone concentrations on July 13,
1979, Philadelphia, PA 6-35
6-14 Diurnal patterns of ozone concentrations, September 20
and 21, 1980, Detroit, MI 6-37
6-15 Diurnal and 1-month composite diurnal variations in ozone
concentrations, Washington, DC, July 1981 6-38
6-16 Diurnal and 1-month composite diurnal variations in ozone
concentrations, St. Louis County, MO, September 1981 6-38
6-17 Diurnal and 1-month composite diurnal variations in ozone
concentrations, Alton, IL, October 1981 (fourth quarter) 6-39
6-18 Diurnal and 1-month composite diurnal variations in
ozone concentrations, N. Little Rock, AR, November 1981
(fourth quarter) 6-39
6-19 Composite diurnal patterns by quarter of ozone
concentrations at a rural agricultural site, Alton, IL,
1981 6-40
6-20 Composite diurnal patterns by quarter of ozone
concentrations at a rural agricultural site, N. Little
Rock, AR, 1981 6-40
6-21 Three-day sequence of hourly ozone concentrations at
Montague, MA, SURE station showing locally generated
midday peaks and transported late peaks 6-42
6-22 Composite diurnal ozone pattern at an Argonne, IL, agricultural
site, August 6 through September 30, 1980 6-44
6-23 Probability that "exposures" and "respites" for specified
concentration cutoffs will persist for indicated or longer
period at Pasadena, CA, based on aerometric data for April
through September, 1979 through 1981 , 6-46
6-24 Probability that "exposures" and "respites" for specified
concentration cutoffs will persist for indicated or longer
period at Pomona, CA, based on aerometric data for April
through September, 1979 through 1981 6-47
6-25 Probability that "exposures" and "respites" for specified
concentration cutoffs will persist for indicated or longer
period at Washington, DC, based on aerometric data for April
through September, 1979 through 1981 6-48
6-26 Probability that "exposures" and "respites" for specified
concentration cutoffs will persist for indicated or longer
period at Dallas, TX, based on aerometric data for April
through September, 1979 through 1981 6-49
xvi
019FFM/A 6/30/84
-------
LIST OF FIGURES
(continued)
Page
6-27 Quarterly composite diurnal patterns of ozone
concentrations at selected sites representing potential for
exposure of major crops, 1981 6-52
6-28 Daily 7-hour and 24-hour average ozone concentrations at
a rural (NCLAN) site in Argonne, IL, 1980 6-53
6-29 Seasonal variations in ozone concentrations as indicated
by monthly averages and the 1-hour maximum in each month
at selected sites, 1981 6-55
6-30 Composite diurnal data for Sunday versus other six days for
July through September 1981, Pomona, CA 6-58
6-31 Composite diurnal data for Sunday versus other six days for
July through September 1981, Lennox, CA 6-59
6-32 Composite diurnal data for Sunday versus other six days for
July through September 1981, Little Rock, AR 6-60
6-33 New York State air monitoring sites for Northeast Corridor
Monitoring Program (NECRMP) 6-66
6-34 Comparison of monthly daylight average and maximum PAN
concentrations at Riverside, CA, for 1967-1968 and 1980 6-80
6-35 Variation of mean 1-hour oxidant and PAN concentrations,
by hour of day, in downtown Los Angeles, 1965 6-87
6-36 Variation of mean 1-hour average oxidant and PAN
concentrations, by hour of day, Air Pollution
Research Center, Riverside, CA, September 1966 6-88
6-37 Diurnal profiles of ozone and PAN at Claremont, CA,
October 12 and 13, 1978, 2 days of a multi-day smog episode 6-90
6-38 Monthly variation of oxidant and PAN concentrations,
Air Pollution Research Center, Riverside, CA,
June 1966-June 1967 6-92
6-39 Average daily profile by month (July 7-October 10) for
PAN and 03 in New Brunswick, NJ, 1979 6-94
6-40 Diurnal plot of PAN and oxidant concentrations at site
just north of Houston, October 26-27, 1977 6-95
6-41 Diurnal plot of PAN and oxidant concentrations at site in
Houston, near junction of 1-10 and 1-45, October 26-27,
1977 „ 6-96
6-42 Diurnal plot of PAN and oxidant concentrations at site in
southeast Houston, October 26-27, 1977 6-97
6-43 Diurnal profile of HCOOH, along with other oxidants and
smog constituents, on October 12 and 13, 1978, at
Claremont, CA 6-100
xvi i
019FFM/A 6/30/84
-------
LIST OF ABBREVIATIONS AND SYMBOLS
AFR
APHA
aq
ASL
atm
avg
b.p.
bz
C
°C
CA
CAMP
CARB
cc
CH,
CO
co2
cm
concn
DBH
DNPH
DOT
E6
ECD
EKMA
EPA
FID
FRM
ft
FTIR
9
approximately
wavelength
air:fuel ratio
American Public Health Association
aqueous
above sea level
atmosphere
average
boiling point
benzene
carbon
degrees Celsius
chromotropic acid
Continuous Air Monitoring Program
California Air Resources Board
cubic centimeter
methane
carbon monoxide
carbon dioxide
centimeter
concentration
tree diameter at breast height
2,4-dinitrophenylhydrazine
Department of Transportation
normal electrode potential
electron-capture detector
Empirical Kinetic Modeling Approach
U.S. Environmental Protection Agency
flame ionization detector
Federal Reference Method
foot
Fourier-transform infrared
gram(s)
019FFM/A
xvm
6/30/84
-------
LIST OF ABBREVIATIONS AND SYMBOLS (continued)
g/mi
GC
GPT
hr
hv
HC
HCN
HCOOH
HFET
Hg
H2°2
H02
HONO
HONO£
HPLC
HPPA
HRP
in
IR
k
KI
km
L
LAAPCD
LCV
In
LST
M
m
mb
grams per mile
gas chromatography
gas-phase titration
hour(s)
photon
hydrocarbons
hydrogen cyanide
formic acid
Highway Fuel Economy Driving Schedule
mercury
hydrogen peroxide
hydroperoxy
nitrous acid
nitric acid
high-pressure liquid chromatography; also,
high-performance liquid chromatography
3-(£-hydroxyphenyl)propionic acid
horseradish peroxidase
water
sulfuric acid
inch(es)
i nfrared
constant
potassium iodide
ki 1ometer
liter(s)
Los Angeles Air Pollution Control District
leuco crystal violet
natural logarithm (base e)
local standard time
molar
meter(s)
millibar(s)
019FFM/A
xix
6/30/84
-------
LIST OF ABBREVIATIONS AND SYMBOLS (continued)
MBTH
mg
mg/m
MGE
min
ml
mm
mM
MMC
m. p.
mph
MS
MSL
MT
MTBE
NA
NAAQS
NADB
NAMS
NAPBN
NAS
NBS
NECRMP
NEDS
NEROS
NH3
NH4N03
NF
nm
NMHC
NMOC
NO
NO
3-methyl-2-benzothiazolinone hydrazone
milligram(s)
milligrams per cubic meter
modified graphite electrode
minute(s)
milliliter(s)
millimeter(s)
millimolar
mean meridional circulation
melting point
miles per hour
mass spectrometry
mean sea level
metric tons
methyl tertiary butyl ether
Not available
National Ambient Air Quality Standard
National Aerometric Data Bank
National Aerometric Monitoring Stations
National Air Pollution Background Network
National Academy of Sciences
National Bureau of Standards
Northeast Corridor Regional Modeling Project
National Emissions Data System
Northeast Regional Oxidant Study
ammonia
ammonium nitrate
National Forest
nanometer
nonmethane hydrocarbons
nonmethane organic compounds
nitric oxide
nitrogen oxides
019FFM/A
xx
6/30/84
-------
LIST OF ABBREVIATIONS AND SYMBOLS (continued)
N02
N03
N20
NR
NYCC
°2
°3
PAN
PBzN
pH
PNA
PPN
ppb
ppm
ppt
PSD
psig
PST
PUFA
RAPS
RTI
S.D.
SAROAD
SBR
SCAB
sec
SLAMS
SMSA
SRM
SSET
STA
STP
SURE
nitrogen dioxide
nitrogen trioxide
nitrous oxide
natural rubber
New York City Driving Schedule
oxygen
ozone
peroxyacetyl nitrate
peroxybenzoyl nitrate
reciprocal of H ion concentration
peroxynitric acid
peroxypropionyl nitrate
parts per billion
parts per million
parts per trillion
Prevention of Significant Deterioration
pounds per square inch gauge
Pacific Standard Time
polyunsaturated fatty acids
Regional Air Pollution Study
Research Triangle Institute
standard deviation
Storage and Retrieval of Aerometric Data
styrene-butadiene rubber
South Coast Air Basin
second(s)
State and Local Air Monitoring Stations
Standard Metropolitan Statistical Area
Standard Reference Material
small-scale eddy transport
seasonal tropopause adjustment
standard temperature and pressure
Sulfate Regional Experiment Sites
019FFM/A
xxi
6/30/84
-------
LIST OF ABBREVIATIONS AND SYMBOLS (continued)
TEL
Tenax GC
TF
tg/yr
THC
TML
TNMHC
TWC
MM
U
UHAC
U.S.
UV
V
v/v
VHAC
VOC
vol %
w/w
WCOT
XAD-2
XO
tetraethyl lead
adsorbent used in NMOC analysis
tropopause-folding events
teragrams per year
total hydrocarbon
tetramethyl 1ead
total nonmethane hydrocarbons
three-way catalyst
microgram per cubic meter
micromolar
uranium
uranium hydroxamic acid chelates
United States
ultraviolet
vanadium
volume - volume
vanadium hydroxamic acid chelates
volatile organic compounds
volume percent
weight - weight
wall-coated open tubular (column)
absorbent used in NMOC analysis
xylenol orange
year(s)
019FFM/A
xxii
6/30/84
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Chapter 3: Precursors to Ozone and Other Photochemical Oxidants
Principal Authors
Mr. Michael W. Holdren
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. Thomas B. McMullen
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Halvor Westberg
Director, Laboratory for Atmospheric Research, and
Professor, Civil and Environmental Engineering
Washington State University
Pullman, WA 99164-2730
Contributing Author
Mr. George M. Duggan
Office of Air Quality Planning and Standards
Strategies and Air Standards Division
MD-12
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
The following people reviewed Chapter 3 at the request of EPA:
Dr. A. Paul Altshuller
Environmental Sciences Research Laboratory
MD-59
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Joseph J. Bufalini
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxiii
019FFM/A 6/30/84
-------
Chapter 3 Reviewers (cont'd):
Dr. Marcia C. Dodge
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Donald Fox
Associate Professor
Department of Environmental Science and Engineering
School of Public Health, 201H
University of North Carolina
Chapel Hill, NC 27514
Mr. Bruce Gay
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Gerald Gipson
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Robert Hall
Industrial Environmental Research Laboratory
MD-65
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Jimmie A. Hodgeson
Professor, Department of Chemistry
407 Choppin Hall
Louisiana State University
Baton Rouge, LA 70803
*Mr Michael W. Holdren
BatteHe, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. Michael R. Kuhlman
BatteHe, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. William A. Lonneman
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxiv
019FFM/A 6/30/84
-------
Chapter 3 Reviewers (cont'd):
Mr. Chuck Mann
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Mr. Thomas B. McMullen
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Edwin L. Meyer
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Kenneth Rehme
Environmental Monitoring Systems Laboratory
MD-77
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Harold G. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Elmer Robinson
Professor of Civil Engineering and Research Meteorologist
Laboratory for Atmospheric Research
Washington State University
Pullman, WA 99164
(Present address:
Director, Mauna Loa Observatory
National Oceanic and Atmospheric Administration (NOAA/CMCC)
Hilo, HI)
Dr. Chester W. Spicer
BatteHe, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. Bruce Tichenor
Industrial Environmental Research Laboratory
MD-54
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxv
019FFM/A 6/30/84
-------
Chapter 3 Reviewers (cont'd):
*Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
*Authors also reviewed portions of this chapter.
xxvi
019FFM/A 6/30/84
-------
Chapter 4: Chemical and Physical Processes in the Formation and Occurrence of
Ozone and Other Photochemical Oxidants
Principal Authors
Dr. Harold G. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Elmer Robinson
Professor of Civil Engineering and Research Meteorologist
Laboratory for Atmospheric Research
Washington State University
Pullman, WA 99164
(Present address:
Director, Mauna Loa Observatory
National Oceanic and Atmospheric Administration (NOAA/CMCC)
Hilo, HI)
Contributing Authors
Dr. Marcia C. Dodge
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
The following people reviewed Chapter 4 at the request of EPA:
Dr. Joseph J. Bufalini
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Dr. Marcia C. Dodge
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxvn
019FFM/A 6/30/84
-------
Chapter 4 Reviewers (cont'd):
Dr. Donald Fox
Associate Professor
Department of Environmental Science and Engineering
School of Public Health, 201H
University of North Carolina
Chapel Hill, NC 27514
Mr. Gerald Gipson
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Michael W. Holdren
BatteHe, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. Michael R. Kuhlman
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. E. L. Martinez
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Thomas B. McMullen
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Edwin L. Meyer
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Dr. Harold G. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxvm
019FFM/A 6/30/84
-------
Chapter 4 Reviewers (cont'd):
*Mr. Elmer Robinson
Professor of Civil Engineering and Research Meteorologist
Laboratory for Atmospheric Research
Washington State University
Pullman, WA 99164
(Present address:
Director, Mauna Loa Observatory
National Oceanic and Atmospheric Administration (NOAA/CMCC)
Hilo, HI)
Mr. Kenneth L. Schere
Environmental Sciences Research Laboratory
MD-80
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Stanley Sleva
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Chester W. Spicer
BatteHe, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Ms. Beverly E. Til ton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
*Authors also reviewed portions of this chapter.
xxix
019FFM/A 6/30/84
-------
Chapter 5: Properties, Chemistry, and Measurement of Ozone and Other
Photochemical Oxidants
Principal Authors
Dr. Margaret M. Dooley
Associate Professor, Department of Chemistry
Louisiana State University
Baton Rouge, LA 70803
*Dr. Jimmie A. Hodgeson
Professor, Department of Chemistry
407 Choppin Hall
Louisiana State University
Baton Rouge, LA 70803
Dr. William A. Pryor
Chairman and Professor, Department of Chemistry
Louisiana State University
Baton Rouge, LA 70803
Dr. M. Rene Surgi
Department of Chemistry
Louisiana State University
Baton Rouge, LA 70803
Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
Contributing Author
Mr. James M. Kawecki
TRC Environmental Consultants, Inc.
701 W. Broad Street
Falls Church, VA 22046
The following people reviewed Chapter 5 at the request of EPA:
Dr. A. Paul Altshuller
Environmental Sciences Research Laboratory
MD-59
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxx
019FFM/A 6/30/84
-------
Chapter 5 Reviewers (cont'd):
Dr. Joseph J. Bufalini
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Marcia C. Dodge
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Donald Fox
Associate Professor
Department of Environmental Science and Engineering
School of Public Health, 201H
University of North Carolina
Chapel Hill, NC 27514
Mr. Bruce Gay
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Michael W. Holdren
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. Michael R. Kuhlman
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. William A. Lonneman
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Kenneth Rehme
Environmental Monitoring Systems Laboratory
MD-77
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Harold G. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxxi
019FFM/A 6/30/84
-------
Chapter 5 Reviewers (cont'd):
Mr. Stanley Sleva
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Chester W. Spicer
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Ms. Beverly E. Til ton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
^Authors also reviewed portions of this chapter.
xxxn
019FFM/A 6/30/84
-------
Chapter 6: Concentrations of Ozone and Other Photochemical Oxidants in Ambient
Air
Principal Authors
Mr. Thomas B. McMullen
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Elmer Robinson
Professor of Civil Engineering and Research Meteorologist
Laboratory for Atmospheric Research
Washington State University
Pullman, WA 99164
(Present address:
Director, Mauna Loa Observatory
National Oceanic and Atmospheric Administration (NOAA/CMCC)
Hilo, HI)
Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Contributing Authors
Mr. George M. Duggan
Office of Air Quality Planning and Standards
Strategies and Air Standards Division
MD-12
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Sandor Freedman
Piedmont Technical Services
Hillsborough, NC 27278
The following people reviewed Chapter 6 at the request of EPA:
Mr. Gerald Akland
Environmental Monitoring Systems Laboratory
MD-56
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxxi i i
019FFM/A 6/30/84
-------
Chapter 6 Reviewers (cont'd):
Dr. Joseph J. Bufalini
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Gary Evans
Environmental Monitoring Systems Laboratory
MD-56
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Donald Fox
Associate Professor
Department of Environmental Science and Engineering
School of Public Health, 201H
University of North Carolina
Chapel Hill, NC 27514
Mr. Gerald Gipson
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Jiramie A. Hodgeson
Professor, Department of Chemistry
407 Choppin Hall
Louisiana State University
Baton Rouge, LA 70803
Mr Michael W. Holdran
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. Michael R. Kuhlman
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. William A. Lonneman
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Thomas McCurdy
Office of Air Quality Planning and Standards
Strategies and Air Standards Division
MD-12
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxx iv
019FFM/A 6/30/84
-------
Chapter 6 Reviewers (cont'd):
Mr. Thomas B. McMullen
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Edwin L. Meyer
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Harold G. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Mr. Elmer Robinson
Professor of Civil Engineering and Research Meteorologist
Laboratory for Atmospheric Research
Washington State University
Pullman, WA 99164
(Present address:
Director, Mauna Loa Observatory
National Oceanic and Atmospheric Administration (NOAA/CMCC)
Hilo, HI)
Dr. Chester W. Spicer
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
^Authors also reviewed portions of this chapter.
xxxv
019FFM/A 6/30/84
-------
2. INTRODUCTION
2.1 PURPOSE AND LEGISLATIVE BASIS OF DOCUMENT
According to the Clean Air Act, the Administrator of the United States
Environmental Protection Agency (EPA) is required to issue, and to revise on a
periodic basis, air quality criteria for certain air pollutants specified in
the Act. Among these are pollutants known as photochemical oxidants. The
term "photochemical oxidants" has historically been defined as those atmospheric
pollutants that are products of photochemical reactions and that are capable
of oxidizing neutral iodide ions (U.S. Environmental Protection Agency, 1978).
Research has unequivocally established that photochemical oxidants in ambient
air consist mainly of ozone, peroxyacetyl nitrate, and nitrogen dioxide, and
of considerably lesser amounts of other peroxyacyl nitrates, hydrogen peroxide,
alkyl hydroperoxides, and nitric and nitrous acids. Other oxidants suspected
to occur in ambient air in trace amounts include peracids and ozonides.
Although it is by definition a photochemical oxidant, nitrogen dioxide is
not included among the oxidants discussed in this document. The formation of
nitrogen dioxide clearly precedes the formation of ozone and related other
oxidants in the ambient air. While nitrogen dioxide is the dominant oxidant
early in the day, ozone and related other oxidants predominate from late
morning or midday through much of the afternoon.
In addition to the differences in patterns of occurrence of nitrogen
dioxide versus ozone and its related oxidants, nitrogen dioxide is known to
exert deleterious effects on human health and welfare. The Clean Air Act
specifies, therefore, that criteria be issued separately for nitrogen dioxide
and other oxides of nitrogen. The second criteria document prepared by EPA on
the oxides of nitrogen was published in 1982 (U.S. Environmental Protection
Agency, 1982a). That document discussed nitric and nitrous oxides, nitrogen
dioxide, nitric and nitrous acid, and nitrosamines.
The purpose of this document is to review and evaluate the scientific
literature on ozone and related oxidants and to document their effects on
public health and welfare. Such documentation provides the Agency with a
scientific basis for deciding whether regulations controlling these pollutants
are necessary and for deriving such ambient air quality standards as may be
needed.
OZONER/B 2-1 6/22/84
-------
According to section 108 of the Clean Air Act, as amended in 1977, a
criteria document shall
...accurately reflect the latest scientific knowledge useful in indicating
the kind and extent of all identifiable effects on public health or
welfare which may be expected from the presence of such pollutant in the
ambient air, in varying quantities.
(Clean Air Act, U.S.C. §§7408 and 7409)
Air quality criteria may be defined, then, as qualitative and quantitative
information that describes the effects of a pollutant on public health and
welfare in terms of the respective exposures that elicited them.
This document is a revision of Air Quality Criteria for Ozone and Other
Photochemical Oxidants (U.S. Environmental Protection Agency, 1978), as spe-
cified in sections 108 and 109 of the Clean Air Act. As used in this document,
the term "photochemical oxidants" refers to ozone, the peroxyacyl nitrates,
and hydrogen peroxide. The oxides of nitrogen are discussed, but only in the
context of their role as precursors to ozone and related oxidants.
2.2 THE OXIDANT PROBLEM
As described in the Clean Air Act, criteria pollutants are those atmos-
pheric pollutants that are ubiquitous and are emitted into the air from numerous
and diverse sources. While ubiquitous, ozone and other photochemical oxidants
are not emitted into the air as primary pollutants. Rather, they are formed
as secondary pollutants in the atmosphere from ubiquitous primary organic and
inorganic precursors that are emitted by a multiplicity of sources. Oxidant
pollution is widespread in this country as the result of a combination of many
factors, such as local meteorological conditions as well as the concentrations,
composition, and patterns of occurrence of the primary pollutants that give
rise to the oxidants.
An overview of the relationships among nitrogen oxides, volatile organic
compounds such as hydrocarbons, and the respective photochemical ocidants is
helpful for understanding material presented in later chapters. Figure 2-1
depicts the idealized patterns of the respective primary and secondary pollut-
ants involved in atmospheric photochemical processes. Although derived from
laboratory data, the pattern depicted in Figure 2-1 nevertheless represents
the general pattern seen in ambient air, as field research has corroborated.
OZONER/B 2-2 6/22/84
-------
Z
O
g
K
UJ
O
z
O
O
HYDROCARBON
NITROGEN DIOXIDE
IRRADIATION TIME
Figure 2-1. Chemical changes occurring during photoirradiation of
hydrocarbon-nitrogen oxide-air systems.
Source: U.S. Environmental Protection Agency (1978).
2-3
-------
Ozone is the photochemical oxidant currently regulated by national ambient
air quality standards. It is well established that ozone produces effects on
public health and welfare that are attributable to its characteristics as an
oxidant. The bulk of health- and welfare-related research has centered on
ozone, largely because it is the major photochemical oxidant found in ambient
air. While data on the health and welfare effects of hydrogen peroxide and of
the peroxyacyl nitrates are few, sufficient data on their effects and on their
concentrations exist to be indicative of their potential, at high enough
concentrations, for affecting public health and welfare. As subsequent chapters
will document, the peroxyacyl nitrates appear to be ubiquitous. The available
data on hydrogen peroxide are considerably fewer; from information on atmos-
pheric photochemistry, however, hydrogen peroxide would be expected to occur,
at least in trace amounts, in those atmospheres in which ozone and the peroxy-
acyl nitrates are found.
2.3 SCOPE AND ORGANIZATION OF THIS DOCUMENT
The atmosphere does not easily lend itself to the partitioning required
for documentation. Nevertheless, certain boundaries are logical for purposes
of discussion as well as for purposes of regulatory decisions. Ozone and its
organic precursors are known to give rise to secondary organic aerosols (see
chapter 5). Likewise, ozone and hydrogen peroxide both appear to participate
in those atmospheric oxidations of nitrogen dioxide (N0_) and sulfur dioxide
(S0~) that lead to visibility degradation in the atmosphere and to the environ-
mental phenomenon known as acidic deposition. The contributions of ozone and
hydrogen peroxide to these atmospheric and environmental phenomena cannot at
present be quantified, however. The Agency has chosen to discuss these topics
in the air quality criteria documents on the oxides of nitrogen and on particu-
late matter and sulfur oxides (U.S. Environmental Protection Agency, 1982a,
1982b). The present document includes brief discussions of the atmospheric
chemistry of ozone and hydrogen peroxide relative to these topics but does not
include information on visibility degradation or on acidic deposition.
For ease of printing, distribution, and review, this document is being
released in five volumes. The first volume contains the summary and conclusions
for the entire document. The second contains the introduction to the document
(chapter 2) and all chapters dealing with the precursors to photochemical
OZONER/B 2-4 6/22/84
-------
oxidants (chapter 3); the formation of photochemical oxidants (chapter 4);
their properties, reactions, and measurement once formed (chapter 5); and
their concentrations in ambient air (chapter 6). Volume III contains the
documentation of the effects of photochemical oxidants on vegetation, ecosys-
tems, and nonbiological materials. Volume IV contains the extensive body of
data available on the toxicologies! effects of ozone and other oxidants in
experimental animals and on in vitro effects on human cells and body fluids.
In Volume V, effects observed in human controlled exposures (chapter 11) and
in epidemiological studies (chapter 12) are presented. In addition, that
volume contains an evaluation of the integrated health data of probable conse-
quence for regulatory purposes (chapter 13).
It must be emphasized that neither control techniques nor control strate-
gies for the abatement of photochemical oxidants are discussed in this document.
Technology for controlling the emissions of nitrogen oxides and of volatile
organic compounds is discussed in documents issued by the Office of Air Quality
Planning and Standards (OAQPS). Likewise, issues germane to the scientific
basis for control strategies, but not germane to the development of criteria,
are addressed in respective documents issued by OAQPS.
In addition, certain issues of direct relevance to standard-setting are
not explicitly addressed in this document: (1) determination of what consti-
tutes an "adverse effect"; (2) assessment of risk; and (3) determination of a
margin of safety. While scientific data contribute significantly to decisions
regarding these issues, their resolution cannot be achieved solely on the
basis of experimentally acquired information. A fourth issue directly pertinent
to standard-setting is addressed partially in chapter 13 of this document;
that is, identification of the population at risk. The selection of the
population at risk is basically a selection by the Agency of the population to
be protected by the promulgation of a given standard. Information is presented
in chapter 13 of this document on factors, including pre-existing disease,
that biologically may predispose individuals and subpopulations to adverse
effects from exposures to ozone. The identification of a population at risk,
however, requires information above and beyond biological predisposition, such
as levels of exposure, activity patterns, and personal habits. Thus, the
identification of the population at risk relative to standard-setting is the
purview of OAQPS and is not fully addressed in this document.
OZONER/B 2-5 6/22/84
-------
This document contains a review and evaluation of literature on ozone and
other photochemical oxidants through early 1984. Emphasis has been placed on
studies in which concentrations were similar to those found in the ambient
air. On this basis, studies in which the lowest concentration employed exceeded
1 ppm have not been included unless the report contained unique data, such as
documentation of a previously unreported effect or of mechanisms of effects.
The one exception is in the areas of mutagenesis, teratogenesis, and reproduc-
tive effects, where, because of their importance to public health and welfare,
results of studies conducted at much higher than ambient levels have been
included.
A general policy exists within EPA of expressing concentrations of gas-
3
phase criteria pollutants in micrograms per cubic meter (ug/m ) as well as the
more widely used parts per million (ppm) or parts per billion (ppb). That
policy has been followed in those chapters in which the bulk of the data have
been obtained from laboratory studies done at room temperature (e.g., chapters
10 and 11). Data reported in ppm for studies done out of doors, such as field
and open-top chamber vegetation studies, ambient air monitoring, and research
on atmospheric chemistry, have not been converted. Conversion of reported ppm
and ppb units is highly questionable in these cases because it assumes standard
or uniform temperatures and pressures. For data in the health chapters, the
3 3
conversion units used are 1 ppm ozone = 1960 M9/m ; 1 PPb PAN = 4947 jjg/m ;
at 1 atmosphere pressure and 25°C.
OZONER/B 2-6 6/22/84
-------
2.4 REFERENCES
U.S. Congress. (1977) The Clean Air Act as amended August 1977. P.L. 95-95.
Washington, DC: U.S. Government Printing Office.
U.S. Environmental Protection Agency. (1978) Air quality criteria for ozone
and other photochemical oxidants. Research Triangle Park, NC: U.S.
Environmental Protection Agency; EPA report no. EPA-600/8-78-004.
U.S. Environmental Protection Agency. (1982a) Air quality criteria for oxides
of nitrogen. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no. EPA-600/8-82-026.
U.S. Environmental Protection Agency. (1982b) Air quality criteria for parti-
culate matter and sulfur oxides. Research Triangle Park, NC: U.S.
Environmental Protection Agency; EPA report no. EPA-600/8-82-029.
OZONER/B 2-7 6/22/84
-------
3. PRECURSORS TO OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
3.1 INTRODUCTION
Ozone and other photochemical oxidants are almost exclusively secondary
pollutants that are formed in the troposphere from primary pollutants emitted
into the ambient air from a variety of sources. Ozone-rich stratospheric air
and lightning constitute the only direct sources of ozone that are known to
contribute to the ambient air concentrations of ozone. Their contributions
are variable and minor. The primary pollutants that serve as precursors to
the formation of ozone and other photochemical oxidants are volatile nonmethane
organic compounds (NMOC) and nitrogen oxides (NO ).
f\
The purpose of this chapter is to present to the reader an overview of
the chemistry, measurement, sources, abundance, and transformation of precursors
in the United States in order (1) to convey an understanding of the oxidant
problem; (2) to convey an appreciation for the complexities of the production
and, hence, the abatement of oxidant pollution; and (3) to convey information
on the current status of and any anticipated changes in the kind, magnitude,
and distribution of precursors to oxidant formation.
3.2 DESCRIPTION AND CHARACTERIZATION OF PRECURSORS
3-2.1 Description and Basic Nomenclature of Nonmethane Organic Compounds
This section briefly describes and defines those hydrocarbons and other
volatile organic compounds commonly found in the ambient air in urban and
rural areas of the United States.
The term "hydrocarbon" has been used since the preliminary investigations
of tropospheric photochemistry to represent those compounds of carbon and
hydrogen that exist as gases in the ambient air and that participate along
with oxides of nitrogen in reactions that form ozone and other photochemical
oxidants. As knowledge of photochemistry has increased, carbon compounds
containing elements such as oxygen and the halogens have been discovered to be
important also in the formation of the urban photochemical complex. Thus, the
term "volatile organic compounds" (VOC) has come to be used to describe stable
organic compounds that exist as gases under normal atmospheric conditions and
that participate in the formation of photochemical oxidants. Recognition that
019WPS/B 3-1 6/26/84
-------
methane (CH^) is virtually unreactive in the photochemical formation of ozone
and other oxidants has given rise to an even more accurate term, "nonmethane
organic compounds" (NMOC), for describing those gas-phase organic compounds in
ambient air that serve as precursors to ozone and other photochemical oxidants.
While these three terms may sometimes appear to be used interchangeably in
this chapter, the terminology used reflects the terminology reported in the
specific literature cited in this chapter, though in some instances differen-
tiations may have been made for purposes of discussion.
As discussed in Section 3.3 below, methods for measuring gas-phase hydro-
carbons are not specific for hydrocarbons but may also detect other gas-phase
organic compounds, though they will not measure them accurately. Where methods
are used that permit speciation of the compounds measured, organic compounds
other than hydrocarbons can be and usually are excluded from the summation of
individual species used to arrive at a total nonmethane hydrocarbon (TNMHC)
concentration. Where researchers have used methods that do not permit specia-
tion, an indefinite and variable fraction of the reported TNMHC concentration
may, in fact, be the result of the presence of nonhydrocarbon organics and
such concentration data are more properly reported as total nonmethane organic
compounds (NMOC).
The discussion that follows is aimed at presenting basic facts on nomen-
clature and characteristics of photochemically reactive volatile organic
compounds that are relevant to the information given in subsequent sections of
this chapter and in the subsequent chapter.
3.2.1.1 Hydrocarbons. Hydrocarbons are compounds consisting of hydrogen and
carbon only. Except for carbides, carbonates, and oxides of carbon, all
compounds of carbon are organic. The volatility of hydrocarbons is related
generally to the number of carbon atoms in each molecule, as well as to tempera-
ture. Hydrocarbons with a carbon number of one to four are gaseous at ordinary
temperatures, while those with a carbon number of five or more are liquid or
solid in pure state. Liquid mixtures of hydrocarbons such as gasoline may
include some compounds that are gases, as well as those that are liquids, in
pure form. Likewise, gas-phase mixtures in ambient air will usually include
compounds that are liquid in their pure form. Hydrocarbons with a carbon
number of about 8 or less are abundant in ambient air, but those with a carbon
number greater than about 12 are generally not present at gaseous concentrations
high enough to be troublesome. A saturated hydrocarbon has each of its carbon
019WPS/6 3-2 6/26/84
-------
atoms bonded to four other atoms; whereas an unsaturated hydrocarbon has two
or more carbon atoms bonded to fewer than four other atoms.
Alkanes. Alkanes, also known as paraffins, are saturated hydrocarbons
having the general formula C H» ?. The first compound in the series is
methane, CH., which is unimportant in urban photochemistry because of its low
reactivity. Alkanes as a class are the least reactive of the photochenrically
important hydrocarbons (U.S. Environmental Protection Agency, 1978). Alkanes
may be straight- or branched-chain compounds, and are a subclass of the open-
chain (acyclic) hydrocarbons known as aliphatic hydrocarbons.
Alkenes. Alkenes, also known as olefins, have at least one unsaturated
bond. The number of hydrogen atoms in the general formula is decreased by two
for each double bond between carbon atoms, and the general formula for alkenes
with one double bond is C H0 . The first compound in the alkene class is
n Zn
ethene, also known as ethylene; the second is propene, also known as propylene.
Compounds with carbon numbers three or higher can have two double bonds between
carbons and are called dienes. The complete name of a diene is formed by
including a prefix with numbers that indicate the location of the double
bonds. Like alkanes, alkenes are aliphatic hydrocarbons and may exist as
straight or branched chains. As a class, alkenes are the most reactive hydro-
carbons in photochemical systems. The reader is referred to the brief discus-
sion on reactivity in section 3.2.2 and to discussions in Pitts et al. (1977)
and Dimitriades (1974).
Terpenes. Terpenes are a naturally occurring subgroup of alkenes having
the formula C-j^H,,.. Among the terpenes identified in ambient air, a- and
B-pinene have been most frequently studied. Both a- and p-pinene contain six-
membered rings, do several other terpenes; but at least one commonly occurring
member of this group, myrcene, is an acyclic or open-chain compound. Isoprene,
also an olefinic hydrocarbon that is naturally occurring, is a hemiterpene
having the formula C[.Hft.
Alkynes. Alkynes are open-chain hydrocarbons that contain one or more
triple bonds. Acetylene, C-H^, is the simplest member of the class and the
class as a whole is often referred to as the acetylenes. The general formula
for the acetylenes is C I-L __, and for each additional triple bond in the
molecule four hydrogen atoms must be removed from the general formula. Acety-
lene is commonly present in ambient air, is thought to be emitted largely from
mobile sources, and has often been measured as an indicator of auto exhaust
019WPS/B 3-3 6/26/84
-------
emissions, since it is relatively unreactive in ambient air and persists in
the atmosphere longer than most other exhaust components.
Alicyclics. Alicyclics are hydrocarbons in which the carbon chains are
arranged in rings (carbocyclic). They can be saturated compounds containing
no double bonds or may be unsaturated compounds containing one to three double
bonds. They may contain six-membered rings, but they do not possess the six-
membered ring containing three double bonds in resonance that is characteristic
of aromatic hydrocarbons. The carbocyclic group may have alky! substituents
or may be attached to more complex groups, including aliphatic chains. The
conventions of nomenclature applied to aliphatic compounds generally apply
also to alicyclics.
Aromatics. Aromatic hydrocarbons include various compounds having atoms
arranged in six-membered carbon rings with only one additional atom (of hydro-
gen or carbon) attached to each atom in the ring. Benzene is the simplest
compound in the series, having no side chains but only six carbon atoms and
six hydrogen atoms, linked by three conjugated double bonds.
Compounds containing the aromatic ring and elements other than carbon and
hydrogen are included with aromatic hydrocarbons in the general classification
"aromatics." The double bonds in aromatics are not nearly as chemically active
as those in alkenes because of an effect called "resonance stabilization." As
a class, aromatics are between alkanes and alkenes in photochemical reactivity.
Benzene, however, is considered to have low photochemical reactivity.
3.2.1.2 Aldehydes. Aldehydes are not true hydrocarbons, since they always
contain at least one oxygen atom. Nevertheless, they constitute probably the
single most abundant group of volatile organic compounds other than hydrocar-
bons in the ambient air. They are photochemically important compounds because,
along with other oxidizable compounds such as the hydrocarbons, they regenerate
free radicals that will react with oxygen in ambient air to form alkylperoxy
or hydroperoxy radicals (National Academy of Sciences, 1977) (see chapter 4).
Aldehydes are characterized by the presence of the formyl functional
group (CHO). As part of the formyl group, a carbonyl group exists, C=0,
having a carbon-oxygen double bond. The carbonyl group is not unique to
aldehydes, since it is found also in ketones and carboxylic acids, as well as
019WPS/B 3-4 6/26/84
-------
in more complex organic molecules such as the sugars. The carbonyl group
forms the basis, however, for one of the analytical methods used for measuring
aldehydes in ambient air (section 3.3),
3.2.1.3 Other Organic Compounds. Other organic compounds found in ambient
air are known to be photochemically reactive in the formation of ozone and
other photochemical oxidants. These other organic compounds do not occur in
ambient air collectively, much less singly, at concentrations that even approach
the concentrations of nonmethane hydrocarbons. Some of them are suspected of
having adverse health effects, however, and are therefore under scrutiny by
the U.S. Environmental Protection Agency at present, and documents dealing with
such compounds are in preparation by the Agency. These compounds are mentioned
here only because they are photochemically reactive, can serve as precursors to
oxidants, and because they contribute a small but indeterminate fraction of
the total NMOC concentrations reported when continuous hydrocarbon analyzers
(section 3.3) are used to determine the occurrence in ambient air of volatile
organic compounds.
Many of the volatile organics in ambient air that are not hydrocarbons
are organic ha!ides, in which one or more hydrogen atoms of a hydrocarbon have
been replaced by a halogen such as chlorine, fluorine, or iodine. When all
the hydrogen is replaced, the resulting compounds are called halocarbons. An
enormous number of relatively simple organic ha!ides are possible, since a
single carbon atom can be attached to one or more halogen atoms.
A number of halogenated hydrocarbons enjoy widespread industrial use as
commercial solvents for extraction and reaction media and for direct applica-
tion as cleaning and degreasing solvents. Many of these have been detected in
ambient air.
3-2.2 Pertinent Chemical and Physical Properties of Nonmethane Organic
Compounds
The chemical and physical properties of nonmethane organic compounds that
are most pertinent to their role as precursors to ozone and other oxidants are
those properties that govern their emission into and persistence in the atmos-
phere (volatility) and their reactivity in atmospheric photochemical reactions.
A major discussion of these properties lies outside the scope of this document,
inasmuch as such a discussion would necessitate a much more thorough review of
019WPS/B 3-5 6/26/84
-------
atmospheric photochemistry than is relevant to the purposes of this document.
The photochemical reactivity of subclasses and individual species of hydrocar-
bons and of other volatile organic compounds is relevant to mechanistic studies
in atmospheric chemistry and to modeling and other oxidant-control-related
research but is not pertinent to the derivation of criteria. By way of illus-
tration, however, of the differences in reactivity, some very general comments
are in order.
The 1978 criteria document for ozone and other photochemical oxidants
summarized reactivity data acquired from the mid-1960s to the mid-1970s (U.S.
Environmental Protection Agency, 1978). Reference to tables of reactivity
schemes given in the 1978 document shows the relatively higher reactivities of
internally double-bonded alkenes, of aliphatic aldehydes and other carbonyl
compounds (such as branched alkylketones and unsaturated ketones), of dienes,
of 1-alkenes, of partially halogenated alkenes, and of alky!benzenes (primary
and secondary monoalkylbenzenes, and di-, tri-, and tetraalkylbenzenes).
Other compounds also have relatively high reactivity but are not expected to
be as abundant in ambient air as the compounds cited above. The reader is
referred to the 1978 criteria document and the references therein for further
information on reactivities of specific compounds. The concentrations at
which respective classes and species of NMOC occur in ambient air are presented
in section 3.5.
The chief basis of most of the proposed reactivity classifications is the
rate of reaction between an organic and the hydroxyl radical (HO- or OH-). A
key reaction of volatile organic compounds in ambient air, regardless of class
of VOC, is their oxidation via attack by HO- (Atkinson et al., 1979; Atkinson
et al., 1982). This reaction is the first step in a chain reaction that is
propagated by various organic peroxy radicals. It is thought to be the predom-
inant loss process for most organics in the troposphere (Atkinson et al. ,
1982), even for alkanes, for which rate constants for their reactions with
HO- are lower than the alkenes of equivalent carbon numbers (Atkinson et al.,
1979).
Alkenes are unique among VOC in ambient air, inasmuch as they exhibit
reactivity toward ozone as well as toward the hydroxyl radical (Niki et al.,
1983). In addition, the reactions of alkenes subsequent to the initial reac-
tion with HO- are much better understood for the alkenes than for alkanes or
aromatics. The reaction scheme accepted at present for the H0--initiated
019WPS/B 3-6 6/26/84
-------
oxidation of trans-2-butene, as an example of the alkenes, is given in chapter 4
in some detail.
Subsequent reactions, following HO- attack, are complex for the longer-
chain alkanes and are at present rather poorly understood for haloalkanes,
haloalkenes, and for aromatics, including the halogenated aromatics. For all
of these compounds, as well as for oxygenated organics, the initial step is
attack by the hydroxyl radical. Whether the next step is hydrogen abstraction,
addition of OH to a double bond, or the formation of an energy-rich OH-organic
adduct depends largely upon the structure of the organic, although temperature
and pressure dependencies have been observed in experimental smog chamber work
(Atkinson et a!., 1979; 1980). In fairly recent work, Killus and Whitten
(1982) proposed that ring opening can occur in toluene (an aromatic) subsequent
to the formation of an OH-toluene adduct, resulting in products unlike those
occurring from HO- attack on alkanes or alkenes.
Reactivities in ambient air are not necessarily the same as in smog
chamber systems. For example, source strength, meteorological variables,
transport, and the age of the air mass containing the VOC are all known to
affect reactivity. The discussion above, however, presents some of the basic
generalizations that are pertinent to the photochemical reactivity of various
classes of VOC in ambient air.
3.2.3 Description and Properties of Nitrogen Oxides
The physical and chemical properties of the nitrogen oxides that serve as
precursors in the formation of ozone and other photochemical oxidants have
been documented in a recent air quality criteria document (U.S. Environmental
Protection Agency, 1982). The most pertinent properties are briefly summarized
here. The role of nitrogen oxides in the formation of oxidants in the tropos-
phere is discussed in chapter 4 and in the document cited above.
The three most abundant oxides of nitrogen In ambient air are nitric
oxide (NO), nitrogen dioxide (NO,,), and nitrous oxide (NO). The latter,
though ubiquitous, is not known to participate in photochemical reactions in
the troposphere. The two important oxides of nitrogen relative to photochemical
processes in the troposphere are NO and N0_. Their importance derives from
their abundance in ambient air and their participation in cyclic reactions
leading to the production of ozone and other oxidants, as described in chapter 4.
The basic reactions of importance are (1) the photolysis of N0? (X < 430 nm);
019WPS/B 3-7 6/26/84
-------
(2) subsequent formation of ozone (CL) from the atomic oxygen produced in the
photolysis of N02 (in the presence of a third, energy-absorbing molecule); and
(3) the subsequent regeneration of N02 by the reaction of NO with 0_. Coupled
with these basic reactions are reactions between NO and free radicals in the
atmosphere (hydroperoxy, alkylperoxy, and acylperoxy) that oxidize NO to NO-,
disturbing the N0-N02 equilibrium that would otherwise exist, and leading,
then, to the buildup of 03 (National Academy of Sciences, 1977). These reac-
tions, and further information on the source of the free radicals, are given
in chapter 4. Basic physical and chemical properties of NO and NO- are given
in Table 3-1.
TABLE 3-1. PHYSICAL AND CHEMICAL PROPERTIES
OF NITRIC OXIDE AND NITROGEN DIOXIDE
Property
NO
NO,
Odor
Taste
None
Pungent
Color
Absorption
A., nma
Other
properties
of note:
None
<230
Reddish-brown
Broad range,
both >400
and <400
Corrosive, strong oxidant.
Photolyzes at A. <430 nm.
Low partial pressure in ambient air.
Uneven number of valence electrons.
Forms dimers (N204).
Visible light \ >4QO nm; ultraviolet \ <400 nm. Solar UV radiation in the
troposphere extends from about A230 nm to about A400 nm.
Source: Derived from National Academy of Sciences (1977) and U.S. Environmen-
tal Protection Agency (1982).
3.3 SAMPLING, MEASUREMENT, AND CALIBRATION METHODS FOR PRECURSORS TO OZONE
AND OTHER PHOTOCHEMICAL OXIDANTS
During the last decade, a number of advances have been made in the metho-
dology for determining nonmethane organic compounds (NMOC) and oxides of nitro-
gen. An overview of these advances will be discussed in this chapter. In the
019WPS/B
3-8
6/26/84
-------
case of NMOCs, early methods did not provide for any speciation of the complex
mixture of organics in ambient air. Nonetheless, these non-speciation methods
were employed and served a useful purpose in providing a data base for early
photochemical modeling studies. As the air quality models grew more sophisti-
cated, however, the need arose for more specific information concerning the
organic composition of the atmosphere. Consequently, methodology was developed
to provide for detailed speciation of NMOCs. In addition to improving the
data base for photochemical modeling, the NMOC speciation techniques have also
been utilized to characterize various sources of pollution (mobile versus
stationary) and have led to the identification and quantification of many com-
pounds not previously identified in the ambient air.
The development of methodology for oxides of nitrogen has likewise
advanced since the original EPA Federal Reference Method for measurement of
nitrogen dioxide (N0_), the Jacobs-Hochheiser technique, was withdrawn in
1973. A number of methods for nitric oxide (NO) and N0~ have been proposed
and evaluated since then. Information on these more recent methods is pre-
sented in this section.
3.3.1 Nonmethane Organic Compounds
Numerous sampling, measurement, and calibration methods have been employed
to determine vapor-phase nonmethane organic compounds (NMOC) in ambient air.
Some of the measurement methods utilize detection techniques that are highly
selective and sensitive to specific functional groups or atoms of a compound
(e.g., formyl group of aldehydes, halogen), while others respond in a more
universal manner; that is, to the number of carbon atoms present in the organic
molecule. In order to present an overview of the most pertinent measurement
methods, nonmethane organic compounds have been arranged in three major classi-
fications in this section. These classifications are "nonmethane hydrocarbons,"
"aldehydes," and "other oxygenated compounds." Each classification and the
associated measurement methods will be discussed. Sampling and calibration
procedures used with these measurement methods will also be described. Refer-
ence will also be made to those analytical methods utilized in more than one
of the above classifications.
3.3.1.1 Nonmethane Hydrocarbons. Nonmethane hydrocarbons constitute the
major portion of nonmethane organic compounds in ambient air (section 3.5).
Traditionally, nonmethane hydrocarbons have been measured by methods that
019SPW/A 3-9 6/28/84
-------
employ a flame ionization detector (FID) as the sensing element. This detector
was originally developed for gas chromatography and employs a sensitive electro-
meter that measures a change in ion intensity resulting from the combustion of
air containing organic compounds. Ion formation has been shown to be essentially
proportional to the number of carbon atoms present in the organic molecule
(Sevcik, 1975). Thus, aliphatic, aromatic, alkenic, and acetylenic compounds
all respond similarly to give relative responses of 1.00 ± 0.10 when corrected
for the number of carbon atoms present (e.g., 1 ppm hexane = 6 ppm C; I ppm ben-
zene = 6 ppm C; I ppm propane = 3 ppm C). Carbon atoms bound to oxygen, nitrogen,
or halogens give reduced relative responses (Dietz, 1967). Consequently, the FID,
which is primarily used as a hydrocarbon measuring method, should more correctly
be viewed as an organic carbon analyzer.
In the following sections, discussion will focus on the various methods
utilizing this detector to measure nonmethane organlcs. Those methods for
total nonraethane organic compounds in which no compound speciation is obtained
will be covered first. Methods for determining individual compounds will then
be discussed.
3.3.1.1.1 Non-speciation methods. The EPA reference method for nonmethane
organic compounds, which was promulgated in 1971, involves the gas chromato-
graphic separation of methane (CH^) from the remaining organics in an air
sample (U.S. Environmental Protection Agency, 1975). Methane is eluted through
the chromatographic column and detected; another sample of air is subsequently
processed without methane separation. Subtraction of the first value from the
second produces a nonmethane organic concentration.
A number of studies of commercial analyzers employing the Federal Reference
Method have been reported (Reckner, 1974; McElroy and Thompson, 1975; Harrison
et a!., 1977; Sexton et al., 1981). In one of the first studies, the analyses
of known synthetic mixtures of hydrocarbons were conducted by 16 users of the
reference method (Reckner, 1974). The nonmethane concentrations tested in this
study were 0.23 and 2.90 ppm C. The results shown in Table 3-2 indicate the
percentage error from the two known concentrations. At the 0.23 ppm level, the
majority of the measurements were in error by amounts greater than 50 percent.
At 2.90 ppm, most of the measurements were in error by only 20 percent or less.
In general, all of the above studies indicated an overall poor performance
of the commercial instruments when either calibration or ambient mixtures
containing NMOC concentrations less than 1 ppm C were used. The major problems
019WPS/B 3-10 6/26/84
-------
TABLE 3-2. PERCENTAGE DIFFERENCE FROM KNOWN CONCENTRATIONS
OF NONMETHANE HYDROCARBONS OBTAINED BY SIXTEEN USERS
Known
concentration,
ppm
0.23
2.90
% difference from given concentration
>100
6
2
50 to 100
4
—
20 to 50
3
3
10 to 20
2
2
0 to 10
1
9
Source: Reckner, 1974.
associated with these instruments have been summarized in a recent technical
assistance document (U.S. Environmental Protection Agency, 1981) for the
calibration and operation of ambient-air nonmethane organic compound analyzers.
Table 3-3 lists these shortcomings. The first three problems have essentially
been corrected through work done at the National Bureau of Standards. The
assistance document further summarizes ways to reduce the effects of existing
problems and Table 3-4 presents these recommendations.
As a result of the above deficiencies, other approaches to the measure-
ment of nonmethane organics have been investigated. One such method, developed
in 1973, utilizes the fact that CH4 requires more heat for combustion than
other organics (Poli and Zinn, 1973). One portion of the air sample passes
through a catalyst bed where all hydrocarbons except CH. are combusted. This
sample stream then enters an FID where the CH. concentration alone is recorded.
The other portion of the sample passes directly to a second FID for a total
organic carbon measurement. By simultaneously processing both signals, an
NMOC value is obtained. Although it provides a continuous measurement of NMOC
levels, this method is also subject to many of the same shortcomings attributed
to the EPA reference method.
Recently, a prototype instrument that measures NMOCs by optical absorption
has been developed (Manos et al. , 1982). The unit oxidizes NMOCs to carbon
dioxide (C0?) and uses a non-dispersive infrared absorption technique to
measure the organic burden indirectly. Ascarite serves to remove C0_ initially
present in air and a hopcalite catalyst selectively oxidizes organics other
than methane to C0? and H?0. Since carbon monoxide (CO) will also oxidize to
C02 during this process, a dual-channel system is utilized to correct for the
019WPS/B 3-11 6/26/84
-------
TABLE 3-3. SUMMARY OF PROBLEMS ASSOCIATED WITH GATHERING
NMOC DATA BY MEANS OF AUTOMATED ANALYZERS
1. Contaminants may be present in compressed-gas cylinders containing
calibration gases.
2. Compressed-gas cylinders of calibration gases sometimes contain the
standard in a nitrogen or argon background. When no oxygen is blended
with these gases, FID sensitivity is altered.
3. The assay of calibration gases contained in compressed-gas cylinders (as
received from the supplier) is sometimes incorrect.
4. There are wide differences in the per-carbon response to different NMOC
species.
5. FID analyzers require hydrogen, which presents a potential operational
hazard.
6. The NMOC concentration is obtained by subtraction of two relatively large
and nearly equal numbers (TOC-CH =NMOC) and thus is subject to large, rela-
tive errors.
7. NMOC analyzers may exhibit excessive zero and span drift during unattended
operation.
8. The complex design of some NMOC analyzers creates unique problems that are
generally not experienced in other pollutant analyzers. Meticulous set-up,
calibration, and operation procedures (which are analyzer-specific) are
difficult to understand and follow.
Source: U.S. Environmental Protection Agency, 1981.
contribution from ambient CO concentrations. This unit performed well during
a brief laboratory evaluation using calibration standards; however, more
extensive laboratory and field tests are needed before the unit can be con-
sidered suitable for NMOC measurement.
Other methods under development and evaluation include oxidation-reduction
schemes in which nonmethane organics are chromatographically separated from
methane and non-organic species and then oxidized to C0?, reduced to CH., and
detected by FID (U.S. Environmental Protection Agency, 1979). In cases where
organic carbon concentrations are greater than 100 ppb, the reduction step in
this method can be eliminated and a non-dispersive infrared analyzer can be
used to detect the C0» formed during the oxidation step (Salo et al., 1975).
019WPS/B 3-12 6/26/84
-------
TABLE 3-4. SUMMARY OF RECOMMENDATIONS TO REDUCE THE EFFECTS OF PROBLEMS
LISTED IN TABLE 3-3
1. Calibration gases should be checked to determine the concentration of
contaminants.
2. Calibration concentrations should be obtained by dynamic dilution of a
pollutant standard with zero-grade air containing oxygen. The dilution
ratio should be sufficiently high (VLOO:1) to ensure that the calibration
sample contains 20.9% ± 0.3% oxygen.
3. All calibration standards contained in compressed-gas cylinders should be
traceable to Standard Reference Materials from the National Bureau of
Standards.
4. The MMOC response should be calibrated to a propane standard.
5. The operator should use documented procedures for hydrogen safety.
6. All channels should be properly calibrated.
7. The FIDs should be operated in accordance with instructions supplied by
the manufacturer and this document.
8. The training of qualified operators should be augmented with a Technical
Assistance Document, which provides detailed calibration and operation
procedures for NMOC analyzers.
Source: U.S. Environmental Protection Agency, 1981.
A unique total NMOC measurement technique has resulted from the initial
work of McBride and McClenny (1980). Their approach involved the cryogenic
preconcentration of nonmethane organic compounds and the measurement of the
revolatilized NMQCs using flame ionization detection. Their procedure is as
follows. A fixed volume of sample is drawn through a trap cooled to liquid
argon temperature (liquid N» can not be used since it will also condense
methane and air). At this temperature all NMOCs are condensed onto the trap
(open tube). After the residual CH. and oxygen are cleared from the trap by
the helium carrier gas, the trap temperature is raised to revolatilize the
NMOC. Using helium as the carrier gas was shown to produce less variation in
response to different organic compounds than direct air injection, which is
employed in conventional NMOC analyzers.
Jayanty et al. (1982) improved upon the original design of McBride and
McClenny (1980) and evaluated the resulting system with a variety of aliphatic
019WPS/B 3-13 6/26/84
-------
and aromatic compounds. The range of detection with the cryogenic trapping
procedure (500 ml of air) was about 50 ppb C and a linear dynamic range of 50
to 5,000 ppb C was attained. Humidity did not generally interfere with the
analysis. Sample precisions of ±5 percent for single- and multiple-component
gas standards and ±10 percent for ambient samples were consistently achieved.
Responses for aromatic compounds, however, were less than expected. The
researchers recommended additional testing and instrument refinement in order
to resolve this problem.
3.3.1.1.2 Speciation methods. The primary separation technique utilized for
NMOC speciation is gas chromatography (GC). Coupled with flame ionization
detection, this analytical method permits the separation and identification of
many of the organic species present in ambient air.
Compound separation is accomplished by means of both packed and capillary
GC columns. If high resolution is not required and large sample volumes are
to be injected, packed columns are employed. The traditional packed column
may contain either (1) a solid polymeric adsorbent (gas-solid chromatography)
or (2) an inert support, coated with a liquid (gas-liquid chromatography).
Packed columns containing an adsorbent substrate are required to separate
C,-C- compounds. The second type of column can be a support-coated or wall-
£. O
coated open tubular capillary column. The latter column has been widely used
for environmental analysis because of its superior resolution and broader
applicability. The wall-coated capillary column consists of a liquid station-
ary phase coated or bonded (cross-linked) to the specially treated glass or
fused-silica tubing. Fused-silica tubing is most commonly used because of its
physical durability and flexibility.
When a complex mixture is introduced into a GC column, the carrier gas
(mobile phase) moves the sample through the packed or coated column (stationary
phase). The chromatographic process occurs as a result of repeated sorption-
desorption of the sample components (solute) as they move along the stationary
phase. Separation results from the different affinities that the solute com-
ponents have for the stationary phase.
The GC-FID technique has been used by numerous researchers to obtain
ambient NMOC data (see section 3,4). In a recent report, Singh (1980) utilized
the cumulative experience of these researchers in order to prepare a guidance
document for the state and local air pollution agencies interested in obtaining
speciation data. In general, most researchers have employed two gas chromato-
graphic units to carry out analyses of NMOC species in ambient air. Organic
019WPS/B 3-14 6/26/84
-------
compounds of C_ through C5 are easily measured on one unit using packed-column
technology, while the other GC separates >C. organics using a capillary column.
Figures 3-1 and 3-2 and Table 3-5 illustrate typical chromatograms of NMOCs
found in urban air. As these figures indicate, all the major peaks eluted
have been identified; on a mass basis, these compounds represent from 65 to 90
percent of the measurable nonmethane organic burden. Identification of GC
peaks is based upon matching retention times of unknowns with those of standard
mixtures. The use of dedicated computer systems facilitates this task, but
close scrutiny of the data is still necessary in order to correct periodic
mis-identification of unknowns resulting from variations in retention time.
Subsequent verification of the individual species is normally accomplished
with gas chromatographic/mass spectrometric (GC/MS) techniques. Compound-
specific detection systems, such as electron capture, flame photometry, and
spectroscopic techniques, have also been employed for peak identifications. A
discussion of these systems, however, is beyond the scope of this report.
Several documents covering these detection systems are available (Lamb et al.,
1980; Riggin, 1983).
Because the organic components of the ambient atmosphere are present at
ppb levels or lower, some means of sample preconcentration is necessary in
order to provide sufficient material for the GC-FIO system. The two primary
techniques utilized for this purpose are cryogenic collection and the use of
solid adsorbents. The more commonly used sorbent materials are generally
divided into three categories: (1) organic polymeric adsorbents, (2) inorganic
adsorbents, and (3) carbon adsorbents. Primary organic polymeric adsorbents
used for NMOC analyses include the materials Tenax GC and XAD-2. These materials
have a low retention of water vapor and, hence, large volumes of air can be
collected. These materials do not, however, efficiently capture highly volatile
compounds such as C? to Cj. hydrocarbons, nor certain polar compounds such as
methanol and acetone. Primary inorganic adsorbents are silica gel, alumina,
and molecular sieves. These materials are more polar than the organic polymeric
adsorbents and are thus more efficient for the collection of the more volatile
and polar compounds. Unfortunately, water is also efficiently collected,
which in many instances leads to rapid deactivation of the adsorbent. Carbon
adsorbents are less polar than the inorganic adsorbents and, as a result,
water adsorption by carbon adsorbents is a less significant problem. The
019WPS/B 3-15 6/26/84
-------
uiui a.
QJ^* 3
Jg-o
<-S c
o
I
a>
a
I
Figure 3-1. Light hydrocarbon chromatogram (C2 to
CB) from the university campus site, Cincinnati,
Ohio, August 24, 1981.
Source: Holdren et al. (1982).
3-16
-------
10 11 1314 15 16 18 25 30
19
LJJ
12
17
20
22
21!
23
24
26
29
28
31
33
34
35
32
Figure 3-2. Heavy hydrocarbon chromatogram !C4 to Ci0) from university campus
site, Cincinnati, Ohio, August 24, 1981.
Source: Holdren et ai. (1982).
3-17
-------
TABLE 3-5. IDENTIFICATION KEY FOR TYPICAL HEAVY HYDROCARBON CHROMATQGRAM,
C4 TO C10 (UNIVERSITY CAMPUS SITE, CINCINNATI, OHIO, AUGUST 24, 1981)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Hydrocarbon
Ethane
Ethyl ene
Acetylene
Propane
Propene
T_so_-Butane
n- Butane
trans-Butane
cis-Butene
iso-Pentane
n-Pentane
2, 3- Dimethyl butane
2-Methyl pentane
3-Methyl pentane
n-Hexane
Methylcyclopentane
2 , 4-Dimethy 1 pentane
Benzene
2-Methyl hexane
3-Methyl hexane
2,2,4-Trimethylpentane
n-Heptane
Ethyl eye lopentane
2 , 4-Dimethy 1 hexane
Toluene
2-Methyl heptane
3-Methyl heptane
jrrOctane
Ethyl benzene
m- + p_-Xylene
o-Xylene
n-Propyl benzene
£- Ethyl toluene
1, 3, 5-Trimethyl benzene
o-Ethyl toluene
1, 2, 4-Tri methyl benzene
Identified NMOC
Unidentified NMOC
Total NMOC
Concentration,
ppb C
43.2
113.5
44.1
28.8
29.8
44.3
115.9
9.7
6.4
137.3
68.6
11.1
42.8
28.0
33.7
27.2
4.2
35.1
25.6
14.6
5.2
15.1
7.7
3.1
73.5
6.7
9.2
6.9
12.5
42.4
15.3
3.6
9.5
7.8
7.9
17.2
1108
762
1870
Source: Holdren et a!., 1982.
019WPS/B 3-18 6/26/84
-------
carbon-based materials also tend to exhibit much stronger adsorption properties
than organic polymeric adsorbents; thus, lighter-molecular-weight species are
more easily retained. These same adsorption effects result, however, in
irreversible adsorption of many compounds. Furthermore, the very high thermal
desorption temperatures required (350 to 400 °C) limit their use and also may
lead to degradation of labile compounds. The commonly available classes of
carbon adsorbents include: (1) various conventional activated carbons;
(2) carbon molecular sieves (Spherocarb, Carbosphere, Carbosieve); and (3) car-
bonaceous polymeric adsorbents (Ambersorb XE-340, XE-347, XE-348).
Although a number of researchers have employed solid adsorbents for the
characterization of selected organic species in air, only a few attempts have
been made to identify and quantitate the range of organic compounds from C2
and above. Westberg et al. (1980) evaluated several carbon and organic poly-
meric adsorbents and found that Tenax-GC exhibited good collection and recovery
efficiencies for >Cfi organics; the remaining adsorbents tested (XAD-4, XE-340)
were found unacceptable for the lighter organic fraction. The XAD-4 retained
>C? organic gases, but it was impossible to completely desorb these species
without partially decomposing the XAD-4. Good collection and recovery effici-
encies were provided by XE-340 only for organics of C. and above.
Ogle et al. (1982) used a combination of adsorbents in series and designed
an automated GC-FID system for analyzing C« through C-,,. hydrocarbons. Tenax
GC was utilized for Cg and above, while Carbosieve S trapped C, through C5
organics. Silica gel followed these adsorbents and effectively removed water
vapor while passing the C« hydrocarbons onto a molecular-sieve 5A adsorbent.
The combined sorbents have been laboratory-tested with a 38-component hydro-
carbon mixture. Good collection and recovery efficiencies were obtained.
Preliminary field tests have also been successful, but a very limited data
base exists. Futhermore, the current chromatographic procedures utilize
packed-column technology. The addition of capillary columns to this system
would permit better resolution of the complex mixtures typically found in
ambient air.
Presently, the preferred method for obtaining NMOC data is cryogenic pre-
concentration (Singh, 1980). Sample preconcentration is accomplished by
directing air through a packed trap immersed in either liquid oxygen (b.p.
-183°C) or liquid argon (b.p. -186°C). For the detection of about 1 ppb C of
an individual compound, a 250-cc air sample is normally processed. The collec-
tion trap is generally filled with deactivated 60/80 mesh glass beads (Westberg
019WPS/B 3-19 6/26/84
-------
et al., 1974), although coated chromatographic supports have also been used
(Lonneraan et al., 1974). Both of the above cryogens are sufficiently warm to
allow air to pass completely through the trap, yet cold enough to collect
trace organics efficiently. This procedure will also condense water vapor.
An air volume of 250 cc at 50 percent relative humidity and 25°C will contain
approximately 2.5 mg of water, which appears as ice in the collection trap.
The possibility that ice will plug the trap and stop the sample flow is of
concern; furthermore, water transferred to the capillary column during the
thermal desorption step may also cause plugging and other deleterious column
effects. Nonetheless, this limitation has not diminished research efforts to
characterize the ambient atmosphere.
In addition to direct sampling via preconcentration with sorbents and
cryogenic techniques, collection of whole air samples is frequently used to
obtain NMOC data. Rigid devices such as syringes, glass bulbs, or metal
containers; and non-rigid devices such as Tedlar and Teflon plastic bags are
often utilized during sampling. The primary purpose of whole air collection
is to store an air sample temporarily until subsequent laboratory analysis is
performed. The major problem with this approach is assuring the integrity of
the sample contents prior to analysis. Of concern is whether sample components
of interest are adsorbed or decomposed through interaction with the container
walls. Sample condensation may also occur at elevated concentrations or when
samples are stored under high pressures (i.e., in metal containers). Contami-
nation from sampling containers is likewise a frequent occurrence (Lonneman et
al., 1981; Seila et al. , 1976). Table 3-6 summarizes the advantages and
disadvantages of the primary collection media for NMOC analysis.
3.3.1.1.3 Calibration. Calibration procedures for NMOC instrumentation
require the generation of dilute mixtures at concentrations expected to be
found in ambient air. Methods for generating such mixtures are classified as
static or dynamic systems.
Static systems are generally preferred for quantitating NMOCs. The most
commonly used static system is a compressed gas cylinder containing the appro-
priate concentration of the compound of interest. These cylinder gases may
also be diluted with hydrocarbon-free air to provide multi-point calibrations.
Calibration and hydrocarbon-free air cylinders are available commercially.
Additionally, some standard gases such as propane and benzene are available
from the National Bureau of Standards (NBS) as certified standard reference
019WPS/B 3-20 6/26/84
-------
TABLE 3-6. SUMMARY OF ADVANTAGES AND DISADVANTAGES OF PRIMARY COLLECTION MEDIA FOR NMOC ANALYSIS'
Sampling technique
Advantages
Disadvantages
1. Solid adsorbents
CO
I
2. Cryogenic pre-
concentrat ion
Many sorbents do not retain H20
vapor; thus, large volumes of air
can be processed.
Integrated samples from a period
of minutes to days are easily
obtained.
Sample cartridges are convenient
for field use.
o A wide range of organic material
can be collected.
o Artifact problems are avoided.
Collected organics are immediately
available for analysis, without
solvent removal or use of high
thermal desorption temperatures.
Collected species are stable; good
recovery efficiencies are obtained.
No single adsorbent can be used to
collect and recover organics of C2
and above.
Contamination and artifact peaks
plague many sorbent systems.
Many adsorbents require high
(>300°C) thermal desorption tem-
peratures, which may lead to
degradation of labile compounds,
artifact peak formation, etc.
Breakthrough volume and collection
and recovery efficiencies must be
determined for each compound of
interest.
Volume of air collected is limited
by amount of moisture condensing.
Liquid argon or oxygen is necessary
for preconcentration.
-------
TABLE 3-6. SUMMARY OF ADVANTAGES AND DISADVANTAGES OF PRIMARY COLLECTION MEDIA FOR NMOC ANALYSIS (continued)
Sampling technique
Advantages
Disadvantages
OJ
I
3. Rigid containers
(Metal canisters)
4. Non-rigid containers
(Teflon and Tedlar
Bags)
o Can be treated to make them
chemically unreactive.
o Durable; easy to clean, transport,
and use.
o Can be pressurized; leakage and
permeation minimized.
o Excellent stability for many trace
species; long-term storage is
possible.
o Readily available.
o Convenient for collecting integrated
samples.
o Good short-term stability of trace
species.
o High initial cost.
o Limited collection volume.
o Difficult to collect integrated
samples.
o Subject to breakage (at seams)
during handling.
o Admits sunlight.
o Slow permeation of chemicals out
of and into plastic bags during
storage.
o Outgassing contamination from bag
material.
o Short storage life.
Derived from Singh (1980); Jayanty and McElroy (1982); Sexton et al,
(1976); Lonneman et al. (1981); Holdren et al. (1982).
(1981); National Academy of Sciences
-------
materials (SRM). Commercial mixtures are generally referenced against these
NBS standards. In its recent technical assistance document for operating and
calibrating continuous NMOC analyzers, the U.S. Environmental Protection
Agency (1981b) recommended propane-in-air standards for calibration. Some
commercially available propane cylinders have been found to contain other
hydrocarbons (Cox et al., 1982), so that all calibration data should be refer-
enced to NBS standards.
Because of the uniform carbon response of a GC-FID system (±10 percent)
to hydrocarbons (Dietz, 1967), a common response factor is assigned to identi-
fied and unknown compounds obtained from the speciation systems. In cases
where these compounds may be oxygenated species, an underestimation of the
actual concentrations will be reported. In order to attain better accuracy,
relative response factors for substituted hydrocarbons should be experimentally
determined. In determining these factors, dynamic calibration systems (permea-
tion tubes, diffusion tubes, syringe delivery systems) are normally employed
to generate vn situ known concentrations of the individual compound of concern.
Although the GC-FID response has been found to be unchanged over a period of
many months, it is recommended that single-point calibration checks be performed
daily to assure the highest quality of the data. Weekly multi-point calibra-
tions are sufficient during normal field programs.
3.3.1.1.4 Comparison of non-speciation versus speciation methods. Speciation
methods involving cryogenic preconcentration (section 3.3.1.1.2) have been
compared with commercially available or prototype continuous NMOC analyzers in
the following studies.
Jayanty et al. (1982) conducted a laboratory comparison between the pro-
totype non-speciation method described earlier (section 3.3.1.1.1) and their
gas chromatographic separation method. Comparison of the two methods for 12
ambient air samples collected in stainless steel canisters showed agreement
within ±15 percent. Ambient air concentrations ranged from 100 to 1000 ppb C.
Lonneman (1979) compared total NMOC and speciation methods during field
studies in Houston in 1978. Samples were collected during 3-hour integrated
time periods (6 to 9 a.m., 1 to 4 p.m.) in Tedlar bags for subsequent analysis.
The correlation coefficients for 150 measurement pairs from five sites averaged
0.74. For data pairs of 500 ppb C and less, an average correlation coefficient
2
(r ) of 0.55 was calculated, with a low value of 0.12 at one site. Lonneman
attributed the low correlations to maintenance and calibration problems in the
019WPS/B 3-23 6/26/84
-------
continuous analyzers and concluded that the results from continuous analyzers
are "at best marginal for use in photochemical model applications."
Holdren et al. (1982) made a similar comparison during a 2-month study at
urban sites in Cincinnati and Cleveland, Ohio. They utilized a GC/cryogenic
trapping technique and compared their results with data from state-operated
NMOC analyzers (SAROAD data). Ambient air samples were collected in Teflon
bags (6 to 9 a.m. integrated collection) and were transferred immediately to
pretreated aluminum cylinders for shipment and analysis at the central labora-
tory. Concentrations of NMOC ranged from 200 to 2100 ppb C. Linear regression
analyses resulted in correlation coefficients that ranged from 0.75 to 0.92
for the four urban sites (total of 67 comparisons). Limiting the comparisons
to concentrations of 500 ppb C and lower resulted in an average correlation
coefficient of 0.10 (at all four sites).
Richter (1983) compared continuous total NMOC with GC speciation results
obtained at seven fixed ground-level sites used in the Northeast Corridor
Regional Modeling Project (NECRMP). The NMOC data were obtained in real time,
while Teflon bags were used to collected integrated samples (6 to 9 a.m.) for
the GC/cryogenic analyses. Over 60 comparisons were available from each site.
Table 3-7 summarizes statistical information obtained from least-squares
analysis of the data (Richter, 1983). As the table indicates, only data from
the East Boston site exhibited a high correlation coefficient. This study
represents the most extensive effort made yet to compare the two NMOC measuring
methods. The participating laboratories paid a great deal of attention to
technical details for correct instrument operation, calibration, etc. All
data were carefully examined by all contractors and only "verified" data were
compared. Yet the above results indicate that much more work is needed to
resolve the differences between the two methods.
3.3.1.2 Aldehydes. Aldehydes play a unique role in the photochemistry of the
troposphere. They contribute to the formation of photochemical oxidants as
precursors to free radicals and occur as products of the photooxidation of
gas-phase hydrocarbons, often as chain terminators. Aldehydes appear to be
second in abundance, next to nonmethane hydrocarbons, among classes of vola-
tile organic compounds found in ambient air. Historically, the major problem
in measuring concentrations of aldehydes in ambient air has been to find an
appropriate monitoring technique that is sensitive to low concentrations and
specific for the various homologues. Early techniques for measuring formalde-
hyde, the most abundant aldehyde, were subject to many interferences and
019WPS/B 3-24 6/26/84
-------
TABLE 3-7. GC/CONTINUOUS NMOC ANALYZER COMPARISONS, LEAST-SQUARES REGRESSIONS
Location
West End Library,
Washington, DC
Read Street,
Baltimore, MD
Essex, MD
Linden, NJ
Newark, NJ
East Boston, MA
Water town, MA
Slope Intercept, ppm C
0.552
0.113
0.835
0.531
0.987
1.108
0.750
-0.552
-0.283
-0.101
—
-0.277
+0.095
-0.568
Standard
error
0.672
0.713
0.599
0.865
0.574
0.327
0.574
r2
0.169
0.0077
0.354
0.141
0.467
0.887
0.475
Source: Richter, 1983.
lacked sensitivity at low concentrations. When used by skillful technicians,
the more recently developed techniques can characterize with relative accuracy
the types and amounts of aldehydes in ambient air. This section describes
those methods currently used for measuring aldehydes in ambient air. These
include the chromotropic acid (CA) method for formaldehyde, the 3-methyl-2-
benzothiazolone (MBTH) technique for total aldehydes, Fourier-transform infrared
(FTIR) spectroscopy, and the high-performance liquid chromatography (HPLC)
method employing 2,4-dinitrophenylhydrazine (DNPH) derivatization.
3.3.1.2.1 Chromotropic acid method. The chromotropic acid method (CA) involves
the collection of formaldehyde in a midget impinger containing an aqueous
mixture of chromotropic and sulfuric acids, followed by the spectrophotometric
measurement of absorbance of the resulting color (Altshuller and McPherson,
1963; U.S. Dept. of Health, Education and Welfare, 1965). A modification
described by Johnson et al. (1981) improved the accuracy and sensitivity of
the CA method by reducing the concentration of sulfuric acid and by eliminating
a heating cycle, relying solely on the heat of solution generated by sulfuric
acid (Altshuller et al., 1961; Olansky and Deming, 1976). Trapping formaldehyde
in a 1 percent bisulfite solution before adding the CA solution increased
019WPS/B 3-25 6/26/84
-------
collection efficiency from 84 percent to 92 percent with no sulfite interfer-
ences.
The CA method for measuring formaldehyde has been widely studied (Salas
and Singh, 1982; Grosjean and Kok, 1981; National Academy of Sciences, 1981;
Tuazon et al., 1980; Lloyd, 1979). Originally developed as a spot-test by
Eegrlve (1937), it was adopted to quantitate formaldehyde spectrophotometri-
cally (Bricker and Johnson, 1945; West and Sen, 1956) and was modified for
ambient air measurements in the early 1960s (Altshuller et al., 1961;
Altshuller and McPherson, 1963; U.S. Dept. of Health, Education, and Welfare,
1965). While used widely today for both occupational and ambient air environ-
ments, its specificity for formaldehyde, which accounts for approximately half
of total ambient air aldehydes (see section 3.5), limits its usefulness for
characterizing aldehyde concentrations in ambient air.
The CA method has been reported to be sensitive to acrolein, acetaldehyde,
phenol, nitrogen dioxide, and nitrates (National Academy of Sciences, 1981;
Krug and Hirt, 1977; U.S. Dept. of Health, Education, and Welfare, 1973;
Sleva, 1965; Altshuller et al., 1961). Recent work, however, indicates that
neither nitrates, nitrites, NO-, nor acetaldehyde at elevated ambient air
levels interfere with the CA analysis (Johnson et al., 1981; Grosjean and Kok,
1981). Relevant data on other interfering agents were not found.
3.3.1.2.2 MBTH method. A spectrophotometric technique for total aldehydes
was developed in the early 1960s by Sawicki and coworkers (1961). Known as
the MBTH method, it involves the reaction of aldehyde with 3-methyl-2-benzo-
thiazolone hydrazone to form an azine that is oxidized by a ferric chloric-
sulfamic acid solution to form a blue cationic dye (Altshuller, 1983a; U.S.
Dept. of Health, Education, and Welfare, 1965; Hauser and Cummins, 1964;
Altshuller and McPherson, 1963; Altshuller and Leng, 1963; Altshuller et al.,
1961).
The MBTH method has a reported sensitivity of 15 ppb for, primarily,
low-molecular-weight aldehydes (National Academy of Sciences, 1981). The
method is subject to interferences by N02 and gives an inconsistent response
to higher-molecular-weight aldehydes (Sawicki et al., 1961; Altshuller et al.,
1961). Nonetheless, the Intersociety Committee of the American Public Health
Association recommends the MBTH colorimetric method for determining total
aldehydes in air (American Public Health Association, 1977). Miksch and
Anthon (1982) devised a sampling and analysis scheme that permitted a single
019WPS/B 3-26 6/26/84
-------
MBTH sample to be used for both formaldehyde and total aliphatic aldehyde
determinations.
3.3.1.2.3 Fourier- transform infrared spectroscopy. Infrared absorption
spectroscopy has been used by several groups to identify and measure aldehydes
in ambient air (Hanst et al., 1982; Tuazon et al., 1978, 1980, 1981b; Hanst et
al., 1975). These studies employed Fourier- transform infrared (FT-IR) spectro-
meters interfaced to multiple reflection cells operating at total optical
paths of up to 1 km. At such pathlengths, a detection limit of a few ppb was
achieved for formaldehyde. The advantages of the long-path! ength FT-IR method
for ambient air aldehyde measurements (i.e., good specificity and direct j_n
situ analysis) are offset by the relatively high cost and lack of portability
of such systems.
3.3.1.2.4 High-performance liquid chromatography (HPLC) 2,4-dinitrophenylhydra-
zine (DNPH) method. This method takes advantage of the long-
established reaction of carbonyl compounds with 2,4-dinitrophenylhydrazine to
form a 2,4-dinitrophenylhydrazone:
RR'C=0 + NH0NHCCH0(NO,), H,0 + RR'C = NNHC,H,(NtU, (3-1)
£ V 5 £. £. £. O 3 £ £.
Since DNPH is a weak nucleophile, the reaction is carried out in the presence
of acid in order to increase protonation of the carbonyl.
The HPLC-DNPH method is currently the preferred way of measuring aldehydes
in ambient air. Atmospheric sampling is usually conducted with micro-impingers
containing an organic solvent and aqueous, acidified DNPH reagent (Papa and
Turner, 1972; Katz, 1977; Cadle, 1979; Smith and Drummond, 1979; Fung and
Grosjean, 1981). After sampling is completed, the hydrazone derivatives are
extracted and the extract is washed with deionized water to remove the remaining
acid and unreacted DNPH reagent. The organic layer is then evaporated to dry-
ness, subsequently dissolved in a small volume of solvent, and analyzed by
reversed-phase liquid chromatographic techniques employing an ultraviolet (UV)
detection system.
Recently, an improved procedure has been reported that is much simpler
than the above aqueous impinger method (Lipari and Swarin, 1982; Kuntz et al . ,
1980). This scheme utilizes a midget impinger containing an acetonitrile
solution of DNPH and an acid catalyst. After sampling, an aliquot of the
original collection solution is directly injected onto the chromatograph.
019WPS/B 3-27 6/26/84
-------
This approach eliminates the extraction step and several sample-handling
procedures associated with the DNPH-aqueous solution; and provides much better
recovery efficiencies. Lipari and Swarin (1982) have reported detection
limits of 20, 10, 5, and 4 ppb for formaldehyde, acetaldehyde, acrolein, and
benzaldehyde, respectively, in 20-liter air samples.
3.3.1.2.5 Calibration of aldehyde measurements. Since they are reactive
compounds, it is extremely difficult to make stable calibration mixtures of
aldehydes in pressurized gas cylinders. Although gas-phase aldehyde standards
are available commercially, the vendors do not guarantee any level of accuracy.
Formaldehyde standards are generally prepared by one of several methods.
The first method utilizes dilute commercial formalin (37 percent formaldehyde,
w/w). Calibration is accomplished by the direct spiking into sampling impin-
gers of the diluted mixture or by evaporation into known test volumes, followed
by impinger collection. Formaldehyde can also be prepared by heating known
amounts of paraformaldehyde, passing the effluent gases through a methanol-
liquid nitrogen slush trap to remove impurities, and collecting the remaining
formaldehyde.
For the higher-molecular-weight aldehydes, liquid solutions can be eva-
porated or pure aldehyde vapor can be generated in dynamic gas-flow systems
(permeation tubes, diffusion tubes, syringe delivery systems, etc.). These
test atmospheres are then passed through the appropriate aldehyde collection
system and analyzed. A comparison of these data, with the direct spiking of
liquid aldehydes into the particular collection system, provides a measure of
the overall collection efficiency.
3.3.1.2.6 Comparison of measurement methods. Several side-by-side comparisons
of the chromotropic acid method (CA) with other methods have been reported.
Grosjean and Kok (1981) compared the CA method (Johnson et a!., 1981) with
HPLC-DNPH (Fung and Grosjean, 1981) and FTIR spectroscopy (Tuazon et al.,
1978). They found fairly close agreement between the CA and HPLC-DNPH methods,
but noted consistently higher results with FTIR. Corse (1981) sampled ambient
air with CA (U.S. Oept. of Health, Education, and Welfare, 1965), MBTH (U.S.
Dept. of Health, Education, and Welfare, 1965), and HPLC-DNPH methods (Kuntz
et al., 1980). An examination of tabulated data from the Corse study shows a
consistent and considerable difference between CA and HPLC measurements. For
25 CA measurements, formaldehyde averaged 8.8 ppb; while HPLC measurements from
the same sampling train averaged 5.4 ppb higher. Overall, formaldehyde levels
019WPS/B 3-28 6/26/84
-------
were approximately 60 percent higher with HPLC than with CA measurements.
Because blanks were not utilized for the HPLC analyses, however, the HPLC data
are subject to uncertainty, since blank corrections can affect results substan-
tially (Altshuller, 1983). During laboratory studies, Kuntz et al. (1980)
reported reasonable agreement (±7%) among the HPLC-DNPH, CA, and FTIR methods
when low ppb levels of formaldehyde, acetaldehyde, propionaldehyde, hexanal,
and benzaldehyde were generated.
3.3.1.3 Other Oxygenated Organic Species. Literature reports describing the
vapor-phase organic compounds occurring in ambient air indicate that the major
fraction of material consists of unsubstituted hydrocarbons (section 3.5).
Aldehydes as a class of volatile organics appear second in abundance. With
the exception of formic acid (Hanst et al. , 1982; Tuazon et al. , 1981, 1980,
1978), other oxygenated species are seldom reported. The lack of oxygenated
hydrocarbon data is somewhat surprising since significant quantities of these
species are emitted directly into the atmosphere by solvent-related industries
(methanol, ethanol, acetone, etc.; see section 3.4). Along with manmade
emissions, natural sources of oxygenated hydrocarbons also contribute to this
total. In addition to direct emissions, it is also expected that photochemical
reactions of hydrocarbons with oxides of nitrogen, ozone, and hydroxyl radicals
will produce significant quantities of oxygenated products.
Difficulties in sample collection and analysis may account for this
apparent lack of data. The adsorptive nature of the surfaces that contact
these oxygenated species often complicates the process of compound quantita-
tion. Presently, the approach used for analysis of oxygenated and other polar
organic compounds has been to decrease adsorption by deactivating the interior
surfaces of analytical hardware. A novel method has been reported in which
the reactive compounds of interest were modified rather than the surfaces with
which these compounds interact (Osman et al. , 1979; Westberg et al. , 1980).
In these studies, the laboratory derivatization of vapor-phase alcohols and
acids (silylation) was investigated to evaluate the potential of such a proce-
dure for stabilizing these polar compounds prior to analysis. Results have
indicated that silylation procedures greatly reduced adsorption of alcohols
and acids and that, qualitatively, the silylated derivatives could be detected
via the GC-F1D system.
019SPW/A 3-29 6/28/84
-------
3.3.2 Nitrogen Oxides
Aside from the essentially unreactive O, only two oxides of nitrogen
occur in ambient air at appreciable concentrations, nitric oxide (NO) and
nitrogen dioxide (N0?). Both compounds, together designated as NO , participate
^ t\
in the cyclic reactions in the atmosphere that lead to the formation of ozone
(section 4.2).
Analytical methods for N02 and NO that are in current use are briefly
described in this section. Older methods, including the former Federal Refer-
ence Method (Jacobs-Hocheiser method), are described in a recent criteria
document on nitrogen oxides prepared by the U.S. Environmental Protection
Agency (1982).
3.3.2.1 Measurement Methods for N00 and NO. In 1976, the continuous chemi-
luminescence method was promulgated as the new Federal Reference Method.
Other acceptable methods for measuring ambient N0? levels, including two
methods designated as equivalent methods, are the Lyshkow-modified Griess-
Saltzman method, the instrumental colorimetric Griess-Saltzman method, the
triethanolamine method, the sodium arsenite method, and the TGS-ANSA method
[TGS-ANSA = triethanolamine, guaiacol (o-methoxyphenol), sodium metabisulfite,
and 8-anilino-1-naphthalene sulfonic acid]. The sodium arsenite method and
the TGS-ANSA method were designated as equivalent methods in 1977. While some
of these methods measure the species of interest directly, others require
oxidation, reduction, or thermal decomposition of the sample, or separation
from interferences, before measurement. Table 3-8 presents a summary of these
methods.
The current Federal Reference Method measures atmospheric concentrations
of NO,, indirectly by first reducing or thermally decomposing the gas quantita-
tively to NO (with a converter), reacting NO with 0,, and measuring the light
intensity from the chemiluminescent reaction. Two types of converters have
been employed for converting N02 to NO. The first type, e.g., a carbon conver-
ter, actually reduces the N02 to NO. The second, e.g., hot stainless steel,
thermally decomposes the NOp, producing NO. Interfering species will depend
on the type of converter used. The reaction of NO and 03 forms electronically
excited N02 molecules that release light energy that is proportional to the
NO concentration (Fontijn et al., 1970). The NO in the air stream is measured
separately and subtracted from the previous NO (NO plus N0?) measurement to yield
/\ (—•
the N0? concentration. Typical commercial chemiluminescence instruments are
019WPS/B 3-30 6/26/84
-------
TABLE 3-8. METHODS FOR MEASURING NITROGEN DIOXIDE'
Name
Description
Figures of merit
Reference
Chemilumi-
nescence
GO
I
CO
Griess-
Sal tzman
N02 is converted to NO; NO is reacted
with 03, emitting light. Designated
as the Federal Reference Method.
N02 reacts with water to form nitrous
acid, which is reacted with an aroma-
tic amine to form a diazonium salt,
to which an organic coupling agent
is added to form an azo dye. The
amount of N02 collected is related
to absorbence of the solution.
Considered suitable for averaging
times >1 hr.
Lower detection limit is 2.5
|jg/m3 (0.002 ppm). Suitable
for XL-hr averaging times.
Average negative bias of ~5%
with S.D. ± 14% of measured
value for 1-hr averaging time
and range of 50 to 300 |jg/m3
(0.027 to 0.16 ppm).
PAN and other nitrogen com-
pounds, including nitroethane
and nitric acid (HN03) may be
converted to NO.
Ammonia may be converted to NO
if high-temperature converter
is used.
Halocarbons may interfere, if
heated carbon converter is
used.
HN03 and PAN can cause appre-
ciable overestimation of N02
during smog conditions.
Usable range: 19 to 9400 |jg/m3
(0.01 to 5.0 ppm)
Maximum negative bias among dif-
ferent laboratories testing
equivalent samples: 15%, with
S.D. of ± 12% of average value.
Katz (1976); and
U.S. Environmental
Protection Agency (1976a,b)
Constant et al. (1974)
Winer et al. (1974)
U.S. Environmental Protection
Agency (1982)
Joshi and Bufalini (1976)
Spicer (1977a); and
Grosjean (1982b)
Saltzman (1954); and
Constant et al. (1975)
-------
TABLE 3-8. METHODS FOR MEASURING NITROGEN DIOXIDE11 (continued)
Name
Description
Figures of merit
Reference
Triethanol- N02 is absorbed in a solution of
amine -triethanolamine and n-butanol
surfactant; analyzed by Griess-
Sal tzman reagent, producing an azo
dye for spectrophotometric measure-
ment. Considered suitable as a
24-hr method.
Sodium Designated as an equivalent method,
arsenite considered suitable for 24-hr
measurement. N02 is absorbed in an
alkaline solution of sodium arsenite,
then acidified with phosphoric acid;
azo dye is formed by addition of sul-
fanil amide N-(l-naphthyl) ethylene-
diamine dihydrochloride.
Collection efficiency in the
range of 30 to 700 ug/m3 (0.01
to 0.37 ppm):
1. When glass frits are used:
80%.
2. When restricted orifices are
are used: 50%.
Bias error at a stoichiometric
factor of 0.764 is <2%.
Collection efficiency from 20 to
750 ug/m3 (0.01 to 0.4 ppm): 82%
Negative bias among different
laboratories testing equivalent
samples: 3% with S.D. of
± 11 ug/m3.
Increase in N02 response from NO
in the range of 50 to 310 ug/m3
(0.04 to 0.25 ppm): maximum
10 ug/m3.
Increase in N02 response from
C02 in the range of 360,000
to 899,000 ug/m3 (200 to 500
ppm): 50 to 250 ug/m3 (0.02
to 0.13 ppm).
Ellis and Margeson, 1974
Scaringelli et al., 1970
Beard and Margeson, 1974
Constant et al., 1975
Beard et al., 1975
Beard et al., 1975
-------
TABLE 3-8. METHODS FOR MEASURING NITROGEN DIOXIDE3 (continued)
Name
Description
Figures of merit
Reference
TGS-ANSA
CO
I
CO
CO
A 24-hr manual method, designated
an equivalent method. Ambient air
is bubbled through a solution of
triethanolamine, o-methoxyphenol,
and sodium metabisulfite. N02 is
converted to nitrite ion, which is
assayed by diazotization and
coupling using sulfanilamide and
the ammonium salt of 8-anilino-
1-naphthalene-sulfonic acid.
Absorbence is read at 500 nm.
[TGS-ANSA = triethanolamine,
guaiacol, sodium metabisulfite
(TGS) and 8-anilino-1-naphthalene
sulfonic acid (ANSA).]
Collection efficiency with addi-
tion of o-methoxyphenol: 95%
At N02 concentration of
100 pg/m3 (0.05 ppm); no
interferences from ammonia
(25 ug/m3); C02 (154,000 pg/m3);
formaldehyde (750 |jg/m3); NO
(734 |jg/m3); and S02 (439 pg/m3)
Lower detection limit,
<15 pg/m3 (0.008 ppm).
Average bias over the range of
50 to 300 ug/m3 (0.03 to 0.16
ppm) is 9.5 (jg/m3 (0.005 ppm).
Interlaboratory standard devia-
ation is ± 8.8 pg/m3 (0.004 ppm).
Nash, 1970; Fuerst and
Margeson, 1974
Fuerst and Margeson, 1974
Mulik et al., 1973
Constant et al., 1974
Adapted from U.S. Environmental Protection Agency (1982).
-------
capable of detecting levels as low as 2.5 ug/m3 (0.002 pprn) (Katz, 1976).
Winer et al. (1974) found that peroxyacetyl nitrate (PAN) and various nitrogen
compounds were also reduced by the converter to NO and that nitroethane and
nitric acid were partially reduced in the system when carbon (reducing) conver-
ter or a molybdenum (thermal decomposition) converter was used. Joshi and
Bufalini (1976) reported positive interferences from halocarbons when a heated
carbon converter was used; they also suggested that stainless steel converters
may be subject to interferences from chlorinated hydrocarbons. Other evidence
suggests that ammonia (NH3) may be converted to NO in high-temperature thermal
converters (U.S. Environmental Protection Agency, 1982). Spicer et al. (1979)
reported that positive interference with chemiluminescent NO- measurements can
be significant under smoggy conditions. Positive interferences resulting from
the presence of PAN and HN03 on afternoons of high oxidant concentrations can
exceed 30 percent of the N02 concentrations. Grosjean (1982) also reported
that positive interferences from nitric acid and PAN during N0» analysis by
chemiluminescence can cause a 50- to 60-percent N0? overestimation during smog
conditions in Los Angeles. In less severe smog, the overestimation should not
be this high.
In addition to the wet chemical methods for measuring NO,,, other tech-
niques have been investigated. Maeda et al. (1980) reported a new chemilumines-
cence method based on the reaction of N0? with luminol (5-amino-2,3-dihydro-l,
4-phthalazine dione), with a detection limit of about 50 parts per trillion
(ppt) and linearity over a range of 0.5 ppb to 100 ppm. Work is under way to
remove the interferences of 0, and SO^. Other workers have endeavored to
improve chemiluminescence analyzers through physical modifications (Ridley and
Hewlett, 1974; Schiff et al., 1979; Stedman et al., 1977). Molecular correla-
tion spectrometry, in which an absorption band of a sample is compared with a
corresponding band stored in the spectrometer, has been applied in analysis of
N0? (Williams and Kolitz, 1968). Instruments processing the second derivative
of sample transmissivity have also been used (Hagar and Anderson, 1970), as
have infrared lasers and infrared spectrometers (Hanst, 1970; Hinkley and
Kelley, 1971; and Kreuzer and Patel, 1971). Tucker et al. (1973, 1975) report-
ed on instruments based on the principle of laser-induced fluorescence at
optical frequencies. Fincher et al. (1977) described detection of I ppb N02
with a technique based on fluorescence by a pulsed xenon flashlamp. Long-
path! ength differential optical absorption spectroscopy has also been employed
to monitor N02 in the troposphere (Platt et al., 1984; 1980).
019SPW/A 3-34 6/28/84
-------
As summarized from the criteria document for nitrogen oxides prepared by
the U.S. Environmental Protection Agency (1982), methods for measuring NO
directly include ferrous sulfate absorption and spectrophotometric measurement
of the resulting ion (Norwitz, 1966), ultraviolet spectroscopy (Sweeny et al.,
1964), and infrared spectroscopy (Lord et al., 1975). Mass spectrometry and
gas chromatography may also be employed. None of the above techniques, however,
is widely used to monitor air quality. The best direct method for measuring NO
is chemiluminescence, employing ozone as the reactant (equation 4-3, chapter 4)
(Fontijn et al. , 1970).
3.3.2.2 Sampling Requirements. When sampling for NO , long residence times
"~'J~ /\
in sampling lines should be avoided. In the ambient air, the rate of photoly-
sis of N0~ (forming NO and 0, and thus 0-) is almost equal to the rate of the
reaction of the NO and 03 to form N0«. In sampling lines, photolysis stops
but NO continues to react with 03, producing NOp. The magnitude of the dark
reaction of NO with 0_ depends, of course, on the concentrations of NO and 0,
in the sample being analyzed, as well as on the residence time of the sample
in the line. This dark reaction has greater practical consequences in some
situations than in others. In moderately polluted urban areas, steady-state
concentrations of NO are almost certainly too low at the period of maximal 0,
to cause significant errors in obtaining NO or 03 measurements. Conversely,
when 0_ is at a minimum, as in the early morning or possibly even in the late
•3
afternoon, NO (and N0?) may be at maximal levels, and no significant errors
would be introduced. If the concentrations of NO and 03 are both low, however,
as in some rural areas, or during those brief periods in polluted areas when
NO and 0., diurnal patterns cross, then significant measurement errors could be
introduced. Values for N0? would be erroneously high and values for 03, if
simultaneous measurements of 03 were being attempted, would be erroneously
low.
Techniques for limiting errors from sampling to given levels of tolerance
are reviewed by Butcher and Ruff (1971). In general, only glass or Teflon
materials should be used in sampling trains. Among absorbents, granules im-
pregnated with triethanolamine are reported to be the best, converting only 2
to 4 percent of the incoming N02 to NO (Intersociety Committee, 1977; Huygen,
1970). The most frequently used oxidizer is chromic oxide on a fire-brick
granule support (Intersociety Committee, 1977; Levaggi et al., 1974).
019SPW/A 3-35 6/28/84
-------
3.3.2.3 Calibration. Procedures for calibrating measurement methods used to
determine NO are critical for obtaining accurate analyses. Monitoring instru-
^\
ments may be calibrated either by measuring a gas of known concentration or by
comparing measurements of gas from a stable source with measurements made by a
primary reference method. Measurement methods for NO and N0» are calibrated
principally by standard reference materials (SRM). In a National Bureau of
Standards study, the initial accuracy and stability of standard mixtures of NO
in nitrogen were found to be high (Hughes, 1975). Other sources include
permeation of compressed NO through membranes to produce dilute NO streams
(U.S. Environmental Protection Agency, 1976), electrolytic generation (Hersch
and Deuringer, 1963), catalytic reduction of NO™ (Breitenbach and Shelef,
1973), and photolysis and rapid dilution of N02 (Guicherit, 1972). Of these
standard reference materials, only compressed NO and the N0? permeation tube
are currently used. The U.S. Environmental Protection Agency (1976a,b) recom-
mended the combined use of permeation tubes and gas-phase titration, using one
technique to check the other. The two standard reference materials available
for generating known concentrations of NO and NO^ are the cylinder of compressed
NO in N? (50 and 100 ppm) and the N02 permeation tube. Both are commercially
available and are traceable to SRMs at the National Bureau of Standards.
The preparation of standard mixtures of NO in nitrogen has been studied
by the National Bureau of Standards (Hughes, 1975). The initial accuracy with
which standards may be prepared, based on either pressure or mass measurements,
is quite good. The stability of mixtures at concentrations above about 50 ppm
was found to be satisfactory; the average change in concentrations over a
7-month period was only 0 to 1 percent.
The permeation tube is the only direct source of dilute NOp mixtures in
widespread use (O'Keefe and Ortman, 1966; Scaringelli et al., 1970). It may
be calibrated by weighing or, though rarely done, by micromanometric measure-
ments. The other common procedure used to calibrate NO- measurement instru-
ments is gas-phase titration. Stable sources of known concentrations of both
NO and 0,. are required. A dilute stream of NO is measured by NO methods. The
0- is added to the stream at a constant rate. The decrease in NO that occurs
•3
through its reaction with the added ozone is equal to the NO- formed. Thus, a
known N0? concentration is produced.
019SPW/A 3-36 6/28/84
-------
3.4 SOURCES AND EMISSIONS OF PRECURSORS
3.4.1 Manmade Sources and Emissions
This section presents information on the manmade sources and emissions of
precursors to ozone and other photochemical oxidants. There are no major or
pervasive primary sources of ozone or of related photochemical oxidants. In
isolated instances, intrusion of stratospheric ozone into the troposphere may
briefly make a significant local contribution to the natural background con-
concentration of ground-level ozone (chapter 4). Human activities generate no
significant direct contribution to ambient ozone concentrations. Consequently,
ozone and other photochemical oxidants are almost exclusively secondary pollu-
tants arising from the emission and reactions of primary pollutants.
It should be noted that the relationship between emissions of the precur-
sors and the resulting ozone levels is neither direct nor constant. Thus, the
magnitudes of source emissions cited here are best viewed as indicating a poten-
tial for ozone production rather than as determining or predicting the extent
of the ozone problem in any particular area.
Three recent documents, which have been extensively reviewed by the
scientific community, discuss sources of nitrogen oxides, aldehydes, and
hydrocarbons and other volatile organic compounds. These are Air Quality
Criteria for Oxides of Nitrogen (U.S. Environmental Protection Agency, 1982);
Formaldehyde and Other Aldehydes (National Academy of Sciences, 1981); and
Review of Criteria for Vapor-Phase Hydrocarbons (Tilton and Bruce, 1981).
Wherever appropriate, material is quoted from these documents; updated material
is included in the few cases in which significant new information has appeared
since their publication.
Many volatile organic compounds besides hydrocarbons may participate in
photochemical smog reactions. The Federal reference method for nonmethane
hydrocarbons uses a flame ionization detector as the sensing element, which is
not specific for hydrocarbons but also responds to varying degrees to other
nonmethane organic compounds (section 3.3). For these reasons, the U.S.
Environmental Protection Agency has, in recent years, subsumed hydrocarbon
source and emissions data into the broader category of volatile organic com-
pounds (VOC). Several compounds are deliberately excluded because their
contribution to the production of photochemical oxidants in the lower tropos-
phere is considered negligible. The principal compound, methane, constitutes
a significant fraction of hydrocarbons in ambient air; the others, ethane,
019SPW/A 3-37 6/28/84
-------
methylene chloride, and several halomethanes and haloethenes, are also excluded
from emission inventories (U.S. Environmental Protection Agency, 1980b).
The preparation of an emission inventory involves the compilation of a
series of estimates and thus is subject to the errors inherent in any inventory.
There are two broad categories of emission inventories: (1) collective nation-
wide estimates based on average annual emissions in the respective general
source categories; and (2) detailed inventories prepared for the purpose of
developing control strategies. The latter is prepared for specific localities,
such as a metropolitan area, and accounts for influences such as industries
that are seasonal, and local meteorology. The data discussed here are in the
first category and were compiled for the purpose of identifying nationwide
trends. In these collective, annual, nationwide inventories, random error
components will tend to cancel out, but systematic error or bias will remain.
The direction of this systematic error is probably toward the low side because
of overlooked sources. This does not impair the use of these estimates for
trend analysis, because as the inventory procedures have been refined, esti-
mates for prior years have been recalculated. Thus, the discussion of trends
in emissions below is based on data in which the remaining biases are consis-
tent for all years.
Inventories for VOC emissions are less reliable than those for most of
the criteria pollutants. Emission inventory procedures for stationary sources
were originally developed to estimate emissions of particulate matter and
sulfur dioxide. As inventories for additional pollutants were compiled,
existing procedures were applied without much change. In some cases, this was
appropriate (e.g., for oxides of nitrogen). Because of the nature of VOC
sources, which produce a higher percentage of fugitive emissions that are hard
to account for, the use of historical procedures for VOC emission sources has
been found to be generally inadequate. Improved procedures have recently been
developed that account more completely for emissions from the ubiquitous
sources of VOC (U.S. Environmental Protection Agency, 1980a).
These annual emission estimates cannot be scaled to shorter time periods
or to individual areas of the country because seasonal variations occur in
some source emissions and because the roster of sources obviously varies from
city to city. The second type of emission inventory mentioned above is appli-
cable where attention to these details is required (U.S. Environmental Protec-
tion Agency, 1981).
019SPW/A 3-38 6/28/84
-------
3.4.1.1 Trends in Emissions of Volatile Organic Compounds. Although not all
volatile organic compounds have the same potential for oxidant formation,
estimates of their total emissions nevertheless provide a gross measure of
compounds available for photochemical production of ozone and other photochemi-
cal oxidants. Figure 3-3 shows national trends in emissions of volatile
organic compounds (VOC) by general source category for the period 1970 through
1981. Total VOC emissions nationwide were 22 percent lower in 1981 than in
1970. The main sources nationwide are industrial processes, which emit a wide
variety of VOCs such as chemical solvents; and transportation, which involves
the emission of VOCs in gasoline vapor as well as in gasoline combustion
products. Industrial process emissions peaked in 1978 and 1979, while emis-
sions from transportation sources decreased about 35 percent despite a 42-percent
increase in total vehicle miles driven. Decreases also occurred in the solid
waste and miscellaneous categories.
3.4.1.2 Trends in Emissions of Nitrogen Oxides. Total national NO emissions
"-'-1 n - -~ •'"-—' '-— T ' " " '-" - '- - " ~T --..J.--— •-. ". -T-.U1- -run ^
in 1982 were almost 12 percent above the 1970 rate, but appear to have declined
slightly from 1978 and 1979 levels (Figure 3-4). The increase over the period
1970 through 1982 may be attributed primarily to two causes: (1) increased
fuel combustion in stationary sources such as power plants; and (2) increased
fuel combustion in highway motor vehicles, as the result of the increase in
vehicle miles driven. Total vehicle miles driven increased by 42 percent over
the 13 years in question. Emissions associated with industrial processes
remained relatively constant, but solid waste and miscellaneous emissions
decreased slightly.
The national trends shown do not reflect the considerable local and
regional differences that exist in the relative amounts of N0x emitted in the
major source categories. For example, motor vehicle emissions in Los Angeles
County, California, increased sixfold from 1940 to 1970 (Los Angeles County,
1971), compared to a threefold national increase.
Although they are minor on a national level, industrial process losses
(NO emissions from noncombustion industrial sources) can be important near
s\
individual local sources. The principal activities in this source category
are petroleum refining and the manufacture of nitric acid, explosives, and
fertilizers.
Aircraft are not considered a major source of NO on the national scale,
/\
but their impact in the immediate vicinity (at a radius of up to 10 miles) of
019WPS/B 3-39 6/26/84
-------
30
20
ro
(fl
«3
LU
O
o
10
1 I
TRANSPORTATION
IND. PROC., STAT. SOURCE
SOUpJ/VASTE
NON-IND~SOLVENTs"
r—r
i
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981
YEAR
Figure 3-3. National trend in estimated emissions of volatile organic com-
pounds, 1970 through 1982.
Source: U.S. Environmental Protection Agency (1982b).
3-40
-------
25
20
O>
". 15
(A
(A
(A
i
Ul
X
i 10
i—i—i—i—i—i i i r
TRANSPORTATION
FUEL COMBUSTION
IND. PROC., SOLID WASTE. MISC. ] ' T~\
I I
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981
YEAR
Figure 3-4. National trend in estimated emissions of nitrogen oxides, 1970
through 1982.
Source: U.S. Environmental Protection Agency (1982b).
3-41
-------
major airports has been discussed by George et al. (1972) and by Jordan and
Broderick (1979).
Figure 3-5 compares the relative trends in mobile source NO and VOC
f\
emissions versus the trend in vehicle miles traveled, both total and in urban
areas, all referenced to the base year 1970 (U.S. Environmental Protection
Agency, 1983; U.S. Department of Transportation, annual publications). By
1982, VOC emissions from mobile sources were only about 60 percent of their
1970 level. Mobile source emissions of NO were about 28 percent higher in
)\
1981 than in 1970, but a small downward trend began in 1979.
3.4.1.3 Geographic Distribution of Manmade Emissions of Volatile Organic
Compounds. An overall picture of the density of VOC emissions for all counties
in the coterminous United States, as of February 1978 (U.S. Environmental
Protection Agency, 1978), is given in Figure 3-6. Areas of high emission
density are apparent in Southern California; in the Northeast Corridor extend-
ing from the greater Washington, D.C., area to the greater Boston area; in
many counties bordering the Great Lakes; along the Gulf Coast of Texas and
Louisiana; and in many other eastern counties. The subset of area-source VOC
emissions is mapped in Figure 3-7; the major component in this category is
emissions from vehicles.
3.4.1.4 Geographic Distribution of Manmade Emissions of Nitrogen Oxides. An
overall picture of the nationwide distribution of NO emissions can be obtained
}\
from the maps of the United States reproduced in Figures 3-8 and 3-9.
Figure 3-8 shows total NO emissions by United States counties as compiled in
/\
the National Emissions Data System (NEDS) file of February 1978 (U.S. Environ-
mental Protection Agency, 1978). Regions of high source emissions are evident
near populous and industrial areas. Figure 3-9 shows the subset of area-source
emissions, which is dominated by emissions from vehicles.
3.4.1.5 Profiles of Emissions of Volatile Organic Compounds. Hundreds of
volatile organic compounds have been detected in the ambient air of urban
areas, most of which can participate in photochemical reactions. Some limita-
tions on the accuracy of emission inventories for these hydrocarbons and other
VOC were mentioned in section 3.4.1. Additional factors such as atmospheric
chemical reactions, variations in meteorological dispersion, and varying
three-dimensional emission source distributions result in complex relationships
between even comprehensive and accurate emission inventories and actual ambient
air concentrations of VOCs.
019WPS/B 3-42 6/28/84
-------
OJ
I
en
Tons / Sq Mi
30 - 100
S 100
Figure 3-7. Area source volatile organic compound emissions by county in
the coterminous United States, 1978.
SASD OAOPS 01 Apr 1983
Source: U.S. Environmental Protection Agency, National Emissions Data
System.
-------
.i-^-c * • <<\
' -
Figure 3-8. Total NOX emissions by county in the coterminous United SASD OAQPS 0) Apr 198J
States, 1978.
Source: U.S. Environmental Protection Agency, National Emissions Data
-------
-\
/ Sq Mi
Figure 3-9. Area source NOX emissions by county in the coterminous
United States, 1978.
SASD OAQPS 01 Apr 1983
Source: U.S. Environmental Protection Agency, National Emissions Data
Systems.
-------
3.4.1.5.1 Ambient air profiles and source reconciliation. The comparison of
chemical species in ambient air with species actually emitted by respective
sources is often referred to as "source reconciliation." Source reconciliation
techniques can be used along with dispersion and other models to help validate
emission inventories. More often, source reconciliation techniques are used
to identify the relative contributions of various sources to the observed
ambient pollutant mixture, based on the characteristic emission profiles of
individual source categories.
Composition profiles of hydrocarbons (HC) in ambient air have been report-
ed and compared with emission profiles of known sources by a number of investi-
gators (Neligan, 1962; Stephens and Burleson, 1969; Lonneman et al. , 1974;
Siddiqi and Worley, 1977; Kopczynski et al., 1975; Mayrsohn and Crabtree,
1975; Mayrsohn et al., 1977; Crabtree and Mayrsohn, 1977; and Seila, 1979).
For example, Lonneman et al. (1974) measured ratios of ethylene, isobutane,
n-butane, isopentane, and n-pentane to the nonreactive compound, acetylene,
which is considered a tracer of auto emissions. Subsequent investigations by
other authors (Kopczynski et al. , 1975) confirmed these as characteristic of
automotive sources and estimated, for example, that, overall, fewer than 50
percent of the alkanes and aromatic hydrocarbons in ambient air in St. Louis
were related to automotive emissions; whereas most of the alkenes in the
evening and early morning hours were automotive-related. Using a multivariate
regression technique, Mayrsohn and Crabtree (1975) estimated that the sources
of the average distribution of nonmethane hydrocarbons (NMHC) in Los Angeles
were as follows: 47 percent automotive exhaust; 31 percent gasoline; 8 percent
commercial natural gas; and 14 percent geogenic natural gas. Thus, automotive-
related sources contributed about 78 percent of the C2 to C-.Q NMHC in 1973 in
Los Angeles. These investigators subsequently extended their analysis to
eight sampling sites between Los Angeles and Palm Springs and obtained similar
results: automotive exhaust, 53 percent; gasoline, 12 percent; gasoline
vapor, 10 percent; commercial natural gas, 5 percent; geogenic natural gas, 19
percent; and liquified petroleum gas, 1 percent. In a similar type of study
in Houston, Texas, Siddiqi and Worley (1977) concluded that both automotive
and industrial sources were significant contributors to ambient air NMHC in
the Houston area, but that in downtown Houston, automotive sources dominated.
There is some question, however, regarding their methods (Lonneman and
Bufalini; 1978) and their interpretation (Crabtree and Mayrsohn, 1977). In a
019WPS/B 3-48 6/28/84
-------
state forest about 38 miles north of Houston, Seila (1979) found that automo-
tive sources accounted for 35 percent of the NMHC burden. Of the nonautomotive
NMHC sources, the author concluded that 41 percent originated in Houston and
59 percent from sources north of Houston. Biogenic NMHC emissions were report-
ed as accounting for only 2 percent of the NMHC loading. (Emissions of NMHC
from vegetation will be discussed in section 3.4.2.)
3.4.1.5.2 Stationary sources of volatile organic compounds. The source
category contributing the largest percentage of VOC emissions in 1982, 39
percent, is Industrial Processes (Table 3-9). The category consists almost
entirely of point sources. The composition of these emissions varies widely,
depending on the process or product and the use of emission reduction equipment
and operating practices.
The second largest VOC source category, Transportation, accounting for
33.5 percent of the annual total in 1982, is discussed in the section below on
mobile sources.
The third largest VOC source category, Miscellaneous, accounts for 13.2
percent of the annual total, over half of which consists of the subcategory,
Miscellaneous Organic Solvents. These emissions generally qualify as area-
source emissions. One group of solvents is widely used in domestic products
such as furniture polish, shoe polish, shaving soap, perfumes, cosmetics,
shampoo, hair spray, hand lotion, rubbing alcohol, and nail polish remover.
The predominant compounds emitted are isopropyl alcohol and ethyl alcohol
(Bucon et al., 1978).
3.4.1.5.3 Mobile sources of volatile organic compounds. Emissions of volatile
organic compounds from the production and marketing of gasolines and motor
oils are classed as stationary source emissions and are included in the 1.5
tg/year of VOCs emitted by the petroleum refining industry (Table 3-9).
Following their sale to vehicle owners, these products generate some 7.7
tg/year of mobile source VOC emissions.
Although a significant portion of mobile source VOC emissions arises from
additional evaporation, the most conspicuous mobile source emissions are the
combustion products. Black et al. (1980) reported emission factors for over
60 individual hydrocarbons in both tailpipe and evaporative emissions of four
passenger cars: a 1963 Chevrolet (model unspecified), a 1977 Mustang, a 1978
Monarch, and a 1979 LTD-II. The vehicle tests involved four gasoline fuels of
varying composition. The four passenger cars for which emission data were
019WPS/B 3-49 6/28/84
-------
TABLE 3-9. NATIONAL ESTIMATES OF VOLATILE ORGANIC COMPOUND EMISSIONS, 1982
Source category
Weight emitted,
tg/yr
Percent
Transportation
Highway vehicles
Aircraft
Rai1 roads
Vessels
Other off-highway vehicles
Transportation: total
Stationary source fuel combustion
Electric utilities
Industrial
Commerci al-i nsti tuti onal
Residential
Fuel combustion: total
Industrial processes
Chemicals
Petroleum (crude and products)
Oil and gas
Industrial solvent use
Other
Industrial: total
Solid waste disposal
Incineration
Open burning
Solid waste: total
Miscellaneous
Forest fires
Other burning
Miscellaneous organic solvent
Miscellaneous: total
Total
(4.8)
(0.2)
(0.2)
(0.4)
(0.5)
6.1
(0.0)
(0.1)
(0.0)
(1.9)
2.0
(1.8)
(1.4)
(1.4)
(2.4)
(0.1)
7.1
(0.3)
(0.3)
0.6
(0.8)
(0.1)
(1.5)
2.4
18.2
33.5
11.0
39.0
3.3
13.2
100.0
Source: U.S. Environmental Protection Agency (1983).
019WPS/B
3-50
6/28/84
-------
given by Black et al. (1980) represent a wide range of exhaust and evaporative
emission control configurations. The authors concluded that evaporative
emissions constituted a significant fraction (one-third to one-half) of total
hydrocarbon emissions from all of the tested vehicles. Evaporative hydrocarbon
emissions were relatively more abundant in alkanes than the tailpipe emissions.
Uncombusted fuel was responsible for most of the aggregate hydrocarbon emissions
above carbon number 4; combustion products dominated below carbon number 4.
(This report presents detailed emission data, to which the reader is referred.)
Generally, catalytic tailpipe control systems were more effective in reducing
the amount of unsaturated than saturated hydrocarbon emissions. Evaporative
control devices theoretically should control the more volatile compounds
(generally <_ C.), but in this instance the impact of the devices on the compo-
sition of emissions was not clear.
Table 3-10 gives hydrocarbon exhaust emission factors (in g/mi) for
gasoline-powered, light-duty vehicles (excluding California and high-altitude
models), for model years 1968 through 1980 and by calendar year of operation.
The effects of the age of the vehicle and of control devices are apparent.
(Note the large drop in emissions with the 1975 model year when catalytic
converters were first installed). The composite crankcase and evaporative
hydrocarbon emission rate for these vehicles declined from 2.53 g/mi in 1968
through 1970 to 0.15 g/mi in 1980 (Fisher, 1980).
The average composition of gasoline vapor, determined from weighted
averages of gasoline blending stocks and vapor pressures of respective com-
pounds, consists primarily of alkanes: n-butane (38.1 vol %), isopentane
(22.9 vol %), ri-pentane (7.0 vol %), and isobutane (5.2 vol %). The remaining
individual alkanes and the collective alkenes and aromatics each account for
less than 5 percent by volume of the evaporative emissions from gasoline
(PEDCo, 1978). It should be noted that various gasoline blends from about six
different blending stocks are used to tailor gasoline characteristics to suit
differing climatic regions of the United States. Evaporative emissions from
diesel vehicles are negligible because of low fuel volatility (Linnel and
Scott, 1962; McKee et al. , 1962). Exhaust emissions from gasoline-fueled
vehicles typically contain fuel components and low-molecular-weight hydrocar-
bons that are not present in the fuel. The predominant hydrocarbons in auto-
mobile exhaust, as reported in three separate studies, are shown in Table 3-11.
019WPS/B 3-51 6/28/84
-------
I
\ TABLE 3-10. HYDROCARBON EXHAUST EMISSION FACTORS3
03 FOR LIGHT-DUTY, GASOLINE-POWERED VEHICLES FOR
co
i
en
ro
ALL AREAS EXCEPT CALIFORNIA AND HIGH-ALTITUDE
(9/miD)
Model
Year
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1970 1971 1972
4.3 5.0 5.
3.5 4.3 5.
2.7 3.5 4.
2.7 3.
2.
7
0
3
5
7
July,
1973
6.3
5.7
5.0
4.3
3.5
2.7
calendar year
1974
6.
6.
5.
5.
4.
3.
2.
9
3
7
0
3
5
7
1975
7.4
6.9
6.3
5.7
5.0
4.3
3.5
1.3
of operation
1976
7.9
7.4
6.9
6.3
5.7
5.0
4.3
1.6
1.3
1977
8.3
7.9
7.4
6.9
6.3
5.7
5.0
1.9
1.6
1.3
1978
8.7
8.3
7.9
7.4
6.9
6.3
5.7
2.2
1.9
1.6
1.3
1979
9.1
8.7
8.3
7.9
7.4
6.9
6.3
2.5
2.2
1.9
1.6
1.3
1980
9.4
9.1
8.7
8.3
7.9
7.4
6.9
2.8
2.5
2.2
1.9
1.6
0.3
aEmission factors for vehicles through model year 1975 and through calendar year
1975 are based on actual surveillance tests of in-use vehicles. Post-1975 calendar
year factors for all model-year vehicles are projected. Deterioration factors used
are: pre-1968, 0.58; 1968-1974, 0.53; 1975-1979, 0.23; and > 1980, 0.23—all
in g/mi per 10,000 miles of travel.
To convert g/mi to g/km divide g/mi by 1.609.
Source: Fisher (1980).
en
•**.
ro
oo
-------
TABLE 3-11. PREDOMINANT HYDROCARBONS IN EXHAUST EMISSIONS
FROM GASOLINE-FUELED AUTOS
Fraction of total HC, vol %
Hydrocarbon (HC)
Methane .
Ethyl ene .
Acetylene,
Propylene
n-Butane
Isopentane
Toluene.
Benzene
n-Pentane
m- + p_-Xylene
1-Butene
Ethane
2-Methylpentane
n-Hexane
Isooctane
All others
62-Car
survey
16.7
14.5
14.1
6.3
5.3
3.7
3.1
2.4
2.5
1.9
1.8
1.8
1.5
1.2
1.0
22.0
15-Fuel
study
18
17
12
7
4
4
5
NAC
NA
NA
3d
NA
NA
NA
NA
30
Engine- variable
study
13.8
19.0
7.8
9.1
2.3
2.4
7.9
NA
NA
2.5
6.0d
2.3
NA
NA
NA
26.9
Variables were air:fuel ratio and spark timing.
Combustion products.
CNA = Data not available.
Includes isobutylene.
Source: National Academy of Sciences, 1976.
019WPS/B
3-53
6/28/84
-------
The use of catalytic converters has had a pronounced effect not only on
the amount of hydrocarbons emitted from automobiles but also on the actual
species of hydrocarbons. In general, oxidation catalysts have resulted in an
increase in the percentage of the less reactive alkanes, especially methane,
and a decrease in the percentage of alkenes and acetylene. Typically, exhaust
from a catalyst-equipped automobile contains about 62 percent alkanes, 17
percent aromatics, 18 percent alkenes, and 3 percent acetylene. This may be
compared with the corresponding typical values for automobiles without conver-
ters: 40, 24, 26, and 11 percent, respectively. Methane levels generally
range from about 10 to 30 percent (Black and Bradow, 1975; Black, 1977).
Exhaust gases from gasoline-fueled vehicles also contain nonhydrocarbon organic
compounds such as aldehydes, ketones, ethers, esters, acids, and phenols,
amounting to as much as one-tenth of the total hydrocarbon content.
Factors other than gasoline composition influence the composition of
exhaust. These include driving patterns, the specific configuration of emis-
sion control devices, ambient temperature and humidity, and, of course, indivi-
dual automobile parameters such as tuning, make, and model year. Fuel addi-
tives can also influence emissions. For example, in one study, tetraethyl
lead increased hydrocarbon emissions by about 5 percent but did not change the
type of emissions (Leihkanen and Beckman, 1971).
Emissions from diesel automobiles are the subject of two recent papers
(Black and High, 1979; Gibbs et al. , 1983). Black and High (1979) described
the complexities of accounting for both the gaseous and the condensed or
particle-bound hydrocarbons in the cooling exhaust stream on its way through
the exhaust system. They reported total hydrocarbon (THC) emissions of 0.29,
0.35, and 0.45 g/mi from a turbo-Rabbit, a Nissan-Datsun, and an Oldsmobile,
respectively. From 15 to 40 percent of these hydrocarbons were associated
with particles by the time the exhaust stream exited the tailpipe. Gibbs et
al. (1983) have reported THC emissions from 19 in-use diesel automobiles,
representing 1977 to 1979 model years, that were tested periodically over a
28-month period. Emissions of THC at the end of the period ranged from 0.17
to 0.88 g/mi for individual vehicles and averaged 0.65 g/mi. It should be
noted here that the population of diesel-powered passenger cars is not growing
as rapidly as was once expected. Sales of diesel-powered cars peaked at
6 percent for 1981 models and dropped to less than 3 percent by 1983 (Automo-
tive News, 1982a,b,; 1983a,b,c).
019WPS/B 3-54 6/28/84
-------
Dietzman et al. (1980, 1981) recently reported on emissions from both
gasoline- and diesel-powered trucks, using the chassis version of the 1983
transient heavy-duty engine test procedure. This procedure permits a variety
of comparisons between emissions from gasoline and from diesel engines.
Hydrocarbon emission rates are slightly higher when minimum quality DF-2 fuel
is used with the three 4-stroke engines tested. No differences were observed
with the 2-stroke (DD 8V-71) engine.
A summary of both exhaust and evaporative emission characteristics of a
variety of fuel types is presented in Table 3-12 (Tilton and Bruce, 1981).
3.4.1.6 Profiles of Emissions of Nitrogen Oxides. Fuel combustion is the
dominant source of NO emissions nationally. Stationary sources contribute
/\
51.8 percent and mobile sources contribute 43.8 percent (1981 estimates, see
Table 3-13). In contrast to their contributions to VOC emissions (Table
3-9), industrial process sources contribute only about 3 percent to the
national total NO emissions.
A
Table 3-14 documents NO/NO ratios in emissions from a variety of source
s\
types. Nitric oxide (NO) is the dominant oxide of nitrogen emitted by most
sources; N0? generally comprises less than 10 percent of the total NO emis-
£. /\
sions. Note, however, that N0? forms upwards of 30 to 50 percent of the total
NO emissions from certain diesel and jet turbine engines under specific load
/\
conditions. Tail gas from nitric acid plants, if uncontrolled, may contain
about 50 percent N0_. The variations in NO/NO ratios by source type reported
£. r*.
in this table may be significant in local situations, as, for example, in the
immediate vicinity of a high-volume roadway carrying a significant number of
diesel-powered vehicles.
The combustion process converts some nitrogen from both the combustion
air and the fuel into nitrogen oxides. To date, the most cost-effective
procedures for reducing NO emissions from stationary combustion sources
x\
involve modification of combustion conditions rather than denitrification of
fuels or treatment of flue gases (Hall and Bowen, 1982).
The principal categories of stationary combustion sources are electric
utility boilers, industrial boilers, and industrial process heaters.
Baseline NO emissions from electric utility boilers were reported by
Bartok et al. (1971, cited in Hall and Bowen, 1982) as 994 ppm NO from coal-
fired units; 589 ppm NO from gas-fired units; and 360 ppm NO from oil-fired
^\ ft.
units. Combustion modifications achieved reductions in gas- and oil-fired
019WPS/B 3-55 6/28/84
-------
TABLE 3-12. SUMMARY OF EMISSION CHARACTERISTICS FOR AUTOS FUELED BY GASOLINE,
DIESEL, AND ALCOHOL-GASOLINE OR ETHER-GASOLINE BLENDS
Auto, control device,
or fuel
Emission characteristics
Gasoline-fueled,
uncontrolled
Gasoline-fueled,
catalytic converters
Lead additives
gasoline
Ethanol-gasoline
blends (relative
gasoline)
to
MTBE-gasoline blends
(relative to gasoline)
Methanol-gasoline blends
(relative to gasoline)
1. Exhaust emissions: about 40% paraffins; 24% aromatics;
26% olefins; 11% acetylene.
2. Main components of exhaust emissions: methane, ethane,
acetylene, ethylene, propylene, C. olefins, toluene,
benzene, n-butane, rrpentane, isopentane, xylene;
aldehydes and some organic acids, ketones, phenols.
3. Evaporative emissions: 70% of carburetor emissions are
light paraffins and olefins; 90% of fuel-tank emissions
are light paraffins and olefins.
1. Exhaust emissions: about 62% paraffins, 17% aromatics,
18% olefins, 3% acetylene. Methane is about 10 to 30%
of exhaust emissions.
2. Catalysts preferentially oxidize unsaturated HC.
3. Lower reactivity per gram HC emissions than from un-
controlled gasoline-fueled.
4. Lower net HC emissions than from uncontrolled gasoline-
fueled.
1. Presence of TEL increased HC emissions.
2. Absence of TEL (or TML) necessitates higher aromaticity
of gasoline to achieve higher octane ratings.
exhaust THC emissions.
evaporative THC emissions.
aggregate (exhaust plus evaporative) THC
in exhaust increased from noncatalyst cars;
from cars with oxidation catalysts; increased
with TWC catalysts.
in ethylene fron noncatalyst cars; no increase
with oxidation catalysts.
in acetic acid with increasing alcohol content.
1. Decreased exhaust THC emissions.
2. Increased evaporative THC emissions, but less than with
alcohol-gasoline blends.
3. Aldehydes increased from cars with TWC catalysts;
decreased from cars with oxidation catalysts.
1. Emission rate of exhaust THC not appreciably changed;
composition changed.
2. Higher methanol in exhaust.
3. Aldehydes in exhaust increased from noncatalyst cars;
no significant change in aldehydes with TWC catalysts;
higher aldehydes with oxidation catalysts.
1.
2.
3.
4.
5.
6.
Decreased
Increased
Increased
emissions
Aldehydes
unchanged
from cars
Increase
from cars
Increases
019WPS/B
3-56
6/28/84
-------
TABLE 3-12. SUMMARY OF EMISSION CHARACTERISTICS FOR AUTO FUELED BY GASOLINE,
DIESEL, AND ALCOHOL-GASOLINE OR ETHER-GASOLINE BLENDS (continued)
Auto, control device
or fuel
Emission characteristics
100% Methanol
(relative to gasoline)
Diesels
(relative to gasoline)
4. Decreases in exhaust NH-, HCN, photochemical reactivity
with increases in % alcohol.
5. Increases in exhaust formic acid with increases in
% alcohol.
1. Higher cold-start HC emissions.
2. Significantly lower hot-start HC emissions.
3. Exhaust emissions: mainly methane, ethane, ethylene.
4. Significantly higher methanol and aldehyde emission
(aldehydes reduced by increasing compression ratio or
adding water to methanol).
1. Almost exclusively exhaust emissions.
2. Emissions: light, cracked HC, mainly methane, ethylene,
acetylene, propylene; also aldehydes (C-.-Cg), including
acrolein), and acetone.
3. Lower reactivity per gram HC emissions.
4. Lower net HC emissions.
5. Higher carbonyl emissions (aldehydes, ketones).
Note: HCN = hydrogen cyanide
NH3 = ammonia
TEL = tetraethyl lead
THC = total hydrocarbon
TML = tetramethyl lead
TWC = three-way catalyst
MTBE = methyl tertiary butyl ether
Source: Tilton and Bruce (1981).
019WPS/B
3-57
6/28/84
-------
TABLE 3-13. NATIONAL ESTIMATES OF EMISSIONS OF NITROGEN OXIDES, 1982
Source category
Transportation
Highway vehicles
Aircraft
Railroads
Vessels
Other off-highway vehicles
Transportation: total
Stationary source fuel combustion
Electric utilities
Industrial
Commerci al - i nsti tuti onal
Residential
Fuel combustion: total
Industrial processes
Nitric acid
Petroleum refining
Other
Industrial: total
Solid waste disposal: total
Miscellaneous: total
Total
Weight emitted,
tg/yr
(7.8)
(0.1)
(0.7)
(0.2)
(0.9)
9.7
(6.2)
(2.7)
(0.3)
(0.4)
9.6
(0.1)
(0.2)
(0.3)
0.6
0.1
0.2
20.2
Percent
48.0
47.5
3.0
0.5
1.0
100.0
Source: U.S. Environmental Protection Agency (1983).
019WPS/B
3-58
6/28/84
-------
TABLE 3-14. NO/NO RATIOS IN EMISSIONS FROM VARIOUS TYPES OF SOURCES
Source type
Uncontrolled tail -gas from nitric
acid plants
Petroleum refinery heaters —
using natural gas
Linear ceramic tunnel kiln
Rotary cement kilns
Steel soaking pit-natural gas
Wood/bark boiler
Black liquor recovery boiler
Carbon monoxide boiler
Large 2-cycle internal combustion
engine—natural gas
Combined cycle gas turbine
Gas turbine electrical generatoi —
#2 fuel oil
Industrial boilers
(variety of fuels)
NO/NO
~ 0.50
0.93-1.00
0.90-1.00
0.94-1.00
0.97-0.99
0.84-0.97
0.91-1.00
0.98-1.00
0.80-1.00
0.83-0.99
0.55-1.00
(no load)-
(full load)
0.90-1.00
Reference
Gerstle and
Peterson, 1966
Hunter et al . ,
1979
Wasser, 1976
Cato et al . ,
1976
Diesel-powered passenger cai—Nissan
0.77-0.91
(idle)-(SOmph)
Braddock and
Bradow, 1976
Diesel-powered passenger car--
Peugeot 204d
Diesel-powered passenger car—
various Mercedes
0.46-0.99
(idle)-(SOmph)
0.88-1.00
Springer and
Stahman, 1977a
Diesel-powered truck and bus--
various engines
0.73-0.98
Springer and
Stahman, 1977b
Mobile vehicles internal gasoline
combustion engine
0.99-1.00
Milliard and
Wheeler, 1979
Aircraft turbines (JT3D, TF30)
0.13-0.28 (idle )
0.73-0.92
(takeoff and cruise)
Souza and
Daley, 1978
Earlier studies (Lozano et al., 1968; Chase and Hurn, 1970) did not report
such high idle concentrations of N02.
Source: U.S. Environmental Protection Agency (1981).
019WPS/B
3-59
6/28/84
-------
units of up to 80 percent, but no more than 50 percent in coal-fired units.
In subsequent tests with coal-fired units, reductions of up to 62 percent were
achieved (Crawford et al., 1974, 1978; cited in Hall and Bowen, 1982). Since
coal has a higher nitrogen content than oil or gas, this may be near the limit
of reductions achievable through combustion modification.
Baseline NO emissions from the varied population of industrial boiler
/\
designs were found to be most dependent on fuel (Cato et al. , 1974, 1976;
cited in Hall and Bowen, 1982). Coal-fired units had the highest average NO
/v
emissions (475 ppm); oil the next highest (120 to 293 ppm, depending on fuel
grade); and natural gas the lowest (139 ppm). Average reductions of 12 to
22 percent, with a maximum reduction of 75 percent, have been demonstrated
using one or more combustion modification techniques. Results were quite
variable, depending on boiler design.
Similarly, industrial process heaters are of diverse design and function,
produce quite varied baseline NO emissions, and respond to adjustments in
combustion conditions with mixed results (Hunter et al., 1978, 1979; cited in
Hall and Bowen, 1982). Most units fired with natural gas, refinery gas, No. 6
oil, or wood produced NO concentrations in the range of 76 to 320 ppm.
Cement kilns heated with natural gas produced some high levels, ranging from
90 to 2250 ppm NO . Emission reductions among the industrial process heaters
/\
studied ranged from nil to 69 percent.
As might be expected, modifying combustion conditions in stationary
combustion units to reduce NO emissions sometimes caused concomitant increases
/\
in CO and particulate emissions. Overall efficiency was sometimes reduced,
sometimes increased.
Emissions of NO from mobile sources, gasoline- and diesel-fueled vehicles
^\
are affected by a number of variables such as speed, load, and air:fuel ratio
(APR), as reported recently by Billiard and Wheeler (1979). They concluded
that gasoline engines show a maximum emission of N02 at "lean" AFRs of about
17:1; whereas diesel engines show a maximum emission at very lean AFRs of
about 70:1 and at low speeds, under which conditions as much as 30 percent of
the NO emissions may be N0?. An active platinum oxidation catalyst increases
/> «•
the N02 in both engine types at a catalyst temperature of about 470°C, but CO
levels above 1000 ppm in gasoline engine exhaust can negate the conversion to
N02.
019WPS/B 3-60 6/28/84
-------
Gibbs et al. (1983) reported on NO emissions from a group of 19 in-use
/\
diesel automobiles representing the 1977 to 1979 model years. Vehicle mileage,
both initial and accumulated, varied considerably, but over the 28-month test
period the average accumulation was 35,000 miles per vehicle. Emissions of
NO at the conclusion of the period (phase 3), using the Federal Test Proce-
/\
dure, ranged from 0.84 g/mi to 3.15 g/mi. The general trend with accumulating
mileage ranged from no change to a 20 percent decrease in NO emissions.
/\
Smith and Black (1980) reported on emissions from four gasoline-powered
passenger cars equipped with three-way catalyst (TWC) control systems. These
catalysts oxidize hydrocarbons and carbon monoxide to carbon dioxide and water
as conventional oxidation catalysts do; and at the same time, reduce NO to
A
nitrogen (N«). Values for NO emissions ranged from 0.41 to 0.89 g/km (0.66
to 1.43 g/mi) for the 1978 Federal Test Procedure; from 0.50 to 1.14 g/km
(0.80 to 1.84 g/mi) for the Congested Freeway Driving Schedule; from 0.49 to
1.05 g/km (0.79 to 1.69 g/mi) for the Highway Fuel Economy Driving Schedule
(HFET); and from 0.41 to 0.93 g/km (0.66 to 1.50 g/mi) for the New York City
Driving Schedule (NYCC). All vehicles met the standard for which they were
designed.
Two other recent papers by Dietzmann et al. (1980, 1981) deal with NO
/\
emissions from heavy-duty gasoline and diesel trucks. Using several test
fuels, the authors found NO emissions from diesel engines ranging from 10.86
J\
g/km (17.47 g/mi) to as much as 26.35 g/km (42.40 g/mi) while the engines were
operating on the 1983 transient cycle chassis test. In another test with
engines using leaded gasoline, NO emissions ranged from 7.64 to 9.68 g/km
J\
(12.29 to 15.58 g/mi). Detailed NO emission rates from four heavy-duty diesel
engines for a number of fuels showed no obvious trends regarding the effect of
different fuels on NO emissions. Instead, differences ranging from 12 to 27
/\
g/km (19.3 to 43.4 g/mi) were seen among types of engines (Dietzman et al. ,
1980, 1981).
Factors influencing seasonal variations in NO emissions from mobile
x\
sources include temperature (about a 35 percent decrease in emissions per
vehicle mile with an ambient temperature increase from 20 to 90°F) (Ashby et
al., 1974), and number of vehicle miles traveled (about 18 percent higher in
summer than in winter, nationwide) (Federal Highway Administration, 1978).
There are also differences between vehicle miles traveled in urban versus
rural areas and among states in different regions of the country (Federal
019WPS/B 3-61 6/28/84
-------
Highway Administration, 1978). Seasonal variations in NO emissions from
stationary sources are also expected since fossil-fueled power plants produce
an estimated 15 percent more NO in the summer than in the spring (U.S. Depart-
J\
ment of Energy, 1978). Greater degrees of variation and different seasonal
patterns have been reported for different regions of the country (California
Board of Sanitation, 1966). Diurnal emission variations, notably those asso-
ciated with motor vehicle traffic, are also important because of their poten-
tial impact upon ambient air quality.
3.4.2 Natural Sources and Emissions
3.4.2.1 Natural Sources and Emissions of Volatile Organic Compounds. This
section presents information on hydrocarbon emissions from biogenic sources.
Sampling of natural sources for determination of hydrocarbon emission rates
necessitates the use of techniques far different from those used to sample
emissions from manmade sources. Whereas the Agency has formulated guidelines
for preparing inventories of manmade emissions (section 3.4.1), comparable
guidelines for preparing inventories of natural emissions do not exist.
Knowledge of natural emissions of hydrocarbons and other VOC is still suffi-
ciently formative to require additional research. Consequently, this section,
unlike the previous section, includes a discussion of techniques used by
researchers to inventory natural (biogenic) hydrocarbon emissions. It also
includes estimates of emission rates for individual species and for all species
in certain areas as well as across the United States.
During the 1970's, studies conducted in several different laboratories
established that monoterpenes, which were known or expected to be present in
ambient forest atmospheres, were quite reactive both in photochemical processes
and in ozonolysis reactions. These laboratory studies demonstrated that
biogenic hydrocarbons will produce ozone when they are irradiated in the
presence of oxides of nitrogen. It has also been established, however, that
natural hydrocarbons are oxidized by ozone under normal atmospheric conditions.
There have been recent as well as earlier reports that ascribe both an ozone-
producing and an ozone-scavenging role to biogenic hydrocarbons (Arnts and Gay,
1979; Roberts et al., 1983). Although the actual atmospheric fate of natural
hydrocarbons is currently not well understood (i.e., oxidation to gas phase
and aerosol products), a good deal of progress has been made toward character-
izing the identities and magnitude of biogenic hydrocarbon emissions. The
019WPS/B 3-62 6/28/84
-------
purposes of this section of the document are to (1) provide a brief summary of
the types of hydrocarbons emitted by natural processes; (2) describe the
experimental methods utilized to measure emission rates; (3) discuss procedures
employed for establishing natural hydrocarbon inventories; and (4) furnish a
tabulation of reported biogenic hydrocarbon emission inventories. For detailed
information, the reader is referred to several excellent review articles
(Altshuller, 1983; Bufalini and Arnts, 1981; Dimitriades, 1981).
Geogenic hydrocarbon emissions—those emanating from natural gas seeps,
volcanoes, and geothermal venting—will not be considered for two reasons.
First, the majority of hydrocarbons emitted by geogenic sources are low-
molecular-weight alkanes that react very slowly in photochemical systems in
the atmosphere. Second, hydrocarbon emission rates from geogenic sources are
poorly understood, such that reliable data for these sources are sparse.
Consequently, the ensuing discussion will be concerned only with biogenic
emissions from vegetation.
3.4.2.1.1 Biogenic VOC emissions. As part of the photosynthesis process, the
leaves of plants produce sugars that are converted to starches for storage, to
cellulose for structural growth, and to a variety of secondary compounds that
participate in the normal metabolism of the plant. The production of isopre-
noid compounds is a normal metabolic process in all green plants. To date,
isoprene and the monoterpenes (section 3.2) are the only biogenic hydrocarbons
identified as emissions from vegetation. These compounds are of interest to
atmospheric scientists largely because they are volatile enough to be released
under normal environmental conditions and because they have been shown to be
potential ozone precursors.
Measurements by Sanadze and Dolidze (1962) and Rasmussen (1964) suggested
that isoprene was emitted by plants. Subsequent work by Rasmussen (1970) and
Evans et al. (1982) associated isoprene emissions with a variety of plant
species. Monoterpenes are emitted by coniferous trees as well as by some
deciduous types of vegetation. The commonly identified monoterpenes are
a-pinene, (3-pinene, camphene, A3-carene, limonene, myrcene, and p-phellandrene.
In addition, the oxygenated monoterpenes, 1,8-cineole and camphor, have been
detected in some plant emission samples. As a general rule, coniferous trees
emit primarily monoterpenes, and deciduous vegetation emits isoprene.
3.4.2.1.2 Biogenic emission rates. Biogenic emission rates have been deter-
mined almost exclusively by enclosure techniques. This procedure involves
019WPS/B 3-63 6/28/84
-------
enclosing the entire plant or a portion of it, such as a branch of a tree, in
a bag or chamber constructed of light, transparent material. If the chamber
is operated in a static mode, a background sample is collected immediately
after enclosure. Emissions are then allowed to accumulate for a measured
period of time, after which an emission sample is withdrawn from the chamber.
The branch or plant is removed from the chamber and the leafy portion that was
enclosed is dried and weighed. The emission rate is calculated from the
difference in concentration between the background and emission samples divided
by the time and the weight of biomass. Emission rates are generally expressed
in units of micrograms per gram dry biomass per hour [jjg (g. , . ) hr"1].
Enclosure chambers can also be operated in a dynamic mode, in which case the
emission rate is obtained by multiplying the concentration of biogenic hydro-
carbons eluted from the chamber by the air flow rate and dividing this product
by the weight of dried biomass.
Tables 3-15 and 3-16 provide a summary of isoprene and monoterpene emis-
sion rates measured by the enclosure method. Because biogenic emission rates
are temperature-dependent, the corresponding temperatures are also listed in
the tables. A comparison of the two tables indicates that during the daytime
isoprene emission rates are generally higher than those reported for monoter-
penes. Isoprene emission rates range from 3 to 233 ug g hr at 30°C,
whereas monoterpene rates vary from less than 1 to 15 |jg g hr at 30°C. No
emission rate information is included in these tables for agricultural crops.
A few emission measurements have been made for corn, tobacco, forage crops,
and pasture; however, the data base is too small to permit development of
reliable emission rate estimates. The biogenic emission rates recorded for
these crops were in the range of 0.5 to 2 pg g hr (Lamb et al. , 1983).
Besides temperature, biogenic emission rates are affected by other environ-
mental factors. Rasmussen (1972) reported that emission rates varied with
species, plant maturity, resin gland integrity, and leaf temperature. Dement
et al. (1975) found that the emission rate of monoterpenes from Salvia mellifera
(California Black Sage) is dependent on the vapor pressures of the terpenes,
the humidity, and the amount of oil present on the surface of the leaf. These
investigators also reported that the emission rate is not directly dependent
on the photosynthetic activity or on the stomatal opening of the plant. This
suggested that the release mechanism was physical and that the terpenes were
volatilized from the surface of the leaf rather than from the inside.
019WPS/B 3-64 6/28/84
-------
TABLE 3-15. ISOPRENE EMISSION RATES
Species
Robinia pseudoacacia
Platanus occidental is
Platanus racemosa
Salix nigra
Sal ix babylonica
Sal ix carol iniana
Washingtonia filifera
Phoenix dactyl ifera
Populus tremuloides
Populus deltoides
Quercus boreal is
Quercus dumosa
Quercus laurifolia
Quercus nigra
Quercus laevis
Quercus virginiana
Quercus incana
Quercus myrti folia
Quercus phellos
Quercus agri folia
Liquidambar styraciflua
Liquidambar styraciflua
Picea engelmannii
Picea sitchensis
Eucalytus global us
Common Emission rate,
name ug g hr
Locust
Sycamore
Sycamore
Wi 1 1 ow
Willow
Willow
Palm
Palm
Aspen
Cottonwood
Oak
Oak
Oak
Oak
Oak
Oak
Oak
Oak
Oak
Oak
Sweetgum
Sweetgum
Spruce
Spruce
Eucalyptus
11
21
11
19
233
12
11
15
38
28
15
35
10
24
23
9
44
15
31
49
14
8
12
3
44
Temp. ,
°C
30
28
30
28
30
30
30
30
28
28
28
30
30
30
30
30
30
30
30
30
28
30
28
28
28
Reference
Winer et al. (1983)
Evans et al. (1982)
Winer et al. (1983)
Evans et al. (1982)
Evans et al. (1982)
Zimmerman (1979)
Evans et al . (1982)
Evans et al . (1982)
Evans et al. (1982)
Evans et al. (1982)
Evans et al. (1982)
Winer et al. (1983)
Zimmerman (1979)
Zimmerman (1979)
Zimmerman (1979)
Zimmerman (1979)
Zimmerman (1979)
Zimmerman (1979)
Zimmerman (1979)
Winer et al. (1983)
Evans et al. (1982)
Zimmerman (1979)
Evans et al. (1982)
Evans et al . (1982)
Evans et al. (1982)
019WPS/B
3-65
6/28/84
-------
TABLE 3-16. MONOTERPENE EMISSION RATES
Species
Common
name
Emission rate,
M9 9 nr
Temp. ,
°C
Reference
Pseudotsuga taxi folia
Pinus ponderosa
Pinus ponderosa
Pinus palustris
Pinus clausa
Pinus elliotti
Pinus elliotti
Pinus sylvestris
Pinus taeda
Pinus taeda
Pinus halepensis
Pinus canariensus
Pinus radiata
Picea abies
Picea engelmannii
Picea sitchensis
Eucalyptus global us
Acer saccharum
Douglas fir
Pine
Pine
Pine
Pine
Pine
Pine
Pine
Pine
Pine
Pine
Pine
Pine
Spruce
Spruce
Spruce
Eucalyptus
Maple
15
5
4
6
11
6
3
6
1
4
0.6
2
0.6
5
3
1
8
2
30 Knoppel et al. (1982)
30 Knoppel et al. (1982)
30 Rasmussen (1972)
30 Zimmerman (1979)
30 Zimmerman (1979)
28 Evans et al. (1982)
30 Zimmerman (1979)
30 Knoppel et al. (1982)
30 Knoppel et al. (1982)
30 Arnts et al. (1978)
30 Winer et al. (1983)
30 Winer et al. (1983)
30 Winer et al. (1983)
28 Evans et al. (1982)
28 Evans et al. (1982)
28 Evans et al. (1982)
28 Evans et al. (1982)
28 Evans et al. (1982)
019WPS/B
3-66
6/28/84
-------
Tingey and his coworkers have conducted extensive studies into the effects
of environmental conditions on emission rates. Live-oak seedlings were used
to study the influence of temperature changes, water stress, light intensity,
and COp levels on isoprene emission rates (Tingey et a!., 1981). Isoprene
emissions decreased to near zero levels in the dark. Water stress did not
alter the emission rate significantly and exposure to subambient CO^ levels
caused a small increase in isoprene emission rate. During daylight, tempera-
ture appears to be the main factor controlling isoprene emissions. Figure 3-10
shows a typical example of the relationship between ambient temperature and
the emission rate of isoprene.
In contrast to isoprene, monoterpene emission rates do not appear to be
influenced by light intensity. The emission rate of a-pinene as well as rates
for four other monoterpenes emitted by slash pine were similar under various
levels of light and darkness (Tingey et al., 1980). A log-linear increase in
emission rates of monoterpenes with temperature was observed, however, in the
slash pine studies.
Peterson and Tingey (1980) have used published data to model isoprene and
monoterpene emission profiles through the diurnal cycle. The predicted emission
rates, which are averages for several species, are shown in Figure 3-11. This
figure illustrates the light-dependence of isoprene emissions and the minimal
fluctuation expected in monoterpene rates.
There has been considerable discussion concerning the validity of emission
rate data obtained by the bag enclosure technique. Confidence in this method
has been limited primarily by uncertainties associated with isolating the
vegetation in an artificial environment. Temperature, humidity, and possibly
C0? concentration are likely to be different inside the enclosure. Also,
emission rates will certainly increase if the vegetation inside the enclosure
has suffered physical damage. Other questions, such as how representative
emission rates are when measured from just one branch and whether emission
rates are commensurate with ambient biogenic hydrocarbon concentrations, are
difficult to answer.
Attempts to validate the bag enclosure method have focused on comparing
enclosure emission estimates with those obtained by alternate procedures.
Biogenic emission fluxes have been determined in hardwood and coniferous
forests using micrometeorological gradient procedures and the enclosure method
(Lamb et al., 1983). In the gradient studies, the flux of a particular biogenic
019WPS/B 3-67 6/28/84
-------
100.0
60.0
O>
O>
w
z
O
55
CO
01
O
I
20.0
10.0
6.0
2.0
1.0
0.6
0.4
10
I I
— PULLMAN, WA
LANCANSTER, PA
15 20 25
TEMPERATURE, °C
30
35
Figure 3-10. Total nonmethane hydrocarbon
emission rate as a function of temperature for
isoprene-emitting hardwoods at two sites.
Source: Flyckt et al. (1980)
3-68
-------
z
o
at o
O c>
UU 3.
Crt <
cooc
ISOPRENE
10,000
1,000
100
-a.m.-
•*- NOON
-p.m.
Figure 3-11. Estimated diurnal cycle of isoprene and
monoterpene emission rates.
Source: Peterson and Tingey (1980).
3-69
-------
hydrocarbon was determined from measurements of vertical eddy diffusivities
and vertical hydrocarbon concentration profiles collected at respective levels
along a tower extending above the forest canopy. Hydrocarbon flux was calcu-
lated on the basis of surface layer theory, where the flux is given as
F = KZ dc/dz (3-2)
and the vertical diffusivity, K , was obtained from vertical wind speed and
temperature profiles. In order to compare the results from the two methods,
the emission rates determined by the enclosure method must be converted to an
emission flux by multiplying by a biomass factor:
Er (ug g"1 hr"1) x Bf (g m~2) = F (ug m"2 hr"1) (3-3)
Biomass factors can be obtained from forest inventory information and from
allometric relationships, which relate vegetation size and density to dry
weight biomass per square meter of surface area (Sollins et al., 1973). At
30°C, isoprene fluxes obtained using the enclosure and gradient profile methods
in a Pennsylvania hardwood forest agreed very closely. The gradient profile
_2 -i
procedure gave a flux of 8,000 ug m hr , while the enclosure technique
yielded 7,300 ug m"2 hr"1 (Lamb et al. , 1983). Good agreement has been
reported, also, for crpinene emission fluxes measured by a micrometeorological
procedure and the enclosure method (Knoerr and Mowry, 1981).
Although the micrometeorological approach yields mass fluxes similar to
the enclosure method, it, too, has certain limitations. The measurement of
small vertical gradients above a forest canopy and the application of surface
layer theory to non-ideal sites can lead to erroneous results. In many res-
pects, the difficulties in measuring mass fluxes from a forest can be avoided
by simulating the forest emissions with an inert tracer release and measuring
ambient concentrations of the tracer and biogenic gases along downwind sample
lines. The tracer approach does not require perturbation of the vegetation,
nor does it rely upon precise gradient measurements. The only requirement is
for an isolated source impacted minimally by upwind biogenic hydrocarbon
contributions. Isoprene fluxes obtained using the tracer procedure in a
central Washington oak grove compared well with flux estimates determined
simultaneously with the enclosure technique (Allwine et al., 1983). Thus, it
019WPS/B 3-70 6/28/84
-------
appears that if the bag enclosure procedure is applied properly representative
biogenic emission rates will be acquired.
3.4.2.1.3 Biogenic emission inventories. The basic information required for
development of a biogenic emission inventory is a knowledge of vegetation
coverage, emission rates, and biomass factors. How these components are
derived and utilized in the inventory process is illustrated through the
following example.
A forested plot consisting of isoprene-emitting hardwoods is chosen for
study. The emission rate of biogenic hydrocarbons is measured for representa-
tive types of vegetation within the plot. This is most easily accomplished
using the enclosure method described previously. Sufficient data must be
acquired so that the relationship between emission rate and temperature can be
clearly defined. Once this has been accomplished, the size distribution and
number of trees in the plot must be determined. Standard procedures that
utilize random sampling techniques are employed for this type of forest survey.
The information contained in the first four columns in Table 3-17 is derived
directly from the forest survey. The next step in the inventory process is to
determine the total leaf biomass for the study area, which is done by means of
regression equations that relate biomass to a more easily measured tree para-
meter. Allometric equations (y = ax ) that relate tree growth (biomass) to
the proportions of trees have been developed by cutting down sample trees and
subjecting them to intensive measurements. Figure 3-12 shows four allometric
relationships that can be used to correlate biomass with tree diameter at
breast height (DBH). Relationship number 4 was used together with the data in
column number 4 of Table 3-17 to obtain biomass values for each tree species
in this sample inventory (Table 3-17, column number 5). Multiplying the tree
frequency by leaf biomass per tree gives a species total. Then, by summing
the individual species totals, a biomass factor for the forest plot can be
established. Finally, the biogenic hydrocarbon flux from the forest plot is
obtained by multiplying this biomass factor by the appropriate emission rate
value determined by the enclosure technique. An area-wide emission estimate
can be obtained by multiplying the flux (ug m~2 hr'1) by the area (m2) of the
forest plot.
Table 3-18 contains a listing of area-wide biogenic emission fluxes that
have been reported for the United States and portions thereof. All of the
emissions data contained in this table have been derived in a manner similar
019WPS/B 3-71 6/28/84
-------
TABLE 3-17. FOREST SURVEY DATA FOR ISOPRENE-EMITTING HARDWOODS
Species
(Col. 1)
Black oak
Chestnut oak
Black locust
Scarlet oak
White oak
Canopy
position
(Col. 2)
Overstory
Understory
Overstory
Understory
Overstory
Understory
Overstory
Understory
Overstory
Understory
Frequency,
trees/ha
(Col. 3)
99.66
14.40
41.53
72.02
16.62
14.40
24.92
0.0
26.98
33.59
Avg DBH
cm
(Col. 4)
33.49
23.91
31.08
13.14
31.61
13.32
36.65
0.0
31.89
19.11
Leaf
, biomass,
kg/tree
(Col. 5)
15.58
8.80
13.73
3.19
14.89
3.62
17.32
0.0
14.34
6.02
Forest total :
Total
biomass,
kg/ha
(Col. 6)
1552.70
126.72
570.21
229.74
247.47
46.94
431.61
0.0
386.89
202.20
= 3794 kg/ha
= 379 g/m2
Source: Lamb et al. (1983)
019WPS/B
3-72
6/28/84
-------
100
S
(9
ui
K
CO
.
i
10
0.1
RELATIONSHIP #1
(Whittaker)
RELATIONSHIP ti
(Monk)
RELATIONSHIP *3
(Whittaker)
RELATIONSHIP #4
(Sollins)
I
I
1 10
LEAF BIOMASS, kilograms
100
Figure 3-12. Comparison of four allometric relationships for determination of
leaf biomass.
Source: Lamb et al. (1983).
3-73
-------
CO
00
TABLE 3-18. AREA-WIDE BIOGENIC EMISSION FLUXES
co
Location
Emission flux,
-2 -1
ug m hr
Comment
Reference
South Coast Air Basin,
California
Lake Tahoe, California
Lake Tahoe, California
San Francisco Bay Area,
California
San Francisco Bay Area,
California
San Francisco Bay Area,
California
Tampa/St. Petersburg,
Florida
Southeastern Virginia
Pennsylvania
Houston, Texas
United States
United States
United States
<780
1950
2438
1388
2265
777
2540
8890
1660
1170
1712
1099
884
Entire basin
Forested area of basin
Daytime
Nighttime
Forested area only
Winer et al. (1983)
JSA, Inc. (1978)
JSA, Inc. (1978)
Sandberg (1978)
Hunsaker (1981)
Hunsaker (1981)
Zimmerman (1979)
Salop et al. (1983)
Flyckt (1980)
Zimmerman (1979)
Marchesani (1970)
Zimmerman (1977)
Zimmerman (1978)
CO
2
-------
to that just described. Needless to say, accuracy of the estimate depends
upon the size of the area for which the inventory has been prepared. Good
biogenic emission estimates can be obtained for small forest plots that are
well characterized in terms of emission rates, tree species, and biomass.
When attempts are made, however, to inventory large areas such as the entire
United States, much more uncertainty is .introduced. As the size of the area
increases, it becomes more difficult to select representative vegetation
classes and to assign proper emission rates to them. Uncertainty in the
assignment of biomass factors also increases as the size of the inventory base
is expanded. In many cases, it is desirable to express emission estimates in
terms of weight of biogenic emissions per day or per year. In these cases, it
must be recognized that isoprene emissions are nearly zero at night and undoub-
tedly quite low in the winter when deciduous trees are without leaves. Also,
since emission rates are temperature-dependent, variations in diurnal and
seasonal temperatures must be considered. Many of these problems in preparing
inventories have been discussed in detail (Altshuller, 1983; Zimmerman, 1981;
Wells, 1981; Box, 1981; Dimitriades, 1981).
Considering all the variables, it is somewhat surprising that the area-wide
emission fluxes listed in Table 3-18 show no more variation than they do.
With the exception of the Southeastern Virginia area, which is a forested
region with high biomass coverage, most values differ by less than a factor of
three.
3.4.2.2 Natural Sources and Emissions of Nitrogen Oxides.
Natural emissions of nitrogen oxides (NO ) originate from the oxidation
/\
of nitrogen gas by electrical discharge in the atmosphere, from the ammonifica-
tion of organic nitrogen during biological decomposition, and from the oxida-
tion of organic nitrogen during forest fires. Nitrogen fixation and electrical
discharge are normal processes of the nitrogen cycle that convert inert nitrogen
gas to biologically useful nitrate or ammonia.
The atmosphere, composed of 79 percent molecular nitrogen (NO, is an
important reservoir for nitrogenous compounds and provides a major link between
terrestrial and aquatic ecosystems for the transport and transformation of
gaseous and particulate forms of nitrogen oxides (NO ). Figure 3-13 depicts
y\
the nitrogen cycle.
Nitrogen fixation is the conversion of inert nitrogen gas to biologically
more usable forms, either by reduction to NH. or oxidation to NO- . It has
019WPS/B 3-75 6/28/84
-------
N2
MN°TROGENR I - \ BIOLOGICAL FIXATION
NITROGEN | - op MOLECULAR MTRQGEN
ATMOSPHERE
ELECTRICAL
AND
PHOTOCHEMICAL
FIXATION
en
NITROGEN
OXIDES:
NO, NO2
VOLCANIC
ERUPTION
WEATHERING
OF ROCKS
FOREST & GRASSLAND FIRES
STORAGE OF
NITROGENOUS
COMPOUNDS IN
SEDIMENTS, SOILS,
AND SEDIMENTARY
ROCKS
ANIMALS IN GRAZING
FOOD CHAIN
J I,
DEATH & WASTES
DETRITUS FOOD CHAIN
AMMONIA/^ NITROGEN
NH3 \1 R =
- -b^
Figure 3-13. The nitrogen cycle (organic phase shaded).
Source: U.S. Environmental Protection Agency (1982)
-------
been estimated that, on a global basis, nitrogen fixation in terrestrial
ecosystems accounts for the production of 139 Tg of fixed nitrogen each year;
leguminous plants account for 35 Tg of this total with the remainder being
produced by free-living nitrogen-fixing microorganisms in forest and grasslands
(Burns and Hardy, 1975). Estimates by other investigators differ considerably
and are shown in Table 3-19. The amount of biologically fixed nitrogen actual-
ly released into the ambient air is unknown.
Lightning flashes in the troposphere can convert N~ to NO via reaction
with monatomic oxygen. Crutzen and Ehhalt (1977) estimated that from 8 to 40
Tg N are fixed by lightning each year. The work of Chameides et al. (1977)
suggests that lightning is a significant source of NO , producing about 30 to
40 Tg NO -N per year. If this estimate is correct, lightning could account
/\
for as much as 50 percent of the total atmospheric NO on a global basis.
/\
Direct observations by Noxon (1976) have indicated that during a lightning
storm ambient concentrations of NO- could possibly be enhanced by a factor of
500 over normal. The enhanced levels declined rapidly after passage of the
storm. Liu et al. (1977) estimated NO production by lightning at 9 Tg per
year.
Biological ammonification is an important process in the renewal of the
limited supply of inorganic nitrogen. During the decomposition process,
organic compounds, such as ami no acids, are converted into NHL and ammonium
O
ions. Volatilization of ammonia from soils may increase the atmospheric
concentration of NO as NHL undergoes atmospheric transformations (National
X O
Academy of Sciences, 1978; Hill, 1971).
Total NO emissions to the atmosphere from terrestrial sources were
s\
reported by Soderlund and Svensson (1976) to be in the range of 8 to 25 Tg N
per year. Soderlund and Svensson (1976) have hypothesized that a net flow of
NO prevails from terrestrial to aquatic systems; losses of NO from aquatic
X X
systems to the atmosphere were considered insignificant.
Although 40 to 108 Tg NO -N per year has been estimated to be released
s\
from terrestrial sources to the atmosphere, the bulk is reabsorbed and only 8
to 25 Tg NO -N escapes to the troposphere (Robinson and Robbins, 1975). Hill
(1971) reported that NO and NO- are absorbed, to some extent, from the atmos-
phere by plants. Using data obtained from the experiments of Marakov (1969),
and those of Kim (1973), Soderlund and Svensson (1976) estimated that soil
contributes to the atmosphere between 1 and 14 Tg N in the form of NO and N0_;
019WPS/B 3-77 6/28/84
-------
TABLE 3-19. GLOBAL ESTIMATES OF NITROGEN TRANSFORMATION
(Tg N/yr)
Range of estimates
Reference1
Biological fixation
(N2 » NH4 )
54 to 270
1. Delwiche (1970).
2. Burns and Hardy (1975).
3. Soderlund and Svensson (1976).
4. Robinson and Robbins (1975).
5. Liu et al. (1977).
6. Sze and Rice (1976).
7. Council for Agricultural Science and Technology (1976).
8. Chameides et al. (1977).
1, 2, 3, 4, 5
Electrochemical fixation
lightning (N2 » NO )
atmospheric (N2 *• fio2)
Biological denitrifi cation
(N03~ > N2)
(N03 > N20)
combined
Industrial denitrification
(Organic -N » NO )
(Other > NO )
/\
Atmospheric denitrification
(NH3 > N0x)
Natural NO emissions
(land anci sea)
NH3 emissions to atmosphere
from land and sea
10
14
96
20
83
14
30
3
40
110
to
to
to
to
to
to
to
to
to
to
40
20
190
340
270
19
36
30
210
850
2,
2,
2,
2,
5,
2,
2,
2,
3,
2,
8
4
3
3,
6,
3,
3,
3
4
3,
4
7
4
5
4
019WPS/B
3-78
6/28/84
-------
losses from aquatic ecosystems to the atmosphere were considered by these
authors to be minor. The principal source of gaseous NO in terrestrial
systems is believed to be the decomposition of nitrates (Soderland and Svensson,
1976).
3.4.2.3 Local Natural Sources of NO . Since the ozone/oxidant-forming poten-
tial of natural sources of precursors depends to some degree upon their local
distributions, it is worthwhile to examine local sources of NO . In this
f\
section, available estimates of emission rates of NO from fresh water, swamps,
J\
soil, and vegetation are presented.
The fate of nitrogen in fresh water has been reviewed by Keeney (1973)
and by Brezonik (1973). Molecular nitrogen is the main product of denitrifica-
tion, a biochemical process; NO (or N-0) is seldom detected (Black and High,
1979). Brezonik (1973) concluded that denitrification did not appear to be a
significant process in Florida lakes. In Smith Lake, Alaska, Goering and
Dugdale (1966) failed to detect NO or N_0, although N2 as a product of denitri-
fi cation was present. Under acid conditions resulting from high concentrations
of polyphenolic substances such as tannins, lignins, and humic acid, the
purely chemical reaction of nitrous acid with organic substances could result
in a significant source of NO (Brezonik, 1973). Direct measurement of this
)\
as a possible source of emissions, however, appears to be lacking.
There is no direct evidence that plants emit any NO into the atmosphere,
although they contain, and exchange with their local environment, significant
quantities of nitrogen in various oxidation states. In addition, data are
lacking on possible NO exchange from forest litter (Ratsch and Tingey, 1978).
y\
Wijler and Delwiche (1954) identified NO as a product of denitrification
processes in soil, although N? was the major initial product. At least two
studies (Renner and Becker, 1970; Payne et al., 1971) have shown that NO is a
specific product of the bacterial reduction of nitrate in soil; N^O, however,
is the terminal product of reduction in several bacterial strains. It may be
expected, from the work of Cady and Bartholomew (1961) on N~ production in
soil, that NO production will increase as soil oxygen decreases. Under anoxic
conditions, however, Cady and Bartholomew found that N»0 was reduced to NO.
Bailey (1976) found that increases in soil temperature resulted in decreases
in NO production.
In addition to biochemical processes, there is evidence that purely
chemical reactions in the soil produce N_ and NO under certain conditions
^ /\
(Delwiche and Bryan, 1976; Porter, 1971).
019WPS/B 3-79 6/28/84
-------
Stevenson et al. (1970) showed evidence that NO, N?0, and N? could be
produced by nitrosation of humic and fulvic acids, lignins, and aromatic
substances at pH 6.0 and 7.0 in the absence of oxygen. Steen and Stojanovic
(1971) found that NO was volatilized from a calcareous soil when high concen-
trations of urea were nitrified with concurrent accumulation of nitrite, and
assumed that nitrosation between nitrous acid and organic matter was the main
pathway by which NO was formed.
Wullstein and Gilmour (1964, 1966) reported that nitrite reacted with
certain reduced transition metals in sterile, moderately acid systems to yield
NO as a primary gaseous product.
Bremner and Nelson (1968) found that N« and N0? and small amounts of N^O
were formed when nitrite was added to neutral and acid soils. They suggested
that the reactions between soil organic constituents and nitrite were responsi-
ble for the formation of N« and N^O, while self-decomposition of HNO» was
responsible for the formation of NO and NO,,. In steam-sterilized raw humus
samples incubated with nitrite, NO was the predominant gaseous reaction product
(Nommick and Thorin, 1971). Nelson and Bremner (1970a, 1970b) found that the
formation of N0? by decomposition of nitrite in pH 5.0 solution was not promoted
by organic or inorganic soil constituents and concluded that most of the N0»
evolved was formed by self-decomposition of HNO^. The amount of NOp formed
was inversely related to soil pH, but pH had little effect on the amount of
nitrite converted to nitrate. These findings led to the conclusion that the
self-decomposition reaction of HN02 was best represented by the equation:
2HN02 > NO + N02 + H20 (3-4)
Although more is known about the natural biological and chemical processes
in the environment that produce emissions of nitrogen oxides than those that
produce VOC, the problems with actually quantifying such emissions exceed the
problems associated with quantifying natural emissions of VOC. As indicated
in section 3.3, the limits of detection of methods for measuring nitrogen
oxides are such that low-level measurements are often not reliable. Techniques
for estimating NO emissions from such sources as lightning and biological
/\
processes are virtually nonexistent. In addition, scaling of emissions from
such sources as bacterial nitrification and denitrification for use in preparing
area-wide emission inventories is not possible. Thus, the emissions reported
019WPS/B 3-80 6/28/84
-------
in this section should be taken as very gross approximations that serve to
identify natural sources of NO and to present the relative magnitude of
emissions from such sources.
3.5 REPRESENTATIVE CONCENTRATIONS OF OZONE PRECURSORS IN AMBIENT AIR
Nonmethane organic compounds (NMOC) and the oxides of nitrogen (NO ) in
/\
the presence of sunlight react to form ozone and other photochemical oxidants.
The reaction sequence is very complex, and therefore, it is difficult to
establish dependable precursor-oxidant relationships. Factors such as absolute
NMOC and NO concentrations, relative NMOC and NO concentrations, NMOC reactiv-
A X
ity, and NO composition are known to effect the photochemical reactions that
/\
produce ozone and other oxidants in ambient atmospheres. The purpose of this
section is to provide a summary of NMOC and NO concentrations recorded at
various urban and nonurban locations in the United States.
3.5.1 Concentrations of Nonmethane Orgam'cs Compounds in Ambient Air
Automated total hydrocarbon analyzers have been employed for many years
to measure ambient hydrocarbon concentrations. The accuracy of data obtained
with these analyzers is questionable, however, because the methodology does
not provide a direct measure of the nonmethane organic fraction. Rather, a
total hydrocarbon value that includes methane is obtained and the methane
concentration must be determined and substracted from the total. The indirect
nature of the measurement, along with calibration difficulties, limits the
usefulness of data obtained with total hydrocarbon analyzers. Consequently,
nonmethane organic data obtained in this way will not be utilized in this
discussion.
Gas chromatographic methods are now available that allow identification
and quantification of individual hydrocarbon species. Total hydrocarbon con-
centrations are then derived by summation. This procedure works well for the
"true" hydrocarbons, which contain only carbon and hydorgen atmos. Low-molecular-
weight aldehydes are particularly troublesome and thus must be measured using
alternate methods. Originally, ambient aldehyde concentrations were determined
using chemical methods that were specific for formaldehyde and the combined
series of aliphatic aldehydes. In more recent years, analytical methods employ-
ing liquid chromatography have been developed for measuring the concentration
of individual aldehydes in ambient atmospheres.
019WPS/B 3-81 6/28/84
-------
For discussion purposes, the ambient NMOC data will be segregated in this
section into two groups: (1) nonmethane hydrocarbons, and (2) oxygenated
hydrocarbons. There is a fairly substantial data base for characterizing
urban nonmethane hydrocarbon concentrations. Measurements of nonurban hydrocar-
bon levels, as well as both nonurban and urban oxygenated hydrocarbons, are
much more limited. In the latter group, aldehydes have received the most
attention. There is insufficient information available for establishing
ambient air concentrations of other classes of oxygenated hydrocarbons such as
alcohols, ketones, acids, and ethers.
3.5.1.1 Urban Nonmethane Hydrocarbon Concentrations. Most of the ambient air
nonmethane hydrocarbon (NMHC) data have been obtained during the 6:00 to
9:00 a.m. time period. Since urban hydrocarbon emissions peak during that
period of the day and atmospheric dispersion is limited, these concentrations
generally reflect maximum diurnal levels. Table 3-20 lists the mean and range
of NMHC concentrations recorded in a number of urban areas throughout the
United States. For most urban areas included in the table, a mean NMHC value
between 400 and 900 ppb C was observed. It is obvious, however, that Houston,
Las Vegas, and Los Angeles exhibit mean values in excess of 1000 ppb C. The
data in Table 3-20 are not meant to serve as a comparison of NMHC levels in
various cities but rather are shown to indicate the mean and range of concentra-
tions that have been reported. Comparisons are invalid because of major
differences in sample numbers, site classifications, and seasonal sampling
periods. It can be seen that the range of NMHC concentrations measured in
some urban areas can vary by as much as two orders of magnitude. For example,
the lowest and highest NMHC concentrations recorded in Milwaukee were 24 ppb C
and 3116 ppb C, respectively. In many cases, the range of values reported in
Table 3-20 might not reflect the true maximum and minimum concentrations that
occur in a particular urban area. Most of the hydrocarbon sampling programs
were of short duration (~1 month) and in some cases were not operated on a
daily basis. For example, the relatively high mean values reported in Las
Vegas are undoubtedly the result of the fact that ambient air samples were
only analyzed for hydrocarbons on days when conditions were appropriate for
oxidant formation. It is probably safe to assume, however, that NMHC levels
during the 6:00 to 9:00 a.m. time period in major urban areas will usually
exceed 50 ppb C but seldom surpass 10,000 ppb C.
OZONER/D 3-82 6/28/84
-------
The hydrocarbon composition of urban atmospheres is dominated by species
in the C2 - C-.Q molecular-weight range. The alkanes are most prominent,
followed by aromatics and alkenes. Based on speciation data obtained in
several urban areas, alkanes generally constitute 50 to 60 percent of the
hydrocarbon burden, aromatics 20 to 30 percent, with alkenes and acetylene
making up the remaining 5 to 15 percent (Sexton and Westberg, 1984). The
alkane fraction is usually dominated by species in the C. - Cg molecular-weight
range: v-butane, rrbutane, j^-pentane, ri-pentane, 2-methylpentane, 3-methylpen-
tane, hexane, etc. Predominant aromatics include benzene, toluene, ethylbenzene,
and the xylenes. Ethylene and propene are normally the most abundant alkenes,
with lesser amounts of the isomeric occurring. Table 3-21 shows a typical
example of the hydrocarbon composition recorded in urban atmospheres.
3.5.1.2 Nonurban Nonmethane Hydrocarbon Concentrations . Nonurban nonmethane
hydrocarbon concentrations are generally one to two orders of magnitude lower
than those measured in urban areas (Ferman, 1981; Sexton and Westberg, 1984).
On an individual species basis, concentrations seldom exceed 10 ppb C; total
hydrocarbon concentrations range up to ~150 ppb C, but usually fall in the range
of about 5 to 100 ppb C. Alkanes comprise the bulk of species present, with
ethane, propane, n-butane, v-pentane, and n-pentane most abundant. Ethylene and
propene are occasionally reported at concentrations of 1 ppb C or less, and
toluene is usually present at ~1 ppb C. Table 3-22 provides a summary of the
range of hydrocarbon concentrations measured at various nonurban locations in
the United States. Samples were carefully selected at most of the sites in
order to guarantee their nonurban character. At the coastal and near-coastal
sites, only those samples collected upwind of manmade sources (onshore advection)
were included. The nonmethane hydrocarbon concentrations reported at coastal
sites (Belfast, Benicia, Miami, and Houston) are definitely lower than those
measured at most of the inland sites. It should be pointed out, however, that
the numbers of samples measured for each of the nonurban locations listed in
Table 3-22 is small. This, coupled with the fact that only a limited range of
hydrocarbons were monitored in some cases, makes intersite comparisons tenuous
at best.
In recent years there has been considerable interest in the ambient air
concentrations of naturally emitted hydrocarbons. The species most often
reported include isoprene, a-pinene, p-pinene, A-carene, and limonene. These
species are generally reported only in nonurban hydrocarbon sampling programs.
OZONER/D 3-83 6/28/84
-------
TABLE 3-20. NONMETHANE HYDROCARBON CONCENTRATIONS
MEASURED BETWEEN 6:00 and 9:00 a.m. IN VARIOUS UNITED STATES CITIES
Mean
City NMHC
(Date) cone. , ppb C
Atlanta (1981)
Baltimore (1980)
Boston (1980)
Cincinnati (1981)
Houston (1976)
Houston (1978)
Las Vegas (1980)
Las Vegas (1983)
Los Angeles (1968)
Los Angeles (1982)
Milwaukee (1981)
Newark (1980)
New York (1969)
Philadelphia (1979)
St. Louis (1973)
Tulsa
Washington, D.C. (1980)
491
659
569
840
1414
1679
2506
2762
3388
2920
324
732
830
669
817
426
671
Range
113 to 1677
51 to 2798
83 to 4750
260 to 1870
356 to 16,350
400 to 4500
689 to 4515
1835 to 4590
_ _ _
390 to 6430
24 to 3116
89 to 6946
_ _ _
305 to 1710
_ _ _
103 to 3684
210 to 2953
Reference
Westberg and Lamb (1983)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Holdren et al. (1982)
Sexton and Westberg (1984)
Lonneman (1979)
Nay lor et al. (1981)
Nay! or et al . (1984)
Lonneman (1977)
Grosjean and Fung (1984)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Lonneman (1977)
Sexton and Westberg (1984)
Lonneman (1977)
Eaton et al. (1979)
Sexton and Westberg (1984)
OZONER/D
3-84
6/28/84
-------
TABLE 3-21. HYDROCARBON COMPOSITION TYPICALLY MEASURED IN URBAN AREAS
(From sample collected in Milwaukee, 1981)
ppb C
Hydrocarbon
ppb C
Hydrocarbon
14.0 Ethane
25.5 Ethylene
16.0 Acetylene
18.5 Propane
9.5 Propene
28.5 i-Butane
65.0 jr-Butane
2.0 1-Butene
3.5 J^-Butene
3.5 t-2-Butene
c-2-Butene
49.0 i-Pentane
24.0 n-Pentane
2.0 1-Pentene
t-2-Pentene
0.5 c-2-Pentene
Cyclopentene
2.0 Cyclopentane
4.5 2,3-Dimethyl butane
13.0 2-Methy1pentane
c-4-Methyl-2-pentene
10.0 3-Methy1pentane
1-Hexene
11.0 n-Hexane
t-2-Hexene
c-2-Hexene
6.5 Methylcyclopentane
4.5 2,4-Dimethylpentane
9.5 Benzene
2.0 Cyclohexane
4.5
5.5
8.0
5.0
8.0
3.0
1.5
1.5
33.5
2.5
2.5
2.5
6.5
18.5
6.5
4.0
3.0
5.0
3.0
22.0
3.0
17.0
7.0
2-MethyIhexane
2,3-Di methylpentane
3-MethyIhexane
2,2,3-Trimethylpentane
n-Heptane
Methylcyclohexane
2,4-Dimethylhexane
2,3,4-Tri methylpentane
Toluene
2,3-DimethyIhexane
2-Methylheptane
3-EthyIhexane
n-Octane
Ethylcyclohexane
Ethylbenzene
£, m-Xylene
Styrene
o-Xylene
n-Nonane
T-Propylbenzene
in- Propyl benzene
£-£thyltoluene
m-Ethyltoluene
o-Ethyl toluene
1,3,5-Tri methyl benzene
1,2,4-Trimethyl benzene
1,2,3-Tri methyl benzene
Methylstyrene
1,3-Di ethyl benzene
1,4-Diethylbenzene
Total identified
hydrocarbons
I Olefin
I Aromatic
I Paraffin
Acetylene
ppb C
46
134
301
16
497
9
27
60
3
ppb C
Total unidentified
hydrocarbons
Total NMHC by
summing individual
species
87
584
Source: Westberg and Lamb (1982)
OZONER/D
3-85
6/28/84
-------
TABLE 3-22. NONMETHANE HYDROCARBON CONCENTRATIONS
MEASURED IN NONURBAN ATMOSPHERES
Location
Species
analyzed
Concentration
range, ppb C
Reference
Belfast, ME
Benicia, CA
Miami, FL
Glascow, IL
Janesville, WI
Houston, TX
Robinson, IL
Smoky Mtns.
Northern Idaho
Virginia
Atlanta (urban)
Whiteface Mtn., NY
Elkton, MO
Southern WA
Eastern TX
North Carolina
Colorado
r - r
C2 C5
C2 - C10
C2 - C10
C2 * C10
C2 * C10
C2 - C10
Terpenes
Isoprene
Isoprene
Terpenes
Isoprene
Isoprene
a-pinene
a-pinene
Terpenes
10 to 22
7 to 14
2 to 23
60 to 150
9 to 24
2 to 24
13 to 113
38 to 149
0.1 to 18
4 to 150
0 to 8
6 to 84
0 to 28
0.1 to 8
0.6 to 13
0 to 8
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Chatfield and Rasmussen (1977)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Cronn (1982)
Holdren et al. (1979)
Ferman (1981)
Westberg and Lamb (1983)
Whitby and Coffey (1977)
Rasmussen et al. (1976)
Allwine et al. (1983)
Seila (1981)
Seila (1981)
Roberts et al. (1983)
OZONER/D
3-86
6/28/84
-------
Because they are present at very low concentrations, it is extremely difficult
to establish positive identifications when the natural hydrocarbons mix with
manmade emissions in an urban area. The one exception is isoprene, which has
been reported in both urban and nonurban sampling programs. Monoterpene
(C-.-H.,..) concentrations in ambient air seldom exceed 20 ppb C. Average concen-
1U ID
trations of orpinene, the most commonly reported monoterpene, are usually
below 10 ppb C. During the summer months, isoprene concentrations as high as
150 ppb C have been measured (Ferman, 1981). Maximum concentrations in the 30
to 40 ppb C range, however, are more common. Ambient concentrations of these
naturally emitted hydrocarbons are very site-dependent. The highest concentra-
tions are observed in or are immediately adjacent to forested areas. Seasonal
variations will exist, as well, because natural hydrocarbon emission fluxes
are directly related to the amount of biomass present and increase with tempera-
ture. In a recent review article, Altshuller (1983) has provided a more
detailed discussion of natural hydrocarbons and their effect on air quality.
3.5.1.3 Nonmethane Hydrocarbon Concentrations Aloft. Hydrocarbon concentra-
tions in the layer above a morning surface inversion and below the afternoon
mixing level are of interest because oxidant precursors in this layer mix with
urban plumes following breakup of the surface inversion. Table 3-23 provides
a listing of mean hydrocarbon concentrations and the range of concentrations
observed aloft in the vicinity of several United States cities. Data included
in Table 3-23 were obtained from samples collected between 6:00 and 10:00 a.m.
at altitudes between 1000 and 5000 ft above the surface. Mean NMHC values
vary from about 10 ppb C to nearly 50 ppb C, with individual samples spanning
the range of approximately 10 to 100 ppb C. These data were acquired during
oxidant study programs and therefore represent hydrocarbon concentrations
aloft during summertime periods when oxidant episodes are most likely to
occur.
The hydrocarbon content of samples collected aloft is dominated by alkanes.
Based on the 150 or so samples summarized in Table 3-23, alkanes were about
75 percent of the total NMHC, aromatics accounted for about 15 percent, and
alkenes the remaining 10 percent.
3.5.1.4 Urban Aldehyde Concentrations. Aldehydes are produced in urban
atmospheres by photochemical reactions. In addition to this jji situ source,
aldehydes are emitted by combustion sources and, thus, enter the atmosphere as
primary pollutants. Most of the ambient aldehyde data are restricted to
OZONER/D 3-87 6/28/84
-------
TABLE 3-23. NONMETHANE HYDROCARBON CONCENTRATIONS IN SAMPLES COLLECTED
ALOFT (1000 to 5000 ft) DURING MORNING HOURS (6:00 to 10:00 a.m.)
Location Mean
and date NMHC cone. , ppb C
Atlanta (1981)
Baltimore (1980)
Boston (1980)
Canton, OH (1974)
Groton, CT (1975)
Houston (1978)
Milwaukee (1981)
New York (1980)
Philadelphia (1979)
Phoenix (1973)
Tulsa
Washington, D.C. (1980)
19
41
19
34
23
35
22
50
32
32
44
36
Range of
NMHC concns. , ppb C
9
11
4
24
13
14
10
11
21
12
13
11
to
to
to
to
to
to
to
to
to
to
to
to
41
90
42
48
41
81
66
88
59
47
73
65
No.
samples
14
28
11
15
8
15
9
18
6
7
10
15
Source: Adapted from Westberg and Allwine (1984).
OZONER/D 3-88 6/28/84
-------
formaldehyde concentrations measured by the chromotropic acid (CA) method and
total aliphatic aldehydes obtained by means of the 3-methyl-2-benzothiazolone
(MBTH) monitoring procedure. Long-path infrared techniques have provided some
ambient aldehyde data for the Los Angeles area. Analytical procedures that
permit the routine characterization of individual aldehydes have only recently
been reported. When ambient air is passed through a dinitrophenylhydrazine
solution, aldehydes are converted to hydrazones that can be separated and
quantitatively determined by liquid chromatography (HPLC). Ambient aldehyde
data that have been acquired from all of these monitoring procedures will be
utilized in the ensuing discussion.
Aldehydes observed in urban atmospheres include formaldehyde, acetaldehyde,
acrolein, chloral, propanal, n-butanal, and benzaldehyde. Formaldehyde concen-
trations have been best characterized because the CA monitoring methodology
was established in the early 1960s. Table 3-24 shows formaldehyde levels
recorded in a number of United States cities. Most of the studies referenced
in this table were of short duration. Consequently, the mean concentrations
that are listed may not be representative of seasonal or annual means. Even
from this limited data base, however, it is apparent that urban formaldehyde
concentrations are low. With the exception of the 1961 Los Angeles data, the
reported mean values fall in the 10 to 30 ppb range, with maximum concentrations
ranging up to 90 ppb. Since mean nonmethane hydrocarbon (NMHC) levels in many
of these same cities range between 400 and 900 ppb C, formaldehyde probably
constitutes less than 3% of the total NMOC (NMHC plus aldehydes) in most urban
areas. This supposition is supported by recent work in Los Angeles which
showed a mean NMHC concentration of approximately 2900 ppb C and, during the
same time period, a mean formaldehyde level of 27.5 ppb C. In this case, the
formaldehyde amounted to less than 1 percent of the total NMOC.
Diurnal monitoring programs in Los Angeles and other cities indicate that
formaldehyde concentrations are elevated during the morning rush-hour traffic
period and again during the afternoon hours when ozone levels are high.
Generally, the highest formaldehyde levels coincide closely with peak ozone
concentrations. In the Los Angeles basin, formaldehyde concentrations in the
20 ppb range have been observed throughout the nighttime hours (Tuazon et al.,
1981).
Acetaldehyde concentrations are generally below formaldehyde levels in a
given urban area. Data collected in Las Vegas during 1980 and 1981 showed an
OZONER/D 3-89 6/28/84
-------
TABLE 3-24. FORMALDEHYDE CONCENTRATIONS IN SEVERAL UNITED STATES CITIES
City
Los Angeles
Los Angeles
Los Angeles
Columbus, OH
Baltimore
Denver
St. Louis
Chicago
Riverside
New York
Pittsburgh
Houston
Las Vegas
Mean Concentration
concentration, range, Measurement
Date ppb ppb method
1961
1978
1981
1980
1980
1980
1980
1981
1980
1981
1981
1978
1981
40
23
28
8
27
12
11
13
19
14
21
10
11
5-160
6-71
4-86
1-24
1-87
7-29
8-19
9-17
10-41
7-46
13-35
0-35
CA
Long-path IR
HPLC
CA
CA
CA/HPLC
CA/HPLC
CA/HPLC
CA/HPLC
CA/HPLC
CA/HPLC
CA
HPLC
Reference
a
b
c
d
d
e
e
e
e
e
e
f
g
Altshuller and McPherson (1963).
bTuazon et al. (1978).
CGrosjean and Fung (1984).
dJoshi et al. (1981).
eSingh et al. (1982).
fJoshi et al. (1979).
gNaylor et al. (1981).
OZONER/D
3-90
6/28/84
-------
average acetaldehyde concentration of 6 ppb, with a range from 0 to 14 ppo.
This compares to an average formaldehyde concentration of 11 ppb during the
same study period. During the fall months of 1981 in Los Angeles, acetaldehyde
concentrations ranged between 2 and 39 ppb, while the formaldehyde levels
varied between 4 and 86 ppb. On a ppb C basis, acetaldehyde comprised 23 per-
cent of average carbonyl content and 0.6 percent of the total NMOC measured in
the 1981 Los Angeles study (Grosjean and Fung, 1984).
Ambient air concentrations of higher-molecular-wsight aldehydes were
reported in both the Los Angeles and Las Vegas studies referred to previously.
In Los Angeles, C_ carbonyls (propanal, acetone, and acrolein) varied in
«3
concentration from 1 to 54 ppb, butanal ranged between 0 and 5 ;?.pb; and benzsl-
dehyde concentrations never exceeded 2 ppb. Propana'i concentrations up to
about 1 ppb and benzaldehyde levels ranging between 0 and 0.6 ppb were measured
in Las Vegas. Chloral (trichloroacetaldehyde) was detected in ambient air in
Las Vegas at an average concentration of 1 ppb, with values for 18 samples
ranging between 0 and 5 ppb. Acrolein, which "is believed to cause eye irrita-
tion, has been measured in Los Angeles air at concentrations up to 15 ppb.
Since the early 1960s, the MBTH method has been employed in urban environ-
ments to measure total aliphatic aldehyde concentrations. The total aldehyde
concentrations determined in this way include formaldehyde, which is generally
the predominant aldehyde present. In studies where simultaneous formaldehyde
(CA method) and total aliphatic aldehyde (MBTH method) data are available,
formaldehyde usually accounts for more than 50 percent of the total aldehydes.
Measurements in Columbus, Ohio, during the summer of 1980 showed a mean formal-
dehyde level of 7.9 ppb and a mean total aldehyde concentration of 13 ppb.
In Houston, Joshi (1979) found an average of 10 ppb formaldehyde and 16.4
total aldehydes. These were short-term studies (1 to 2 months) conducted
during the oxidant season. On an annual basis, the formaldehyde contribution
is even more significant. For example, the annual mean total aldehyde concen-
tration (MBTH 24-hour method) averaged across all Houston monitoring stations
in 1974 was identical to the mean formaldehyde concentration.
In summary, it appears that urban aldehyde concentrations can vary from a
few ppb up to about 200 ppb. Formaldehyde is present in highest concentration,
followed by acetaldehyde. In polluted atmospheres, acrolein, propanal, butanal,
and benzaldehyde have been measured at concentrations less than 15 ppb. Where
concurrent nonmethane hydrocarbon (NMHC) data are available, aldehydes average
about 3 percent of the total nonmethane organic (NMOC) species present.
OZONER/D 3-91 6/28/84
-------
3.5.1.5 Aldehyde Concentrations in Rural Atmospheres. Formaldehyde levels in
remote atmospheres have been reported to range between 0.1 and 10 ppb. Mean
concentrations representative of global background conditions vary from 0.3 to
0.5 ppb (Duce et al. , 1983). Very few total aldehyde measurements have been
made at rural locations in the United States. Breeding et al. (1973) reported
values of 1 to 2 ppb in rural Illinois and Missouri. It is expected that
certain higher-molecular-weight aldehydes (Cfi-C.._) emitted by natural vegeta-
tion should be present in the gas phase; however, ambient concentrations of
these species have not been reported.
3.5.2 Atmospheric Concentrations of Nitrogen Oxides
Ambient air levels of nitrogen oxides have been monitored throughout the
United States for a number of years. Since N0? is the only oxide of nitrogen
for which a NAAQS has been promulgated, it has received the greatest attention.
The National Aerometric Data Bank at EPA, which receives ambient air data from
federal, state and local monitoring stations, contains the most comprehensive
collection of aerometric data for the United States. The air quality criteria
document for oxides of nitrogen (U.S. Environmental Protection Agency, 1982)
included a thorough discussion of seasonal and annual trends in ambient NO
J\
concentrations for a large number of United States cities. This type of
information will not be repeated in this present discussion. The emphasis
here will be on NO measurements that can be related to the diurnal photochemi-
cal processes that produce ozone.
3.5.2.1 Urban NO.. Concentrations. Concentrations of NO , like hydrocarbon
A "~ "
concentrations, tend to peak in urban areas during the early morning period
when atmospheric dispersion is limited and automobile traffic is dense. Most
of the NO is emitted as nitric oxide (NO) and, thus, in the absence of chemi-
^k
cal reactions NO would be expected to be the predominant oxide of nitrogen
present. Nitric oxide is converted rapidly, however, to NO,, by ozone and
peroxy radicals produced in atmospheric photochemical reactions. Since both
the abundance of ozone and the photochemical activity vary diurnally and from
day to day, the relative concentrations of NO and NOp can fluctuate signifi-
cantly. As a general rule, urban NO concentrations peak during the 6:00 to
9:00 a.m. period in the morning. This is followed by a rapid decrease caused
by the photochemical conversion of NO and NO^ and increased atmosphere mixing.
Nitric oxide levels remain low during the daytime period and then usually
OZONER/D 3-92 6/28/84
-------
begin to build up again through the nighttime hours. Nitrogen dioxide concen-
trations typically increase during the mid-morning hours and then abate as the
afternoon progresses. Levels of NO^ begin to increase again following the
late afternoon rush-hour period and often continue to climb during the night-
time.
The average NO concentration in urban areas of the United States is
about 70 ppb, with nitric oxide and nitrogen dioxide contributing about equally
(Logan, 1983). Monitoring data for 1975 through 1980 showed that peak 1-hr
H0? concentrations equalled or exceeded 400 ppb in Los Angeles and several
other California locations, as well as at sites in Kentucky (Ashland) and
Michigan (Port Huron). Cities with one peak hourly concentration exceeding
270 ppb during those years include Phoenix, St. Louis, New York City, Spring-
field, IL, Cincinnati, Saginaw, and Southfield, MI, and more than a dozen sites
in California. Reported hourly concentrations in excess of 140 ppb were quite
common nationwide during the years between 1975 and 1980 (U.S. Environmental
Protection Agency, 1982).
Urban NO concentrations during the 6:00 to 9:00 a.m. period are of
primary importance in terms of oxidant production. Average NO levels recorded
in several urban areas during this morning period are listed in Table 3-25.
Most of these data were collected in special, field study programs that were
designed to provide information concerning ozone and ozone-precursor relation-
ships. Thus, concurrent 6:00 to 9:00 a.m. hydrocarbon samples were also
obtained, which permits the calculation of hydrocarbon-NO ratios in each of
these urban areas. These data are included in Table 3-25.
As can be seen in Table 3-25, mean 6:00 to 9:00 a.m. NO concentrations
in the 10 cities listed fall in the range of about 50 to 150 ppb. Hydrocarbon
concentrations (ppb C) exceeded the oxides of nitrogen levels by a factor of 5
to 16 during this same time period. Smog chamber experiments indicate that
significant quantities of ozone can be produced when HC/NO ratios are in this
/\
range. Threfore, when meteorological conditions are appropriate, it is expected
that an ozone buildup will occur in plumes emanating from these cities.
Indeed, ozone production has been observed in the vicinity of most of the
cities referenced in Table 3-25.
3.5.2.2 Nonurban NO Concentrations. Concentrations of NO in "clean" remote
environments are usually below 0.5 ppb (Logan, 1983). For example, median
concentrations measured on Niwot Ridge in Colorado are about 0.3 ppb in the
OZONER/D 3-93 6/28/84
-------
TABLE 3-25. AVERAGE 6:00 to 9:00 a.m. NO CONCENTRATIONS AND
HC/NO RATIOS IN URBAN AREAS
City
Atlanta
Baltimore
Boston
Houston
Linden, NJ
Los Angeles
Milwaukee
St. Louis
Tulsa
Washington, D.C.
Average NO ,
ppb
57
85
63
125
59
147
66
77
46
94
Average
HC/NO
/\
9
10
10
13
16
10
5
8
13
14
References
Westberg and Lamb (1983)
Richter (1983)
Ri enter (1983)
Westberg et al . (1978)
Richter (1983)
U.S. Environmental
Protection Agency (1978)
Westberg and Lamb (1983)
EPA (1978)
Eaton et al . (1979)
Richter (1983)
OZONER/D
3-94
6/28/84
-------
summer and 0.24 ppb in winter. In exceptionally clean air, NO concentrations
J\
as low as 0.015 have been recorded (Bellinger et al., 1982). Slightly higher
NO concentrations have been reported at other remote locations in the western
United States and Canada. Kelly et al. (1982) deduced a mean NO concentration
?\
of about 1 ppb from measurements in South Dakota. At the South Dakota site,
nitric oxide generally contributed less than 20 percent of the total NO .
s\
Measurements of NO during the 1970s at rural locations in Montana (Decker
A
et al., 1978) and Saskatchewan (McElroy and Kerr, 1977) yielded average concen-
trations similar to those recorded in South Dakota.
In the northeastern United States, nonurban NO concentrations appear to
X
exceed those in the west by about a factor of ten. A median NO concentration
/\.
of 6.6 ppb was derived from data collected at nine rural sites utilized in the
Sulfate Regional Experiment (SURE) program (Mueller and Hidy, 1983). Median
concentrations at the individual stations, which extended eastward from the
Ohio River Valley to the Atlantic Coast, varied from 2 to 11 ppb. Measurements
at nonurban sites in Pennsylvania and Louisiana during the summer of 1975
showed mean hourly NO concentrations of 4.7 and 4.1 ppb, respectively (Decker
/\
et al., 1978). Nitric oxide composed approximately 40 percent of the total NO
^\
at these latter two nonurban sites.
In summary, it appears from the limited amount of data available that NO
/\
concentrations in unpopulated, rural regions of the western United States
average 1 ppb or less. At nonurban locations in the more industralized eastern
United States, average NO concentrations can exceed 10 ppb.
3.6 SUMMARY
3.6.1 Nature of Precursors to Ozone and Other Photochemical Qxidants
Photochemical oxidants are products of atmospheric reactions involving
volatile organic compounds (VOC), oxides of nitrogen (NO ), hydroxyl radicals,
/\
oxygen, and sunlight. They are almost exclusively secondary pollutants,
formed in the atmosphere from their precursors by processes that are a complex
function of precursor emissions and meteorological factors.
Although vapor-phase hydrocarbons (compounds of carbon and hydrogen only)
are the predominant organic compounds in the ambient air that serve as precur-
sors to photochemical oxidants, other volatile organic compounds are also
photochemically reactive in those atmospheric processes that give rise to the
OZONER/D 3-95 6/28/84
-------
oxidants. In particular, halogenated organics (e.g., haloalkenes) that parti-
cipate in photochemical reactions are present in ambient air, although at lower
concentrations than the hydrocarbons. They are apparently oxidized through the
same initial step involved in the oxidation of the hydrocarbons; that is, attack
by hydroxyl radicals (H0«)- Alkenes, haloalkenes, and aliphatic aldehydes
are, as classes, among the most reactive organic compounds found in ambient
air. Alkenic hydrocarbons and halocarbons are unique among VOC in ambient air
in that they are susceptible both to attack by HO- and to ozonolysis (oxida-
tion by ozone) (Niki et al., 1983). Methane, halomethanes, and certain
haloethenes are of negligible reactivity in ambient air and have been classed
as unreactive by the U.S. Environmental Protection Agency (1980).
The oxides of nitrogen that are important as precursors to ozone and
other photochemical oxidants are nitrogen dioxide (N02) and nitric oxide (NO).
Nitrogen dioxide is itself an oxidant that produces deleterious effects, which
are the subject of a separate criteria document (U.S. Environmental Protection
Agency, 1982). Nitrogen dioxide is an important precursor to ozone and other
photochemical oxidants (1) because its photolysis in ambient air leads to the
formation of oxygen atoms that combine with molecular oxygen to form ozone;
and (2) because it reacts with acetylperoxy radicals to form peroxyacetyl
nitrate (PAN), a relatively potent phytotoxicant and lachrymator. Although
ubiquitous, nitrous oxide (N?0) is unimportant in the production of oxidants
in ambient air because it is virtually inert in the troposphere. (In the stra-
tosphere, where the wavelength distribution is different, N_0 is photolyzed.)
Since methane is considered only negligibly reactive in ambient air, the
volatile organic compounds of importance as oxidant precursors are usually
referred to as nonmethane hydrocarbons (NMHC) or, more properly, as nonmethane
organic compounds (NMOC).
3.6.2 Measurement of Precursors to Ozone and Other Photochemical Oxidants
Numerous analytical methods have been employed to determine nonmethane
organic compounds (NMOC) in ambient air. To present an overview of the most
pertinent information, measurement methods for the organic species may be
arranged in three major classifications: nonmethane hydrocarbons, aldehydes,
and other oxygenated compounds.
OZONER/D 3-96 6/28/84
-------
Nonmethane hydrocarbons have been determined primarily by methods that
employ a flame ionization detector (FID) as the sensing element. Early methods
for the measurement of total nonmethane hydrocarbons did not provide for
speciation of the complex mixture of organics in ambient air. These methods,
still in use for anaysis of total nonmethane organic compounds, are essentially
organic carbon analyzers, since the response of the FID detector is essentially
proportional to the number of carbon atoms present in the organic molecule
(Sevcik, 1975). Carbon atoms bound, however, to oxygen, nitrogen, or halogens
give reduced relative responses (Dietz, 1967). This detector has been utilized
both as a stand-alone continuous detection system (non-speciation) and also
with gas chromatographic techniques that provide for speciation of the many
organics present in ambient air. A number of studies of non-speciation analyz-
ers have indicated an overall poor performance of the commercial instruments
when calibration or ambient mixtures containing NMOC concentrations less than
1 ppm C were analyzed (e.g., Reckner, 1974; McElroy and Thompson, 1975; Sexton
et al., 1981). The major problems associated with the non-speciation analyzers
have been summarized in a recent technical assistance document published by
the U.S. Environmental Protection Agency (1981). The document also presents
ways to reduce some of the existing problems.
Because of the above deficiencies, other approaches to the measurement of
nonmethane hydrocarbons are currently under development. The use of gas
chromatography coupled to an FID system circumvents many of the problems
associated with continuous non-speciation analyzers. This method, however,
requires sample preconcentration because the organic components are present at
part-per-billion (ppb) levels or lower in ambient air. The two main preconcen-
tration techniques in present use are cryogenic collection and the use of
solid adsorbents (McBride and McClenny, 1980; Jayanty et al., 1982; Westberg
et al., 1980; Ogle et al. , 1982). The preferred preconcentration method for
obtaining speciated data is cryogenic collection. Speciation methods involv-
ing cryogenic preconcentration have also been compared with continuous non-
speciation analyzers (e.g., Richter, 1983). Results indicate poor correlation
between methods at ambient concentrations below 1 part-per-million carbon (ppm
C).
Aldehydes, which are both primary and secondary pollutants in ambient
air, are detected by total NMOC and NMHC speciation methods but can not be
quantitatively determined by those methods. Primary measurement techniques
OZONER/D 3-97 6/28/84
-------
for aldehydes include the chromotropic acid (CA) method for formaldehyde
(Atlshuller and McPherson, 1963; Johnson et al., 1981), the 3-methyl-2-benz-
othiazolone (MBTH) technique for total aldehydes (e.g., Sawicki et al., 1961;
Hauser and Cummins, 1964), Fourier-transform infrared (FTIR) spectroscopy
(e.g., Hanst et al., 1982; Tuazon et al., 1978, 1980, 1981), and high-perform-
ance liquid chromatography employing 2,4-dinitrophenyl-hydrazine derivatization
(HPLC-DNPH) for aldehyde speciation (e.g., Lipari and Swarin, 1982; Kuntz et
al., 1980). The CA and MBTH methods utilize wet chemical procedures and
spectrophotometric detection. Interferences from other compounds have been
reported for both techniques. The FTIR method offers good specificity and
direct i_n situ analysis of ambient air. These advantages are offset, however,
by the relatively high cost and lack of portability of the instrumentation.
On the other hand, the HPLC-DNPH method not only offers good specificity but
can also be easily transported to field sites. A few intercomparison studies
of the above methods have been conducted and considerable differences in
measured concentrations were found. The data base is still quite limited at
present, however, and further intercomparisons are needed.
Literature reports describing the vapor-phase organic composition of
ambient air indicate that the major fraction of material consists of unsubsti-
tuted hydrocarbons and aldehydes. With the exception of formic acid, other
oxygenated species are seldom reported. The lack of oxygenated hydrocarbon
data is somewhat surprising since significant quantities of these species are
emitted into the atmosphere by solvent-related industries and since at least
some oxygenated species appear to be emitted by vegetation. In addition to
direct emissions, it is also expected that photochemical reactions of hydro-
carbons with oxides of nitrogen, ozone, and hydroxyl radicals will produce
significant quantities of oxygenated products. The adsorptive nature of the
surfaces that contact these oxygenated species during sample collection and
analysis may account for the apparent lack of data. Attempts have been made
to decrease adsorption by deactivating the reactive surface or by modifying
the compound of interest (Osman et al. , 1979; Westbert et al., 1980). Addi-
tional research efforts should focus on this area.
Aside from the essentially unreactive N20, only two oxides of nitrogen
occur in ambient air at appreciable concentrations: nitric oxide (NO) and
nitrogen dioxide (N0_). Both compounds, together designated as N0x, partici-
pate in the cyclic reactions in the atmosphere that lead to the formation of
ozone.
OZONER/D 3-98 6/28/84
-------
The preferred means of measuring NO and N0~ is the chemiluminescence
method (U.S. Environmental Protection Agency, 1976). The measurement principle
is the gas-phase chemi luminescent reaction of 0- and NO (Fontijn et al.,
1970). While NO is determined directly in this fashion, N02 is detected
indirectly by first reducing or thermally decomposing the gas quantitatively
to NO with a converter. The reaction of NO and 0_ forms excited NO- molecules
O (L
that release light energy that is proportional to the NO concentration. Al-
though the NO chemiluminescence is interference-free, other nitrogen compounds
do interfere when directed through the NO- converter. The magnitude of these
interferences is dependent upon the type of converter used (Winer et al. ,
1974; Joshi and Bufalini, 1976). Other NO and NO- measuring methods have also
been summarized in this chapter. None of the other techniques is widely used to
monitor air quality.
3.6.3 Sources and Emissions of Precursors
The photochemical production of ozone, the principal component of "smog,"
depends both on the presence of precursors, volatile organic compounds (VOCs)
and nitrogen oxides (NO ), that are emitted by manmade and by natural sources,
J\
and on suitable conditions of sunlight, temperature, and other meteorological
factors. Because of the intervening requirement for meteorological conditions
conducive to the photochemical generation of ozone, emission inventories are
not as direct predictors of ambient concentrations in the case of secondary
pollutants such as ozone and other oxidants as they are for primary pollutants.
Emissions of manmade VOCs (excluding several relatively unreactive com-
pounds such as methane) in the United States have been estimated at 18.2 tg/yr
for 1982. Trends in manmade VOC emissions for 1970 through 1982 were shown in
Figure 3-3 (U.S. Environmental Protection Agency, 1983). The annual emission
rate for manmade VOCs has decreased some 28 percent during this period. The
main sources nationwide are industrial processes, which emit a wide variety of
VOCs such as chemical solvents; and transportation; which includes the emission
of VOCs in gasoline vapor as well as in gasoline combustion products. Estimates
of biogenic emissions of organic compounds in the United States are highly
inferential but data suggest that the yearly rate is the same order of magni-
tude as manmade emissions. Most of the biogenic emissions actually occur
during the growing season, however, and the kinds of compounds emitted are
different.
OZONER/D 3-99 6/28/84
-------
Emissions of manmade NO in the United States estimated at 20.2 tg/yr for
1982. Annual emissions of manmade NO were some 12 percent higher in 1982
than in 1970, but the rate leveled off in the late 1970s and exhibited a small
decline from about 1980 through 1982. The increase over the period 1970
through 1982 had two main causes: (1) increased fuel combustion in stationary
sources such as power plants; and (2) increased fuel combustion in highway
motor vehicles, as the result of the increase in vehicle miles driven. Total
vehicle miles driven increased by 42 percent over the 13 years in question.
Trends in manmade NO emissions over 1970 through 1982 were shown in Figure
/\
3-5 (U.S. Environmental Protection Agency, 1983). Estimated biogenic N0x
emissions are based on uncertain extrapolations from very limited studies, but
appear to be about an order of magnitude less than manmade emissions.
3.6.4 Ambient Air Concentrations of Precursors
3.6.4.1 Hydrocarbons in Urban Areas. Most of the available ambient air data
on the concentrations of nonmethane hydrocarbons (NMHC) in urban areas have
been obtained during the 6:00 to 9:00 a.m. period. Since hydrocarbon emissions
are at their peak during that period of the day, and since atmospheric disper-
sion is limited that early in the morning, NMHC concentrations measured then
generally reflect maximum diurnal levels. Representative data for urban areas
show mean NMHC concentrations between 0.4 and 0.9 ppm.
The hydrocarbon composition of urban atmospheres is dominated by species
in the C? to C,fi molecular-weight range. The paraffinic hydrocarbons (alkanes)
are most prominent, followed by aromatics and alkenes. Based on speciation
data obtained in a number of urban areas, alkanes generally constitute 50 to
60 percent of the hydrocarbon burden in ambient air, aromatics 20 to 30 percent,
with alkenes and acetylene making up the remaining 5 to 15 percent (Sexton and
Westberg, 1984).
3.6.4.2 Hydrocarbons in Nonurban Areas. Rural nonmethane hydrocarbon concentra-
tions are usually one to two order of magnitude lower than those measured in
urban areas (Ferman, 1981; Sexton and Westberg, 1984). In samples from sites
carefully selected to guarantee their rural character, total NMHC concentra-
tions ranged from 0.006 to 0.150 ppm C (e.g., Cronn, 1982; Seila, 1981; Holdren
et al., 1979). Concentrations of individual species seldom exceeded 0.010 ppm
OZONER/D 3-100 6/28/84
-------
C. The bulk of species present in rural areas are alkanes; ethane, propane,
n-butane, j_so-pentane, and ri-pentane are most abundant. Ethylene and propene
are sometimes present at <0.001 ppm C, and toluene is usually present at
~0.001 ppm C. Monoterpene concentrations are usually about <0.020 ppm C.
During the summer months, isoprene concentrations as high as 0.150 ppm C have
been measured (Ferman, 1981). The maximum concentrations of isoprene usually
encountered, however, are in the range of 0.030 to 0.040 ppm C.
3.6.4.3 Aldehydes in Urban Areas. Aldehydes observed in urban atmospheres
include formaldehydes, acetaldehyde, chloral, propanal, n-butanal, and benz-
aldehyde. Formaldehyde concentrations are the best characterized of these
aldehydes because the chromotropic acid methodology for formaldehyde was
established in the early 1960s. With the exception of early data from Los
Angeles (1961), reported concentrations of formaldehyde in urban areas fall in
the 0.01 to 0.03 ppm range, with maximum concentrations ranging up to 0.09
ppm.
Comparing these concentrations with concentrations of NMHC in urban
areas, it is apparent that formaldehyde probably constitutes less than 3
percent of the total NMOC in most urban areas. Acetaldehyde concentrations
are generally lower than formaldehyde in a given urban area. Concentrations
of total aldehydes in urban atmospheres can vary from a few ppb up to about
0.2 ppm (200 ppb). In polluted atmospheres, acrolein, propanal, butanal, and
benzaldehyde have each been measured at concentrations <0.015 ppm.
3.6.4.4 Aldehydes in Nonurban Areas. Very few total aldehyde measurements have
been made in rural areas. Breeding et al. (1973) reported values for total
aldehydes of 0.001 to 0.002 ppm in rural Illinois and Missouri. Formaldehyde
levels in remote atmospheres apparently range from 0.1 to 10 ppb, with global
background formaldehyde concentrations varying from 0.3 to 0.5 ppb (Duce et
al., 1983).
3.6.4.5 Nitrogen Oxides in Urban Areas. Concentrations of NO , like hydrocar-
"• •""' ^ ' " ""- ^•.,WH>^^-^ J^
bon concentrations, tend to peak in urban areas during the early morning, when
atmospheric dispersion is limited and automobile traffic is dense. Most NO^
is emitted as nitric oxide (NO), but the NO is rapidly converted to N02 by
ozone and peroxy radicals produced in atmospheric photochemical reactions.
The relative concentrations of NO versus N02 fluctuate day-to-day, depending
on diurnal and day-to-day fluctuations in ozone levels and photochemical
activity.
OZONER/D 3-101 6/28/84
-------
Urban NO concentrations during the 6:00 to 9:00 a.m. period in 10 cities
J\
ranged from 0.05 to 0.15 ppm in studies done in the last 5 to 7 years (e.g.,
Westberg and Lamb, 1983; Richter, 1983; Eaton et al., 1979). Concurrent NMHC
measurements for these 10 cities showed that NMHC/NO ratios ranged from 5 to
16.
3.7 REFERENCES
Allwine, G.; Lamb, B.; Westberg, H. (1983) Application of atmospheric tracer
techniques for determining biogenic hydrocarbon fluxes from an oak forest.
In: Proceedings of Forest Environmental Measurements Conference; October;
Oak Ridge, TN.
Altshuller, A. P. (1983a) Measurements of the products of atmospheric photo-
chemical reactions in laboratory studies and in ambient air—relationships
between ozone and other products. Atmos. Environ. 17:2383-2427.
Altshuller, A. P. (1983b) Review: Natural volatile organic substances and
their effect on air quality in the United States. Atmos. Environ.
17:2131-2165.
Altshuller, A. P.; Cohen, I. R.; Meyer, M. E.; Wartburg, A. F., Jr. (1961)
Analysis of aliphatic aldehydes in source effluents and in the atmosphere.
Anal. Chim. Acta 25:101-117.
Altshuller, A. P.; Leng, L. J. (1963) Application of 3-methyl-2-benzothiazolone
hydrazone method for atmospheric analysis of aliphatic aldehydes. Anal.
Chem. 35:1541-1542.
Altshuller, A. P.; McPherson, S. P. (1963) Spectrophotometric analysis of alde-
hydes in the Los Angeles atmosphere. J. Air Pollut. Control Assoc. 13:109-
111.
American Public Health Association, Intersociety Committee. (1977) Tentative
method of analysis for formaldehyde content of the atmosphere (MBTH-
Colorimetric Method—Application to Other Aldehydes). In: Katz, M. , ed.
Methods of Air Sampling and Analysis, 2nd Ed. Washington, DC: American
Public Health Association.
American Public Health Association, Intersociety Committee. (1977) Tentative
method of analysis for nitrogen dioxide content of the atmosphere. In:
Katz, M., ed. Methods of Air Sampling and Analysis, 2nd Ed. Washington,
DC: American Public Health Association.
Arnts, R. R.; Gay, B. W. (1979) Photochemistry of some naturally emitted hydro-
carbons. Research Triangle Park, NC: U.S. Environmental Protection Agency;
EPA report no. EPA-600/3-79-081.
OZONER/D 3-102 6/28/84
-------
Arnts, R. R.; Seila, R. L.; Kuntz, R. L.; Mowry, F. L.; Knoerr, K. R.; Dudgeon,
A. C. (1978) Measurement of alpha-pinene fluxes from a loblolly pine
forest. In: Proceedings of the 4th Joint Conference on Sensing of En-
vironmental Pollutants, American Chemical Society. Washington, DC: American
Chemical Society.
Ashby, H. A.; Stahman, R. C. ; Eccleston, B. H. ; Hum, R. W. (1974) Vehicle
emissions—summer to winter. In: Society of Automotive Engineers meeting;
October; Toronto, Canada. SAE technical paper 741053.
Atkinson, R.; Darnall, K. R. ; Lloyd, A. C.; Winer, A. M. ; Pitts, J. N. , Jr.
(1979) Kinetics and mechanisms of the reactions of the hydroxyl radical with
organic compounds in the gas phase. In: Pitts, J. N., Jr.; Hammond, G. S.;
Gollnick, K. ; Grosjean, D. , eds. Advances in Photochemistry, Vol. 11.
New York, NY: John Wiley & Sons.
Atkinson, R., et al. (1980) A smog chamber and modeling study of the gas phase
NO -air photooxidation of toluene and cresols. Int. J. Chem. Kinet.
12: 779-837.
Atkinson, R.; Aschmann, S. M.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr.
(1982) Kinetics of the reactions of OH radicals with n-alkanes at 299 ± K.
Int. J. Chem. Kinet. 14: 781-788.
Automotive News, 1982 Market Data Book Issue. (1982a) Sales of diesel-powered
cars in the U.S. Detroit, MI: Crain Communications, Inc.; April 28;
p. 74.
Automotive News, 1982 Market Data Book Issue. (1982b) Automotive news analysis
of new-car registrations. Detroit, MI: Crain Communications, Inc.; April
28; p. 8.
Automotive News, 1983 Market Data Book Issue. (1983a) Sales of new diesel-
powered cars in U.S. Detroit, MI: Crain Communications, Inc; April 27;
p. 70.
Automotive News, 1983 Market Data Book Issue. (1983b) Domestic car sales.
Detroit, MI: Crain Communications, Inc.
Automotive News. (1983c) Diesel-car sales plunge 48 pet. Detroit, MI: Crain
Communications, Inc.; May 2; p. 70.
Bartok, W., et al. (1971) Systematic field study of NO emission control methods
for utility boilers. U.S. Environmental Protection Agency, Research Triangle
Park, NC. Available from NTIS, Springfield, VA. NTIS PB 210-739. Cited
in Hall and Bowen (1982).
Beard, M. E. ; Margeson, J. H. (1974) An evaluation of arsenite procedure for
determination of nitrogen dioxide in ambient air. Research Triangle Park,
NC: U.S. Environmental Protection Agency; EPA report no. EPA-650/4-74-048.
Beard, M.E.; Suggs, J. C.; Margeson, J. H. (1975) Evaluation of effects of NO,
C02, and sampling flow rate on arsenite procedure for measurement of N02 in
ambient air. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no. EPA-650/4-75-018.
OZONER/D 3-103 6/28/84
-------
Black, F. M. (1977) The impact of emissions control technology on passenger
car hydrocarbon emission rates and patterns. In: Proceedings of the
International Conference on Photochemical Oxidant Pollution and Its Con-
trol. Research Triangle Park, NC: U.S. Environmental Protection Agency;
pp. 1053-1068; EPA report no. EPA-600/3-77-001b.
Black, F. M.; Bradow, R. L. (1975) Patterns of hydrocarbon emissions from 1975
production cars. Presented at Society of Automotive Engineers' National
Fuels and Lubricants Meeting; Houston, TX. SAE paper no. 750681.
Black, F. M.; High, L. E. (1979) Diesel hydrocarbon emissions, particulate and
gas phase. Presented at Society of Automotive Engineers' Symposium on
Diesel Particulate Emissions Measurement Characterization; February;
Detroit, MI.
Black, F. M.; High, L. E. ; Lang, J. M. (1980) Composition of automotive evapo-
rative and tailpipe hydrocarbon emissions. J. Air Pollut. Control Assoc.
30:1216-1221.
Bollinger, M. J. ; Parrish, D. D. ; Hahn, C.; Albritton, D. L.; Fehsenfeld, F.
C. (1982) NO measurements in clean continental air. In: Proceedings of
2nd Symposium on Composition of the Nonurban Troposphere; May; Williamsburg,
VA.
Box, E. 0. (1981) Foliar biomass: data base of the International Biological
Program and other sources. In: Bufalini, J. J. ; Arnts, R. R. , eds.
Atmospheric Biogenic Hydrocarbons. Vol. 1. Ann Arbor, MI: Ann Arbor
Science; pp. 107-119.
Braddock, J. N.; Gabele, P. A. (1977) Emission patterns of diesel-powered
passenger cars. Part II. Presented at Society of Automotive Engineers'
Automotive Engineering Congress and Exposition; February-March; Detroit, MI.
SAE paper no. 770168.
Breeding, R. J. ; Lodge, J. P., Jr.; Pajte, J. B. ; Sheesley, D. C. ; Klonis, H.
B. ; Fogle, B.; Anderson, J. A.; Englert, T. R.; Haagenson, P. L. ; McBeth,
R. B. ; Morris, A. L.; Pogue, R.; Wartburg, A. F. (1973) Background trace
gas concentration in the central United States. J. Geophys. Res. 78:
7057-7064.
Breitenbach, L. P.; Shelef, M. (1982) Development of a method for the analysis
of N02 and NH3 by NO-measuring instrument. J. Air Pollut. Control Assoc.
23:128-131.
Bricker, C. E.; Johnson, H. R. (1945) Spectrophotometric method for determining
formaldehyde. Ind. Eng. Chem., Anal. Ed. 17:400-402. Cited in: National
Research Council. (1981) Formaldehyde and other aldehydes. Washington, DC:
National Academy Press.
Bucon, H. W. ; Macko, J. F. ; Taback, H. J. (1978) Volatile organic compound
(VOC) species data manual. Prepared by KVB Engineering, Inc., Tustin,
CA. Research Triangle Park, NC: U.S. Environmental Protection Agency;
EPA report no. EPA-450/3-78-119.
OZONER/D
3-104
6/28/84
-------
Butcher, S. S.; Ruff, R. E. (1971) Effect of inlet resistance time on analysis of
atmospheric nitrogen oxides and ozone. Anal. Chem. 43:1890-1892.
California Board of Air Sanitation. (1966) The oxides of nitrogen in air
pollution. Sacramento, CA: California Department of Public Health.
Campau, R. M.; Neerman, J. C. (1966) Continuous mass spectrometric determina-
tion of nitric oxide in automotive exhaust. Presented at Automotive En-
gineering Congress; January; Detroit, MI. SAE paper no. 660116.
Cato, G. A., et al. (1974) Field testing: application of combustion modifica-
tions to control pollutant emissions from industrial boilers-phase I, U.S.
Environmental Protection Agency, EPA-650/2-74-078a, NTIS PB 238-920. Cited
in Hall and Bowen (1982).
Cato, G. A.; Muzio, L. J. ; Shore, D, E. (1976) Field testing: application of
combustion modifications to control pollution emissions from industrial
boilers—phase II. Research Triangle Park, NC: U.S. Environmental Pro-
tection Agency; EPA report no. EPA-600/2-76-086a.
Chase, J. 0.; Hum, R. W. (1970) Measuring gaseous emissions from an aircraft
turbine engine. Presented at National Air Transportation Meeting; April;
New York, NY.
Chatfield, R. B.; Rasmussen, R. A. (1977) Hydrocarbons, fluorocarbons, and
oxidants observed at a rural site near St. Louis. In: Proceedings from
National AIChE Meeting; March; Houston, TX.
Constant, P. C., Jr.; Sharp, M. C.; Scheil, G. W. (1974) Collaborative test of
the TGS-ANSA method for measurement of nitrogen dioxide in ambient air.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-650/4-74-046.
Constant, P. C., Jr.; Sharp, M. C.; Scheil, G. W. (1975) Collaborative testing
of methods for measurements of N02 in ambient air. Vol. I. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-650/4-74-019a.
Corse, E. W. (1981) Aldehyde analyses for the 1981 Atlanta study. Report to
U.S. Environmental Protection Agency, Research Triangle Park, NC, under
contract 68-02-2566, September 1981.
Cox, R. D.; Balfour, W. D.; Langley, G. J. (1982). Quality control for ambient
level hydrocarbon sampling and analyses. Presented at 75th Annual Meeting
of the Air Pollution Control Association; June; New Orleans, LA.
Crabtree, J. H. ; Mayrsohn, H. (1977) Urban and industrial air pollution in
Houston, Texas—I. Hydrocarbons (Discussion). Atmos. Environ. 12:1245.
Crawford, A. R. , et al. (1974) Field testing: application of combusion modifi-
cations to control NO emissions from utility boilers. U.S. Environmental
Protection Agency, EPA-650/2-74-066, NTIS PB 237-344. Cited in Hall and
Bowen (1982).
OZONER/D
3-105
6/28/84
-------
Crawford, A. R., et al. (1978) Control of utility boiler and gas turbine pollu-
tant emissions by combustion modification-phase I. U.S. Environmental Pro-
tection Agency, EPA-600/7-78-036a, NTIS PB 281-078. Cited in Hall and
Bowen (1982).
Cronn, D. R. (1982) Monitoring of ambient halocarbons and natural hydrocarbons
in the Smoky Mountains. In: Proceedings from 2nd Symposium on Composi-
tion of the Nonurban Troposphere; May; Williamsburg, VA.
Decker, C. E. ; Ripperton, L. A.; Worth, J. J. B.; Vukovich, F. M. ; Bach, W.
D. ; Tommerdahl, J. B. ; Smith, F.; Wagoner, D. E. (1978) Study of the
formation and transport of ambient oxidants in the western Gulf Coast and
north-central and northeast regions of the United States. Research
Triangle Park, MC: Research Triangle Institute.
Dement, W. A.; Tyson, B. J. ; Mooney, H. A. (1975) Mechanism of monoterpene
volatilization in Sal via mellifera. Phytochemistry 14:2555.
Dietz, W. A. (1967) Response factors for gas chromatographic analyses. J. Gas
Chromatogr. 5:68-71.
Dietzman, H. E.; Parness, M. A.; Bradow, R. L. (1980) Emissions from trucks by
chassis version 1983 transient procedures. Presented at Society of Auto-
motive Engineers' Meeting. SAE Paper No. 801371.
Dietzman, H. E. ; Parness, M. A.; Bradow, R. L. (1981) Emissions from gasoline
and diesel delivery trucks by chassis transient cycle. ASME Energy
Sources Technology Conference and Exhibition. January; Houston, TX.
Dimitriades, B. (1974) Proceedings of the Solvent Reactivity Conference.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA report
no. EPA-650/3-74-010.
Dimitriades, B. (1981) The role of natural organics in photochemical air
pollution: Issues and research needs. J. Air Pollut. Control Assoc.
31:229-235.
Duce, R. A.; Mohnen, V. A.; Zimmerman, P. R.; Grosjean, D.; Cautreels, W. ;
Chatfield, R. B.; Jaenicke, R.; Ogren, J. A.; Pellizzari, E. D. ; Wallace, G.
T. (1983) Organic material in the global troposphere. Rev. Geophys.
Space Phys. 21: 921-952.
Eaton, W. C.; Decker, C. E.; Tommerdahl, J. B.; Dimmock, F. E. (1979) Study of
the nature of ozone, oxides of nitrogen, and nonmethane hydrocarbons in
Tulsa, Oklahoma. Research Triangle Park, NC: Environmental Protection
Agency; EPA report no. EPA-450/4-79-008a.
Eegrive, E. (1937) Reactions and reagents for the detection of organic com-
pounds. Z. Anal. Chem. 110:220-225. Cited in: National Research Council.
(1981) Formaldehyde and Other Aldehydes. Washington, DC: National Academy
Press. 1981.
Evans, R. C. ; Tingey, D. T.; Gumpertz, M. L.; Burns, W. F. (1982) Estimates of
isoprene and monoterpene emission rates in plants. Bot. Gaz. (Chicago)
143:304.
OZONER/D 3-106 6/28/84
-------
Federal Highway Administration. (1978) Traffic Volume Trends, Tables 5A, 5B,
and 9A. November-December (1978) Computer printout available from Federal
Highway Administration, Highway Statistics Division, 400 7th St., SW,
Washington, D.C., 20590.
Ferman, M. A. (1981) Rural nonmethane hydrocarbon concentrations and composition.
In: Bufalini, J. J.; Arnts, R. R., eds. Atmospheric Biogenic Hydrocarbons.
Ann Arbor, MI: Ann Arbor Science Publishers, Inc.
Fincher, C. L.; Tucker, A. W.; Birnbaum, M.; Paur, R. J.; McClenny, W. A.
(1977) Fluorescence ambient N02 monitor with flashlamp pumping. Appl.
Optics 16:1359.
Fisher, T. M. (1980) The evolution of automotive emission controls. Presented
at the 73rd annual meeting of the Air Pollution Control Association;
June; Montreal, Canada. Pittsburgh, PA: Air Pollution control Association;
preprint no. 80-1.1.
Flyckt, D. L.; Westberg, H. H.; Holdren, M. W. (1980) Natural organic emissions
and their impact on air quality. Presented at 73rd Annual Air Pollution
Control Association meeting, June; Montreal, Quebec. Paper 80-69.2.
Fontijn, A.; Sabadell, A. J.; Ronco, R. J. (1970) Homogeneous chemiluminescent
measurement of nitric oxide with ozone: Implications for continuous
selective monitoring of gaseous air pollutants. Anal. Chem. 42:575-579.
Fuerst, R. G. ; Margeson, J. H. (1974) An evaluation of TGS-ANSA procedure for
determination of nitrogen dioxide in ambient air. Research Triangle
Park, NC. U.S. Environmental Protection Agency; EPA report no.
EPA-650/4-74-047.
Fung, K.; Grosjean, D. (1981) Determination of nanogram amounts of carbonyls
as 2,4-dinitrophenylhydrazones by high-performance liquid chromatography.
Anal. Chem. 53:168-171.
George, R. E.; Nevitt, J. S.; Verssen, J. A. (1972) Jet aircraft operations:
Impacts on the air environment. J. Air Pollut. Control Assoc. 22:
507-515.
Gerstle, R. W. ; Peterson, R. F. (1966) Atmospheric emissions from nitric acid
manufacturing processes. Cincinnati, Ohio: National Center for Air
Pollution Control; Public Health Service publication no. 999-AP-27.
Gibbs, R. E.; Hyde, J. D.; Whitby, R. A.; Choudhury, D. R. (1983) Charac-
terization of emissions and fuel economy of in-use diesel automobiles.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-600/3-83-087.
Graedel, T. E. (1978) Chemical Compounds in the Atmosphere. New York, NY:
Academic Press.
Grosjean, D.; Fung, K. (1984) Hydrocarbons and carbonyls in Los Angeles Air.
J. Air Pollut. Control Assoc. 34: 537-543.
OZONER/D 3-107 6/28/84
-------
Grosjean, D. (1982) Distribution of nitrogenous pollutants during photochemi-
cal smog episodes. Presented at 75th annual meeting of the Air Pollution
Control Association; June; New Orleans, LA.
Grosjean, D.; Kok, G. L. (1981) Interlaboratory comparison of methods for
measuring formaldehyde and other aldehydes in ambient air. Final Report;
Prepared for Coordinating Research Council, Atlanta, GA.
Guicherit, R. (1972) Indirect determination of nitrogen oxides by chemilumine-
scent technique. Atmos. Environ. 6:807-814.
Hagar, R. N. , Jr.; Anderson, R. C. (1970) Theory of the derivative spectro-
meter. J. Opt Soc. Am. 50:1444-1449.
Hall, R. E. ; Bowen, J. S. (1982) State-of-the-art combustion modification NO
control for stationary combustion equipment. In: Schneider, T.; Grant,x
L. ; eds. Air Pollution by Nitrogen Oxides. Amsterdam, The Netherlands:
Elsevier Scientific Publishing Co.
Hanst, P. L. (1970) Infrared spectroscopy and infrared lasers in air pollution
research and monitoring. Appl. Spectrosc. 24:161-174.
Hanst, P. L.; Wong, N. W.; Bragin, J. (1982) A long-path infrared study of Los
Angeles smog. Atmos. Environ. 16:969-981.
Hanst, P. L. ; Wilson, W. E.; Patterson, R. K.; Gay, G. W., Jr.; Chaney, L. W.
(1975) A spectroscopic study of California smog. Research Triangle Park,
NC: U.S. Environmental Protection Agency; EPA report no. EPA-650/4-75-006.
Harrison, J. W.; Timmons, M. L.; Denyszyn, R. B.; Decker, C. E. (1977) Evalua-
tion of the EPA Reference Method for measurement of non-methane hydrocar-
bons. Research Triangle Park, NC: U.S. Environmental Protection Agency;
EPA report no. EPA-600/4-77-033.
Hersch, P.; Deuringer, R. (1963) Galvanic monitoring of ozone in air. Anal.
Chem. 35:897.
Hilliard, J. C. ; Wheeler, R. W. (1979) Nitrogen dioxide in engine exhaust.
Presented at Passenger Car meeting; June; Dearborn, MI. SAE Paper No.
790691.
Hilliard, J. C. ; Wheeler, R. W. (1974) Nitrogen dioxide in engine exhaust.
Paper no. DP 18471, Ricardo Consulting Engineers, Sussex, England.
Hinkley, E. D. ; Kelley, P. L. (1971) Detection of air pollutants with tunable
diode lasers. Science (Washington, DC) 171:635-639.
Holdren, M. W.; Westberg, H. H.; Zimmerman, P. R. (1979) Analysis of monoter-
pene hydrocarbons in rural atmospheres. JGR J. Geophys. Res. 84: 5083-
5088.
Holdren, M.; Spicer, C.; Sticksel, P.; Nepsund, K.; Ward, G.; Smith, R. (1982)
Implementation and analysis of HC grab samples from Cleveland and Cincin-
nati, Ozone Monitoring Study. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, EPA report no. EPA-905/4-82-001.
OZONER/D 3-108 6/28/84
-------
Hughes, E. E. (1975) Development of standard reference materials for air
quality measurement. ISA Trans. 14:281-289.
Hunsaker, D. (1981) Selection of biogenic hydrocarbon emission factors for
land cover classes found in the San Francisco Bay Area. Air Quality
Tech. Memo 31, Association of Bay Area Governments, California. January.
Hunter, S. C., et al. (1978) Evaluation of combustion modifications for emissions
reduction on industrial processes. ASME Paper No. 78-WA/APO8. Cited in
Hall and Bowen (1982).
Hunter, S. C. ; Carter, W. A.; McCroy, M. W. ; Cherry, S. S. ; Buening, H. J.
Application of combustion modifications to industrial combustion equipment.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA~600/7-79-015a.
Hunter, S. C., et al. (1979) Application of advanced combustion modification to
industrial process equipment-process heater subscale tests. In Proceedings
of the Third Stationary Source Combustion Symposium, Vol. Ill, U.S. Envi-
ronmental Protection Agency, EPA-600/7-79-050c, NTIS PB 292-541. Cited
in Hall and Bowen (1982).
Huygen, I. C. (1970) Reaction of nitrogen dioxide with Griess type reagents.
Anal. Chem. 42:407-409.
Jayanty, R. K. M. ; Blackard, A.; McElroy, F. F. ; McClenny, W. A. (1982)
Laboratory evaluation of nonmethane organic carbon determination in
ambient air by cryogenic preconcentration and flame ionization detection.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-600/4-82-019.
Johnson, D. F. ; Kok, G. L.; Sonner, R. J. (1981) Improved chromatographic acid
technique for the determination of formaldehyde. In: Grosjean, D; Kok,
G. L. , eds. Interlaboratory comparison study of methods for measuring
formaldehyde and other aldehydes in ambient air. Final report prepared
for Coordinating Research Council, Atlanta, GA.
Jordan, B. C. ; Broderick, A. J. (1979) Emissions of oxides of nitrogen from
aircraft. J. Air. Pollut. Control Assoc. 29:119-124.
Joshi, S. B. ; Adger, R. A.; Barnes, D. S.; Kaiserman, M. (1981) Aldehyde
sampling field study. Northrop Services, Inc. Environ. Sci. Tech. Report
81-02, Research Triangle Park, NC.
Joshi, S. B. (1979) Houston field study-1978: formaldehyde and total aldehydes
monitoring study. Northrop Services, Inc. Report to the U.S. Environmen-
tal Protection Agency. Contract no. 68-02-2566. Research Triangle Park,
NC.
Joshi, S. B.; Bufalini, J. J. (1978) Halocarbon interferences in chemilumines-
cent measurements of NO . Environ. Sci. Technol. 12:597-599.
}\
JSA, Inc. (Jones and Stokes Associates, Inc.) (1978) Draft environmental
impact statement for the South Tahoe Public Utilities District, Waste
Water Facilities Planning Program.
OZONER/D 3-109 6/28/84
-------
Katz, M. Nitrogen compounds and oxidants. (1976) In: Stern, A. C. , ed. Air
Pollution, vol. III. New York, NY: Academic Press, Inc.
Kelly, N. A.; Wolff, G. T.; Ferman, M. A. (1982) Background pollutant measure-
ments in air masses affecting the eastern half of the United States - I.
Air masses arriving from the northwest. Atmos. Environ. 16: 1077-1088.
Killus, J. P.; Whitten, G. Z. (1982) A mechanism describing the photochemical
oxidation of toluene in smog. Atmos. Environ. 16: 1973-1988.
Knoerr, K. R. ; Howry, F. W. (1981) Energy balance/Bowen ratio technique for
estimating hydrocarbon fluxes. In: Bufalini, J. J.; Arnts, R. R., eds.
Atmospheric Biogenic Hydrocarbons. Vol. 1. Ann Arbor, MI: Ann Arbor
Science.
Knoppel, H. ; Versino, B. ; Peil, A.; Schauenburg, H.; Vissers, H. (1982) Quanti-
tative determination of terpenes emitted by conifers. In: Versino, D.;
Ott, H., eds. Physicochemical Behavior of Atmospheric Pollutants. Boston,
MA: D. Reidel Publishing Company; pp. 89-98.
Kopczynski, S. L. ; Lonneman, W. A.; Winfield, T. ; Seila, R. (1975) Gaseous
pollutants in St. Louis and other cities. J. Air Pollut. Control Assoc.
25:251-255.
Kreuzer, L. G. ; Patel, C. K. N. (1971) Nitric oxide air pollution: Detection
by optoacoustic spectroscopy. Science (Washington, DC) 173:45-47.
Krug, E. L. R.; Hirt, W. E. (1977) Interference of nitrate in the determination
of formaldehyde by the chromotropic acid method. Anal. Chem. 49:1865-1867.
Cited in: National Academy of Sciences. (1981). Formaldehyde and Other
Aldehydes. Washington, DC: National Academy Press.
Kuntz, R. ; Lonneman, W.; Namie, G.; Hull, L. A. (1980) Rapid determination of
aldehydes in air analyses. Anal. Lett. 13:1409-1415.
Lamb, B. K. ; Westberg, H. H.; Quarles, T.; Flyckt, D. L. (1983) Natural
hydrocarbon emission rate measurements from selected forest sites.
Research Triangle Park, NC: U.S. Environmental Protection Agency, EPA
report no. EPA-600/3-84-001.
Lamb, S. I.; Petrowski, C.; Kaplan, I. R.; Simoneit, B. R. T. (1980) Organic
compounds in urban atmospheres: A review of distribution, collection and
analysis. J. Air Pollut. Control Assoc. 30:1098-1115.
Leihkanen, H. E. ; Beckman, E. W. (1971) The effect of leaded and unleaded
gasolines on exhaust emissions as influenced by combustion chamber de-
posits. Presented at Society of Automotive Engineers' National Combined
Fuels and Lubricants, Powerplant, and Truck Meetings, October; St. Louis,
MO. SAE paper no. 71083.
Levaggi, D. A.; Kothny, E. L.; Bel sky, T.; deVera, E. R.; Mueller, P. K.
(1974) Quantitative analysis of nitric oxide in presence of nitrogen
dioxide at atmospheric concentrations. Environ. Sci. Techno!. 8:348-350.
OZONER/D 3-110 6/28/84
-------
Linnell, R. H.; Scott, W. E. (1962) Diesel exhaust composition and odor studies.
J. Air Pollut. Control Assoc. 12:510,515,545.
Lipari, F.; Swarin, S. (1982) Determination of formaldehyde and other aldehydes
in automobile exhaust with an improved 2,4-dinitrophenylhydrazine method.
J. Chromatogr. 247:297-306.
Lloyd, A. C. (1979) Chemical kinetic data needs for modeling of the lower tro-
posphere. Gaithersburg, MD: U.S. Department of Commerce, National
Bureau of Standards; special publication 557.
Logan, J. A. (1983) Nitrogen oxides in the troposphere: global and regional
budgets. JGR J. Geophys. Res. 88: 10785-10807.
Lonneman, W. A. (1977) Ozone and hydrocarbon measurements in recent oxidant
transport studies. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/3-77-001a.
Lonneman, W. A. (1979) Ambient hydrocarbon measurement in Houston. In: Pro-
ceedings of Ozone/Oxidants: Interactions with the Total Environment;
October; Houston, TX. Pittsburgh, PA: Air Pollution Control Associa-
tion.
Lonneman, W. A.; Bufalini, J. J. (1978) Discussion of Siddiqi and Worley
(1977). Atmos. Environ. 12:963-964.
Lonneman, W. A.; Bufalini, J. J. ; Kuntz, R. L. ; Meeks, S. A. (1981) Contamina-
tion from fluorocarbon films. Environ. Sci. Techno!. 15:99-103.
Lonneman, W. A.; Kopczynski, S. L. ; Darley, P.E.; Sutterfield, F. D. (1974)
Hydrocarbon composition of urban air pollution. Environ. Sci. Techno!.
8:229-236.
Lord, H. C. ; Egan, W. D. ; Johnson, F. L.; Mclntosh, L. C. (1975) Measurement
of exhaust emission in piston and diesel engines by dispersive spectrum.
J. Air Pollut. Control Assoc. 25:940-942.
Los Angeles County, California. (1971) Profile of Air Pollution Control. Los
Angeles Air Pollution Control District.
Lozano, E. R. ; Melvin, W. W. ; Hochheiser, S. (1968) Air pollution emissions
from jet engines. J. Air Pollut. Control Assoc. 18:392-394.
Maeda et al. (Citation to be supplied).
Manos, D. M. ; Volltrauer, H. N.; Allen, J.; Burch, D. E.; Chudy, J. S.; Pembrook,
J. D. (1982) Totally optical ambient NMHC monitor. Report under Contract
No. 68-02-3292, U.S. Environmental Protection Agency, Research Triangle
Park, November.
Marchesani, V. J.; Towers, T. ; Wohlers, H. C. (1970) Minor sources of air
pollutant emission. J. Air Pollut. Control Assoc. 20:19.
Mayrsohn, H.; Crabtree, J. H. (1975) Source reconciliation of atmospheric
hydrocarbons. Atmos. Environ. 10:137-143.
OZONER/D 3-111 6/28/84
-------
Mayrsohn, H. ; Crabtree, J. H.,; Kuramoto, M. ; Sothern, R. D. ; Mano, S. H.
(1977) Source reconciliation of atmospheric hydrocarbons, 1974. Atmos.
Environ. 11:189-192.
McBride, J. M.; McClenny, W. A. (1980) Analysis of NMOC by cryogenic preconcen-
tration and flame ionization detection. Draft Report. U. S. Environmen-
tal Protection Agency, Research Triangle Park, NC.
McElroy, C. T. ; Kerr, J. B. (1977) A limb scan measurement of atmospheric
nitrogen dioxide. Presented at fall meeting of the American Geophysical
Union; San Francisco, CA.
McElroy, F. F. ; Thompson, V. L. (1975) Hydrocarbon measurement discrepancies
among various analyzers using flame-ionization detectors. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/4-75-010.
McKee, H. C.; Clark, J. M.; Wheeler, R. J. (1962) Measurement of diesel engine
emissions. J. Air Pollut. Control Assoc. 12:516-521, 546.
Miksch, P. R.; Anthon, D. W. (1982) Recommendation for combining the standard
analytical methods for the determinations of formaldehyde and total
aldehydes in air. Am. Ind. Hyg. Assoc. J. 43:362-365.
Mueller, P. K.; Hidy, G. M. (1983) The sulfate regional experiment (SURE):
report of findings. Palo Alto, CA: Electric Power Research Institute;
report no. EA-1901.
Mulik, J. D. ; Fuerst, R. G. ; Meeker, J. R.; Guyer, M.; Sawicki, E. (1973) A
new 24-hour manual method of collection for colorimetric analysis of
atmospheric N02. Presented at the 165th National Meeting of the American
Chemical Society; April; Dallas, TX.
Nash, T. (1970) An efficient absorbing reagent for nitrogen dioxide (1970)
Atmos. Environ. 4:661-665.
National Academy of Sciences, National Research Council. (1981) Formaldehyde
and Other Aldehydes. Washington, D.C.: National Academy of Sciences.
National Academy of Sciences, National Research Council. (1977) Nitrogen
Oxides. Washington, DC: National Academy of Sciences.
National Academy of Sciences, National Research Council. (1976) Vapor-Phase
Organic Pollutants. Washington, DC: National Academy of Sciences.
Naylor, M. H.; Lee, D. C.; Smith, M. J. (1981) Southeast Las Vegas valley air
quality measurement project 1980-81. Final Report. Las Vegas, Nevada:
Clark County Health District.
Neligan, R. E. (1962) Hydrocarbons in the Los Angeles atmosphere. Arch.
Environ. Health 5:581.
Niki, H. ; Maker, P. D. ; Savage, C. M.; Breitenbach, L. P. (1983) Atmospheric
ozone-olefin reactions. Environ. Sci. Techno!. 17: 312A-322A.
OZONER/D 3-112 6/28/84
-------
Norwitz, G. A. (1966) Colorimetric method for the determination of oxides of
nitrogen. Analyst 91:553-558.
Ogle, L. D.; Hall, R. C. ; Crow, W. L. ; Jones, A. E. (1982) Development of
preconcentration and chromatographic procedures for the continuous and
unattended monitoring of hydrocarbons in ambient air. Presented at the
184th National Meeting of the American Chemical Society; September;
Kansas City, MO.
Olansky, A. S.; Deming, S. N. (1976) Optimization and interpretation of absor-
bance response in the determination of formaldehyde with chromotropic
acid. Anal. Chim. Acta 83:241-249.
Osman, M.; Hill, H. H.; Holdren, M. W.; Westberg, H. H. (1979) Vapor-phase sily-
lation of alcohols for air analysis. In: A. Zlatkis, ed. Advances in
Chromatography, Chromatography Symposium-University of Houston, Houston, TX.
Ozolins, G. ; Rehmann, G. (1968) Air Pollution Emission Inventory of Northwest
Indiana. A Preliminary Survey. Cincinnati, Ohio: U.S. Department of
Health, Education and Welfare, Public Health Service.
Peterson, E. W.; Tingey, D. T. (1980) An estimate of possible contribution of
biogenic sources to airborne hydrocarbon concentrations. Atmos. Environ.
14:79.
Pitts, J. N. , Jr.; Winer, A. M. , Darnall, K. R.; Lloyd, A. C. ; Doyle, G. J.
(1977) Hydrocarbon reactivity and the role of hydrocarbons, oxides of
nitrogen, and aged smog in the production of photochemical smog. In:
Proceedings of the International Conference on Photochemical Oxidant
Pollution and Its Control, Vol. II. U.S. Environmental Protection Agency,
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-600/3-77-016b.
Platt, U.; Perner, D. ; Winer, A. M. ; Harris, G. W. ; Pitts, J. N. , Jr. (1980)
Detection of NO- in the polluted troposphere by differential opitcal ab-
sorption. GeopHys. Res. Lett. 7: 89-92.
Platt, U.; Winer, A. M.; Biermann, H. W.; Atkinson, R.; Pitts, J. N., Jr. (1984)
Measurement of nitrate radical concentrations in continental air. Environ.
Sci. Techno!., in press.
Poli, A. A.; Zinn, T. L. (1973) A new continuous monitor for methane and
nonmethane hydrocarbons. Anal. Instrum. 11:135-142.
Rasmussen, R. A.; Chatfield, R. B.; Holdren, M. W.; Robinson, E. (1976) Hydro-
carbon levels in a midwest open-forested area. Technical report to
Coordinating Research Council, Inc. 219 Perimeter Center Parkway,
Atlanta, GA 30346.
Rasmussen, R. A. (1964) Terpenes: their analysis and fate in the atmosphere.
St. Louis, MO: Washington University. Ph.D. thesis.
Rasmussen, R. A. (1970) Isoprene: Identified as a forest-type emission to
the atmosphere. Environ. Sci. Techno!. 4:667.
OZONER/D 3-113 6/28/84
-------
Rasmussen, R. A. (1972) What do the hydrocarbons from trees contribute to air
pollution? J. Air Pollut. Control Assoc. 7:537.
Reckner, L. R. (1974) Survey of users of EPA-Reference Method for measurement
of non-methane hydrocarbon in ambient air. Research Triangle Park, NC:
U.S. Environmental Protection Agency, EPA report no. EPA-650/4-75-008.
Richter, H. G. (1983) Analysis of organic compound data gathered during 1980
in northeast corridor cities. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; EPA report no. EPA-450/4-83-017.
Ridley, B. A.; Hewlett, L. C. (1974) An instrument for nitric oxide measurement
in the stratosphere. Rev. Sci. Instrum. 45:
Riggin, R. M. (1983) Technical assistance document for sampling and analysis
of toxic organic compounds in ambient air. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-600/4-83-027.
Ripperton, L.; White, 0.; Jeffries, H. (1967) Gas phase ozone-pinene reactions.
Presented at the 147th National Meeting of the American Chemical Society,
Division of Water, Air, and Waste Chemistry.
Roberts, J. M.; Fehsenfeld, F. C.; Albritton, D. L.; Sievers, R. E. (1983)
Measurement of monoterpene hydrocarbons at Niwot Ridge, Colorado. J.
Geophys. Res. 88: 10667.
Salas, L. J.; Singh, H. B. (1982) Measurement of formaldehyde and acetaldehyde
in the ambient air. Submitted to Environ. Sci. Techno!.
Salo, A. E.; Whitz, S.; MacPhee, R. D. (1975) Determination of solvent vapor
concentrations by total combustion analysis: A comparison of infrared
with flame ionization detectors. Paper No. 75-33.2. Presented at the
annual meeting of the Air Pollution Control Association; June; Boston,
MA.
Salop, J.; Wakelyn, N. T.; Levy, G. F.; Middleton, E. M.; Gervin, J. C. (1983)
The application of forest classification from Landsat data as a basis for
natural hydrocarbon emission estimation and photochemical oxidant model
simulations in Southeastern Virginia. J. Air Pollut. Control Assoc.
33:17.
Sanadze, G. A.; Dolidze, G. M. (1962) CJ.HQ (isoprene) type in volatile emission
from leaves of plants. Soobshch. TVKad. Nauk Gruz. SSR 57:1222. [Trans.]
Sandberg, J. S.; Basso, M. J. ; Okin, B. A. (1978) Winter rain and summer
ozone: A predictive relationship. Science (Washington, DC) 200:1051.
Sawicki, E.; Mauser, T. R.; Stanley, T. W.; Elbert, W. (1961). The 3-raethyl-
2-benzothiazolone hydrazone test. Sensitive new methods for the detection,
rapid estimation, and determination of aliphatic aldehydes. Anal. Chem.
33:93-96.
Scaringelli, F. ; Rosenberg, P. E. ; Rehme, K. A. (1970) Comparison of the
permeation devices and nitrite ion as standards of colorimetric determina-
tion of nitrogen dioxide. Environ. Sci. Techno!. 4:924-929.
OZONER/D 3-114 6/28/84
-------
Schiff, H. I.; Pepper, D. ; Ridley, B. A. (1979) Tropospheric NO measurements
up to 7 km. JGR J. Geophysical Res. 84: ,
Seila, R. L. (1981) Nonurban hydrocarbon concentrations in ambient air north
of Houston, TX. In: Bufalini, J. J. ; Arnts, R. R. , eds. Atmospheric
Biogenic Hydrocarbons. Ann Arbor, MI: Ann Arbor Science Publishers,
Inc.
Seila, R. L. (1979) Nonurban hydrocarbon concentrations in the ambient air
north of Houston, Texas. Research Triangle Park, NC: U.S. Environmental
Agency; EPA report no. EPA-600/3-79-010.
Seila, R. L. ; Lonneman, W. A.; Meeks, S. A. (1976) Evaluation of polyvinyl
fluoride as a container material for air pollution samples. J. Environ.
Sci. Health Part A 2:121-130.
Sevcik, J. (1975) Detectors in Gas Chromatography. Amsterdam, The Netherlands:
Elsevier Scientific Publishing Company.
Sexton, K.; Westberg, H. (1984) Nonmethane hydrocarbon composition of urban
and rural atmospheres. Atmos. Environ, (in press).
Sexton, F. W. ; McElroy, F. F. ; Michie, R. M., Jr.; Thompson, V. L. (1981) A
comparative evaluation of seven automatic ambient non-methane organic
compound analyzers. Research Triangle Park, NC: U.S. Environmental
Protection Agency; cited in: U.S. Environmental Protection Agency (1981b),
EPA report no. EPA-600/4-81-015.
Siddiqi, A. A.; Worley, F. L., Jr. (1977). Urban and industrial air pollution
in Houston, Texas. I. Hydrocarbons. Atmos. Environ. 11:131-143.
Singh, H. B.; Salas, L. J.; Stiles, R. E. (1982) Distribution of selected
gaseous organic mutagens and suspect carcinogens in ambient air. Environ.
Sci. Technol. 16: 872-879.
Singh, H. B. (1980). Guidance for the collection and use of ambient hydrocar-
bon species data in development of ozone control strategies. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-450/4-80-008.
Sleva, S. F. (1965) Determination of formaldehyde-chromotropic acid method,
In: Interbranch Chemical Advisory Group. Selected methods for the
measurement of air pollutants. U.S. Department of Health, Education, and
Welfare; Public Health Service. Publication No. 999-AP-ll. Washington,
DC: U.S. Government Printing Office.
Smith, L. R. ; Black, F. M. (1980) Characterization of exhaust emissions from
passenger cars equipped with three-way catalyst control systems. June;
Dearborn, MI. SAE technical paper no. 800822.
Sollins, P.; Reichle, D. E. ; Olson, J. S. (1973) Organic matter budget and
model for a southern Appalachian Liriodendron Forest. EDFB-IBP-73-2.
OZONER/D 3-115 6/28/84
-------
Souza, A. F. ; Daley, P. S. (1978) U.S. Air Force Turbine Engine Emission
Survey. Volume III. Engine Model Summaries. CEEDO-TR-78-34. Civil and
Environmental Engineering Development Office, Air Force Systems Command,
Tyndall Air Force Base, Florida 32403.
Spicer, C. W.; Jones, P. W. (1977) The fate of aromatic hydrocarbons in photo-
chemical smog systems: Toluene. J. Air Pollut. Control Assoc. 27:
1122-1124.
Spicer, C. W.; Joseph, D. W.; Sticksel, R. R. ; and Ward, G. F. (1979) Ozone sources
and transport in the Northeastern United States. Environ. Sci. Techno!.
13: 975-985.
Springer, K. J. ; Stahman, R. C. (1977a) Diesel car emission—emphasis on
participate and sulfate. Warrendale, PA: Society of Automotive Engineers,
Inc.; SAE paper no. 770254.
Springer, K. J. ; Stahman, R. C. (1977b) Unregulated emissions from diesels
used in trucks and buses. Presented at SAE International Automotive
Engineering Congress and Exposition; February-March; Detroit, MI. SAE
paper no. 770258.
Stedman, D. H.; Harvey, R. B.; Eastman, J.; Fington, D. J.; Wallace, J. (1977)
Ozone formation over Lake Michigan. Presented at the 4th Joint Conference
on Sensing of Environmental Pollutants; November; New Orleans, LA.
Stephens, E. R. ; Burleson, F. R. (1969) Distribution of light hydrocarbons in
ambient air. J. Air Pollut. Control Assoc. 19:929.
Sweeny, M. P.; Swartz, D. J. ; Rost, G. A.; MacPhee, R.; Chao, J. (1964) Contin-
uous measurements of oxides of nitrogen in auto exhaust. J. Air Pollut.
Control Assoc. 14:249-254.
Tilton, B. E. ; Bruce, R. M. (1981) Review of criteria for vapor-phase hydro-
carbons. Research Triangle Park, NC. U.S. Environmental Protection
Agency; EPA report no. EPA-600/8-81-022.
Tingey, D. T. ; Evans, R.; Gumpertz, M. (1981) Effects of environmental con-
ditions on isoprene emission from live oak. Planta 152:565.
Tingey, D. T. ; Manning, M.; Grothaus, L. C.; Burns, W. F. (1980) Influence of
light and temperature on monoterpene emission rates from slash pine.
Plant Physio!. 65:797.
Tuazon, E. C. ; Graham, A. M. ; Winer, A. M. ; Easton, R. R. ; Pitts, J. N. , Jr.
(1978) A kilometer pathlength Fourier-transform infrared system for the
study of trace pollutants in ambient and synthetic atmospheres. Atmos.
Environ. 12:865-875.
Tuazon, E. C. ; Winer, A. M. ; Graham, R. A.; Pitts, J. N., Jr. (1980) Atmos-
pheric measurements of trace pollutants by kilometer-pathlength FT-IR
spectroscopy. Adv. Environ. Sci. Technol. 10:259-300.
OZONER/D 3-116 6/28/84
-------
Tuazon, E. C. ; Winer, A. M.; Pitts, J. N., Jr. (1981) Trace pollutant concen-
trations in a multiday smog episode in the California South Coast Air
Basin by long pathlength Fourier transform infrared spectrometry. Environ.
Sci. Techno!. 15: 1232-1237.
Tucker, A. W. ; Birnbaum, M.; Fincher, C. L. (1975). Atmospheric N02 determi-
nation by 442 mm laser-induced fluorescence. Appl. Optics 14:[pages].
Tucker, A. W. ; Petersen, A. B.; Birnbaum, M. (1973). Fluorescence determina-
tion of atmospheric NO and N02. Appl. Opt. 12:[pages].
U.S. Department of Energy. (1978). Energy Data Reports. Washington, DC:
U.S. Department of Energy. Publication no. EIA-0049/1.
U.S. Department of Health, Education, and Welfare (1973). Manual of analyt-
ical methods, Vol. 1. Washington, DC: U.S. Government Printing Office.
Cited in: National Academy of Sciences. (1981) Formaldehyde and Other
Aldehydes. Washington, DC: National Academy Press.
U. S. Department of Health, Education, and Welfare (1965). Selected methods
for the measurement of air pollution. Washington, DC: U.S. Government
Printing Office. Publication no. PHS 999-AP-ll.
U.S. Environmental Protection Agency. (1975). Reference method for determina-
tion of hydrocarbons corrected for methane. In: Code of Federal Regu-
lations. Title 40, Part 50.11, Appendix E. U.S. Government Printing
Office, Washington, DC. pp. 21-27.
U.S. Environmental Protection Agency. (1976a). Ambient air monitoring refer-
ences and equivalent methods. Code of Federal Regulations, Title 40,
Part 53.
U.S. Environmental Protection Agency. (1976b). Nitrogen dioxide measurement
principle and calibration procedure. Notice of final rulemaking. Fed.
Regist. 41:52686-5294, December 1.
U.S. Environmental Protection Agency, National Emissions Data System, file for
1978. Office of Air Quality Planning and Standards, Research Triangle
Park, NC.
U.S. Environmental Protection Agency. (1978a) Air quality criteria for ozone
and other photochemical oxidants. Research Triangle Park, NC: U.S.
Environmental Protection Agency; EPA report no. EPA-600/8-78-004.
U.S. Environmental Protection Agency (1978b) Procedures for quantifying
relationships between photochemical oxidants and precursors: supporting
documentation. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no EPA-450/2-77-021b.
U.S. Environmental Protection Agency. (1980a). Procedures for the preparation
of Emission Inventories for Volatile Organic Compounds, Vol. 1, 2nd Ed.
Research Triangle Park, NC: U.S. Environmental Protection Agency: EPA
report no. EPA-450/2-77-028,
OZONER/D 3-117 6/28/84
-------
U.S. Environmental Protection Agency. (1980b). Final emission inventory re-
quirements for 1982 ozone state implementation plans. Research Triangle
Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-450/4-80-016.
U.S. Environmental Protection Agency. (1981a). Procedures for emission inven-
tory preparation. Vol. I. Emission inventory fundamentals. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-450/4-81-026a.
U.S. Environmental Protection Agency. (1981b). Procedures for emission inven-
tory preparation. Vol. II. Point sources. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-450/4-81-026b.
U.S. Environmental Protection Agency. (1981c). Procedures for emission inven-
tory preparation. Vol. III. Area sources. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-450/4-81-02Sc.
U.S. Environmental Protection Agency. (1981d). Procedures for emission inven-
tory preparation. Vol. IV. Mobile sources. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-450/4-81-026d.
U.S. Environmental Protection Agency. (1981e). Procedures for emission inven-
tory preparation. Vol. V. Bibliography. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-450/4-81~026e.
U.S. Environmental Protection Agency. (1981f). Technical assistance document
for the calibration and operation of automated ambient non-methane organic
compound analyzers. Research Triangle Park, NC: U.S. Environmental
Protection Agency, EPA report no. EPA-600/4-81-015.
U.S. Environmental Protection Agency. (1982). Air quality criteria for
oxides of nitrogen. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/8-82-026.
U.S. Environmental Proetection Agency. (1983). National air pollution emission
estimates, 1940-1982. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-450/4-83-024,
Wasser, J. H. (1976). Emission characteristics of small gas turbine engines.
In: Proceedings of the stationary source combustion symposium, volume
III: field testing and surveys. Research Triangle Park, NC: U.S. Envi-
ronmental Protection Agency; EPA report no. EPA-600/2-76-152c.
Wells, C. G. (1981). Methods of estimating foliar biomass pp. 107-114. In:
Bufalini, J. J. ; Arnts, R. R., eds. Atmospheric Biogenic Hydrocarbons,
Vol. 1. Ann Arbor, MI: Ann Arbor Science.
West, P. W; Sen, B. Z. (1956) Spectrophotometric determination of nitrate
using chromotropic acid. Z. Anal. Chem. 153:177-183.
Westberg, H.; Allwine, G.; Robinson, E. (1978) Measurement of light hydrocarbons
and studies of oxidant transport beyond urban areas - Houston study 1976.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-600/3-78-062.
OZONER/D 3-118 6/28/84
-------
Westberg, H.; Lamb, B. (1982) Milwaukee ozone study - 1981. Final Report for
Contract No. NRA 98571. Wisconsin Department of Natural Resources,
Madison, WI.
Westberg, H. ; Lamb, B. (1983) Ozone production and transport in the Atlanta,
GA region. Final report for EPA grant no. CR 809221; Environmental
Sciences Research Laboratory, U.S. Environmental Protection Agency;
Research Triangle Park, NC.
Westberg, H.; Allwine, G. (1984) Analysis and compilation of organic pollutant
data aloft. EPA Report (in press).
Westberg, H. H.; Holdren, M. W.; Hill, H. H. (1980) Analytical methodology for
the identification and quantification of vapor phase organic pollutants.
CRC-APRAC Project No. CAPA-11-71, Coordinating Research Council, New
York, NY.
Westberg, H. ; Lamb, B.P; Lonneman, W. (1983). Oxidant production and trans-
port in the Atlanta, GA, region. Presentation at the 76th annual meeting
of the Air Pollution Control Association; June; Atlanta, GA.
Westberg, H. H.; Rasmussen, R. A.; Holdren, M. W. (1974) Gas chromatographic
analysis of ambient air for light hydrocarbons using a chemically bonded
stationary phase. Anal. Chem. 46:1852-1854.
Whitby, R. A.; Coffey, P. E. (1977) Measurement of terpenes and other organics
in an Adirondack mountain pine forest. J. Geophys. Res. 82: 5928-5934.
Williams, D. T.; Kolitz, B. L. (1968). Molecular correlation spectrometry.
Appl. Opt. 7:607-616.
Wimmer, D. B.; McReynolds, L. A. (1962). Nitrogen oxides and engine combus-
tion. In: SAE Transactions 1962. New York, New York: Society of Auto-
motive Engineers; pp. 733-744.
Winer, A. M. (1983) Investigation of the role of natural hydrocarbons in
photochemical smog formation in California. Final Report, Contract No.
AO-056-32, California Air Resources Board, Sacramento, CA.
Winer, A. M. ; Peters, J. M. ; Smith, J. P.; Pitts, J. N., Jr. (1974) Response
of commercial chemiluminescent N0-N02 analysis to other nitrogen-contain-
ing compounds. Environ. Sci. Technol. 8:1118-1121.
Zimmerman, P. R. (1979). Determinations of emission rates of hydrocarbons
from indigenous species of vegetation in the Tampa/St. Petersburg, Florida,
area. Final Report, Appendix C. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; EPA report no. EPA-904/9-77-028.
Zimmerman, P. R. (1980). Natural sources of ozone in Houston: natural or-
ganics. In: Proceedings of Specialty Conference on Ozone/oxidants-
Interactions with the Total Environment. Pittsburgh, PA: Air Pollution
Control Association.
OZONER/D 3-119 6/28/84
-------
Zimmerman, P. R. (1977). Procedures for conducting hydrocarbon emission in-
ventories of biogenic sources and some results of recent investigations.
In: Emission Inventory/Factor Workshop. Vol. 2. Research Triangle Park,
NC: U.S. Environmental Protection Agency, Office of Air Quality Planning
and Standard; EPA report.
OZONER/D 3-120 6/28/84
-------
4. CHEMICAL AND PHYSICAL PROCESSES IN THE FORMATION AND OCCURRENCE OF OZONE
AND OTHER PHOTOCHEMICAL OXIDANTS
4.1 INTRODUCTION
In the preceding chapter, the nature and identity of precursors to ozone
and other photochemical oxidants were discussed, sources of those precursors
were identified, the amount of precursors emitted into the atmosphere from
various sources was estimated, and the concentrations of precursors actually
measured in the ambient air were described.
The present chapter provides an overview of the complex chemical and
physical processes by which ozone and other photochemical oxidants are formed
from their precursors. In addition, the present chapter describes the physical
processes that result in the transport and dispersion of ozone and the other
oxidants once they are formed. That discussion also includes a brief summary
of the processes by which stratospherically formed ozone can be brought into
the troposphere, through the boundary and sub-laminar layers, and to the
surface.
The subsequent chapter (Chapter 5) presents summary information on the
reactions of ozone and other photochemical oxidants in ambient air and in
biological systems. The present chapter includes, however, a brief discussion
of the relationship of ozone and the other oxidants to atmospheric phenomena
that result from the formation of secondary organic and inorganic aerosols, a
process to which ozone and other oxidants contribute indirectly.
4.2 CHEMICAL PROCESSES
The photochemistry of the polluted atmosphere is exceedingly
complex. Even if one considers only a single hydrocarbon pollutant,
with typical concentrations of nitrogen oxides, carbon monoxide,
water vapor, and other trace components of air, several hundred
chemical reactions are involved in a realistic assessment of the
chemical evolution of such a system. The actual urban atmosphere
contains not just one but hundreds of different hydrocarbons, each
with its own reactivity and oxidation products.
(National Academy of Sciences, 1977)
In order to understand the effects of ozone and other photochemical oxi-
dants on humans, vegetation, and other receptors, however, it is not necessary
019VP/D 4-1 5/2/84
-------
to comprehend all the complex chemical reactions that take place during forma-
tion of these species in the atmosphere. It is sufficient to understand how
photochemical oxidants result from the action of sunlight on precursor compounds.
For a complete and detailed discussion of the many complex reactions thought
to take place in polluted atmospheres, the reader is referred to Demerjian
et al. (1974) and to Atkinson and Lloyd (1984). For a simple explanation of
the origin of these secondary pollutants, the following summarizes current
understanding of the photochemical processes leading to their production.
4.2.1 Formation of Ozone and Oxidants
The concentrations of ozone and oxidants found in urban areas and in
downwind and rural receptor regions are the net result of at least three
general processes: first, the initial emission, dispersion, and then transport
of precursors of ozone and oxidants; second, the photochemical reaction pro-
cesses that occur in the atmosphere as the dispersion and transport take place;
and third, the scavenging processes along the trajectory that act to reduce the
concentrations of both precursors and the resulting ozone and oxidants. This
section discusses briefly the chemical reactions that take place, while the
following section (4.3) discusses the meteorological and climatological
processes that influence the formation and distribution of ozone and other
photochemical oxidants.
In the troposphere, ozone is formed indirectly through the action of
sunlight on nitrogen dioxide (N0~). Sunlight decomposes N0» into nitric oxide
(NO) and an oxygen atom:
N02 + sunlight —> NO + 0 (4-1)
The oxygen atom (0) liberated in this process combines with an oxygen molecule
to produce ozone:
0 + 02 + M > 03 + M (4-2)
In the absence of any competing reactions, the ozone formed in reaction (4-2)
combines with the NO liberated in reaction (4-1) to regenerate an NO™ molecule:
03 + NO * N02 + 02 (4-3)
019VP/D 4-2 5/2/84
-------
As a result of the above three reactions, an equilibrium or steady-state
condition is established among NO, N0«, and 0,, and the concentration of Q~ in
the atmosphere is governed by the expression,
[03] = K [NOjJ, (4-4)
where K is a constant that depends on the sunlight intensity. Typically, K in
the lower troposphere is less than or equal to 0.025 ppm. It is apparent from
equation (4-4) that very little buildup of ozone can occur until most of the
NO that has been emitted into the atmosphere is converted to NOp. Equations
(4-1) through (4-3), however, cannot by themselves explain the buildup of
ozone, since for each molecule of NO oxidized to N0? in equation (4-3) a
molecule of ozone is also destroyed.
An alternate pathway of conversion of NO to NOp that does not destroy 0~
is needed to explain the high ozone levels observed in the urban environment.
Such an alternate pathway is available through the oxidation of reactive
organic species such as hydrocarbons. In the atmosphere, these species can be
oxidized by hydroxyl radicals (HO-)- There are a number of sources of HO
radicals in the lower troposphere. One such source is nitrous acid. This
species, which is formed in the tailpipes of automobiles, reacts with sunlight
in the atmosphere to produce HO radicals. The HO radicals formed in this and
other processes react with hydrocarbons to generate an alkyl radical (R'):
HO- + hydrocarbon > R- + H^O (4-5)
This hydrocarbon radical, R*, quickly picks up an oxygen molecule to form a
peroxy radical, ROp*:
R. + Q2 > R0j£- (4-6)
The next reaction in the series is generally thought to be a conversion of NO
to NOp, at the same time producing an alkoxy radical:
ROZ- + NO > RO- + NO;, (4-7)
019VP/D 4-3 5/2/84
-------
The alkoxy radical reacts with an oxygen molecule to produce a hydroperoxy
radical and an aldehyde:
RO- + Oj, » R'CHO + HO*- (4-8)
Finally, the cycle is completed when the hydroperoxy radical oxidizes another
molecule of NO to N02 and forms a hydroxy radical, which then begins the cycle
over again:
HOy + NO > HO- + NO^ (4-9)
Other, more complex reactions may also occur (see Atkinson and Lloyd, 1984).
This description is highly simplified, but the reactions listed "above
contain the main features of the NO-to-N02 oxidation by organic and other
radicals that leads to subsequent ozone formation. The essential ingredients
are sunlight, NO or N02, and organic compounds. The latter two are emitted
in abundance in urban areas as shown in section 3.4. The number and kind of
hydrocarbon species formed during the course of atmospheric photochemical
reactions is quite large. Present analytical techniques permit identification
and measurement of individual hydrocarbon species at part-per-billion concen-
trations. In a recently published compendium, Graedel (1978) has documented
the emission into or detection in ambient air of more than 1000 organic
compounds.
4.2.2 Initiation and Termination of Photochemical Reactions
The initial source of radicals is an important aspect of the photochemis-
try of polluted atmospheres. Although the rate and yield of oxidant formation
depend on many factors, the length of the induction period before the accumula-
tion of oxidant begins depends heavily on the initial concentration of radicals.
In smog chambers, the photolysis of nitrous acid, HONO, may be the most impor-
tant initial source of radicals (Pitts et a!., 1977). There is also evidence
to suggest that HONO may be a source of free radicals in the atmosphere as
well (Perner and Platt, 1979; Harris et al., 1982). Nitrous acid has been
observed in urban atmospheres at concentrations up to 8 ppb (Platt et al.,
1980; Harris et al., 1982).
019VP/D 4-4 5/2/84
-------
Probably the most important source of radicals in the atmosphere is the
photolysis of aldehydes:
RCHO + sunlight > HCO- + R- (4-10)
The HCO radicals liberated in this process rapidly react with oxygen:
HCO- + Oj, > HO*- + CO (4-11)
As explained in the previous section (4.2.1), the subsequent reaction of HO^*
(and ROp- as well) with NO is the major route to the oxidation of nitric oxide
in ambient air. Aldehydes are emitted from many sources, including automobiles
(section 3.4.1). They are also formed as secondary pollutants in smog.
During the course of the overall smog formation process, the supply of
free radicals is maintained by several sources, but the dominant one appears
to be the photolysis of aldehydes formed in the atmosphere from the initial
hydrocarbons. Since the reactions of free radicals with NO form a cyclic
process, any additional source of radicals will add to the supply of radicals
and will increase the rate of the cycle. Conversely, any reaction that removes
free radicals will slow the rate of the cycle.
Although these photochemical reactions require sunlight, the presence of
sunlight does not mean that the reactions continue indefinitely. Terminating
reactions gradually remove NO and N0? from the reaction mixtures such that the
cycles would slowly come to an end unless fresh NO emissions were injected
/\
into the atmosphere.
Termination of the chain reactions frequently leads to the formation of
other oxidants as well as relatively stable organic nitrates in the atmosphere.
Nitrous acid (HONO), nitric acid (HN03), peroxynitric acid (HOON02), hydrogen
peroxide (H*®?^' peroxyacetyl nitrate, (CH^COOpNOp), other peroxyacyl nitrates,
organic hydroperoxides, and organic peracids have all been observed either in
smoggy atmospheres or in irradiated laboratory mixtures (National Academy of
Sciences, 1977). These compounds are almost always found in very low concentra-
tions in ambient air and may actually occur only as intermediates in the
photochemical degradation of organic compounds. Even though they occur at low
concentrations, however, many of these play significant or even critical roles
in atmospheric chemistry (Pitts et a!., 1983).
019VP/D 4-5 5/2/84
-------
Figure 4-1 (Niki et al., 1978) shows some of the reactions involved in
the HO--initiated photooxidation of cis- or trans-2-butene in an atmosphere
containing NO and NO^. It can be seen that this involves a chain process and
that the major organic product is acetaldehyde. As indicated by Niki et al.
(1978) and Carter et al. (1979), the pathway leading to formic acid does not
occur to any observable extent.
4.2.3 Limitations to Ozone Accumulation
The nature of photochemical systems can be partially explained by consider-
ing their behavior as a function of the initial concentrations of NO and
X
hydrocarbons, as well as the ratio of these two reactants, i.e., the NMOC/NO
ratio. At low NMOC/NO ratios (usually ratios of less than about 1:1 to 2:1),
/\
the rate at which NO is converted to NO,, is influenced by the availability of
organic compounds. In such a hydrocarbon-deficient and NO -rich system, few
s\
organic free radicals are available to effect the conversion of NO to NO,,.
The oxidation of NO proceeds at such a slow rate that a high NOp/NO ratio may
not be achieved by sunset and the buildup of ozone may therefore be limited.
At moderately high NMOC/NO ratios (usually greater than about 5:1 to 8:1),
/\
sufficient organic radicals are available to oxidize all of the NO. The
NOp/NO ratio, therefore, is favorable to 0- accumulation.
At very high NMOC/NO ratios, NO will be oxidized rapidly to N0?. The
/\ £
large number of organic radicals present in this system, however, will then
quickly consume a substantial portion of the N0?. Nitrate formation will
increase, which will effectively lower the N02/N0 ratio and limit 0, buildup.
Ozone formation also will be limited as a result of the reaction of 0- with
excess olefins.
4.2.4 Recent Work on Photochemical Smog Reactions
The hydrocarbons so important in the chemistry of the polluted troposphere
are alkanes (paraffins), alkenes (olefins), and aromatics (section 3.2). In
addition, the oxygenated hydrocarbons such as aldehydes, ketones, dicarbonyls,
and perhaps some alcohols are also important, although they are always found
in much smaller concentrations in ambient air than the hydrocarbons (section
3.5).
The photooxidation reactions of the alkanes is fairly well understood
(Hampson and Garvin, 1978; Atkinson et al., 1982). The only significant
019VP/D 4-6 5/2/84
-------
CH3CH = CHCH3
HO (1)
CH3CH(OH)CHCH3
02 (2)
CH3CH«OH)CH(CH3)06
NO (3)
CH3CH(OH)CH(CH3)O + IMO2
(4)
(5a) O2 (5b)
CH3CH(OH)00
NO (6b)
CH3CHIOH)O + NO2
(7b)
i'
CH3
Figure 4-1. Reaction scheme for the HO-initiated ox-
idation of 2-butene-IMO system. (aThe overall
stoichiometry for CH3 oxidation is CH3 + 2O£ +
2NO — CH2O + HO + 2IMO2; Niki et al., 1972;
Demerjian et al., 1974.)
Source: Niki et al. (1978).
019VP/D
4-7
5/2/84
-------
reaction involving alkanes is oxidation by HO radicals. These radicals abstract
a hydrogen atom from the alkane to produce an alkyl radical and H^O as shown
in reaction 4-5. Alkenes, the most reactive class of hydrocarbons in the
lower troposphere, undergo reaction with both HO* and 0~. The reaction path-
ways important in the alkene-OH* reaction are shown in Figure 4-1. The reac-
tion of alkenes with ozone leads to the formation of a number of free radicals
and stable products. Taking trans-2-butene as an example, ozone adds to the
double bond to form a molozonide. The molozonide then undergoes rapid decom-
position to form an aldehyde and a biradical intermediate:
o
A
CH3CH=CHCH3 + 03 —> CH3CH-CHCH3 —» CH3CHO + 00-C (4-11)
(Molozonide) CH3
The biradical formed in this process can undergo a number of reactions. It
can thermally decompose to yield a variety of free-radical intermediates or it
can react with a number of species, including NO, N02, S02, H20, and aldehydes
(Su et al., 1980; Niki et al., 1981; Dodge and Arnts, 1978).
Knowledge about the reactions of the aromatic compounds in the atmosphere
is not nearly as complete as knowledge about the reactions of alkanes and
alkenes. Although aromatics comprise between 20 and 30 percent of the VOC
carbon in urban atmospheres, the reaction intermediates and final reaction
products of the aromatics are not well known. Different research groups
(Killus and Whitten, 1982; Atkinson et al., 1980) have constructed aromatic
mechanisms that model, with a certain success, smog chamber systems containing
aromatics and NO . This is not to say that these mechanisms accurately de-
scribe the chemistry involved, but that within the bounds of uncertainty that
exist in the reactions, their products, and associated rate constants, these
models can be "tuned" to predict NO, N09, and 0- behavior. Much additional
c. o
work, however, is needed to describe the chemical processes that take place.
Laboratory studies show that under ambient conditions, H0« attack on
aromatics is initially the primary path of reaction (Hendry et al., 1978;
Atkinson et al., 1979). In terms of 03 produced, their reactivity increases
from benzene to toluene to the xylenes. The aromatic rings are cleaved
(Killus and Whitten, 1982; Atkinson et al., 1980) and one of the end products
019VP/D 4-8 5/2/84
-------
of the reaction pathways is peroxyacetyl nitrate (PAN) (Darnall et al., 1979).
A noteworthy finding in the past 5 years is that some nitrogenated pro-
ducts of the photochemical smog system, such as PAN and possibly peroxynitric
acid (HOONCK) (Kamens et al., 1981), have a greater mechanistic role than
thought earlier. Previous investigations of PAN chemistry (Pate et al., 1976;
Cox and Roffey, 1977; Hendry and Kenley, 1977) had shown that PAN can ther-
mally decompose to a peroxyacyl radical and N0« and that the rate of decompo-
sition is extremely temperature-dependent. Therefore, significant levels of
PAN can build up early in the day when temperatures are relatively low. In
the late afternoon, when ambient temperatures are higher, the decomposition of
PAN can proceed at a rapid rate, liberating N0? molecules that can lead to
enhanced 0- production. Despite, however, the tendency toward thermal decom-
position at afternoon temperatures, the rate of formation of PAN may be so
high in the afternoon that PAN concentrations may reach their peak at that
time (Tuazon et al., 1981).
The natural hydrocarbons, i.e., organic compounds emitted from vegetation,
can also react with NO in the presence of sunlight to form 0~, and perhaps
/\ O
other oxidants. On a global basis, the quantities of natural hydrocarbons
emitted are higher than those from manmade sources (section 3.4.2). Their
concentrations in ambient air are much lower, however, especially in the
vicinity of urban areas (sections 3.4.2 and 3.5). More important, recent
reports (Arnts and Gay, 1979; Arnts et al., 1981; Roberts et al., 1983) continue
to confirm that natural hydrocarbons play the dual role of ozone precursor and
ozone scavenger and that the latter role seems to be the more important one.
Thus, natural hydrocarbons are not thought to contribute much ozone or other
oxidants to urban environments. The contribution of natural hydrocarbons to
the formation of ozone generally and the contribution to urban ozone specifically
remain areas of debate. The published literature on the role of natural hydro-
carbons has recently been critically reviewed by Altshuller (1983).
4.2.5 Relationship of Ozone to Aerosol-Related Phenomena
In addition to having direct adverse effects on human health and on
vegetation, ecosystems, and nonbiological materials, ozone can contribute
indirectly to visibility degradation and to acidic deposition via its partici-
pation in the formation of both organic and inorganic aerosols (National
019VP/D 4-9 5/2/84
-------
Academy of Sciences, 1977; U.S. Environmental Protection Agency, 1982a). It
is well established that the source of the vast majority of manmade sulfate
aerosol in the atmosphere is the oxidation of sulfur dioxide (SOp), ultimately
to sulfuric acid (H2S04). The correlations between elevated levels of ozone
and of sulfate aerosol in ambient air have been noted by several investigators
in field studies concerned with visibility reduction by aerosols. Wilson
(1978) and Gillani et al. (1981) have pointed out that atmospheric mixing
intensity and the background 0~ concentration are the two most important
factors in determining S02 oxidation at relative humidities lower than 75
percent. It is also clear, however, that the rate of reaction of CL with SCL
•J £
is far too slow to account for observed formation rates of sulfate aerosol
(U.S. Environmental Protection Agency, 1982a).
Of the many possible gas-phase reactions of S02, only a few appear to
have any significance in the production of sulfate aerosol. The reaction of
HO* with SCL appears to be the dominant pathway for the oxidation of S0?
(Calvert and Stockwell, 1983; Calvert and Mohnen, 1983). A recent analysis by
Stockwell and Calvert (1983) implicates the formation of HOS02 radicals from
the reaction of HO- with SO^, followed by reaction with 0^, as the favored
mechanism for the formation of SO,:
HO- + S02 (+ M) > HOS02- (+ M); (4-12)
HOS02- + 02 » H02- + S03. (4-13)
From the reaction of HOp* with NO, HO- is regenerated:
HQ2. + NO > HO- + N02- (4-14)
and the cycle begins again (see equation 4-9, section 4.2.1). The importance of
the reaction of HO- with SO,, in the atmosphere is supported by observations of
power plant plumes, in which no aerosol is formed at night when the HO- concen-
tration in the ambient air is negligible; and none is formed during the day
before the plume is well mixed with ambient air (the ambient air contains much
higher concentrations of HO- and 03 than the plume) (Blumenthal et al., 1981;
Davis et al., 1979).
In addition to sources of HO- already discussed (equation 4-11 and the
photolysis of HONO), the photolysis of 0,, in relatively clean background air,
019VP/D 4-10 5/2/84
-------
and the subsequent reaction of the liberated oxygen atom with a water molecule
provide an important source of HO* radicals for the SO,, oxidation (Hegg and
Hobbs, 1980). Similarly, the reaction of 0- with olefins in polluted air
leads to the formation of HO* radicals. One can therefore conclude that 0,,
though it does not react directly with SO,, at an appreciable rate, plays an
important indirect role in the transformation of SO,, to sulfate aerosol
via homogeneous oxidation of S02 in both clean and polluted atmospheric systems.
That contribution can not be quantified, however, at least at present (U.S.
Environmental Protection Agency, 1982a).
Despite the poor quality of the available data, the possibility that
nitrate aerosol is a contributor to visibility reduction should not be neglec-
ted. The role of 07 in the formation of this aerosol species is briefly
O
considered here. The principal manmade nitrogen emissions of interest here
are NO and N02, the vast majority of which are NO. This species is relatively
insoluble in water (section 3.2) and does not react with water in any signifi-
cant manner. Thus, NO must be converted to some more highly oxidized form in
order to participate in the formation of particulate nitrate.
The oxidation of NO to N0? can occur through thermal oxidation at very
high concentrations of NO such as those in and very near the stacks of power
plants (U.S. Environmental Protection Agency, 1982b). This generates only a
small portion of the N02 formed in the atmosphere, however. As explained
earlier, the most important reactions leading to formation of NO^ in ambient
air are:
NO + 03 > N02 + 02 (4-15)
NO + H02- * N02 + HO- (4-14)
Thus, if N02 is a precursor of nitrate aerosol, 03 plays a significant direct
role in its formation by oxidizing NO, and an indirect role by leading to
formation of H02« radicals as discussed above. Following oxidation of NO, the
N0? can be converted in the gas phase to nitric acid vapor through either of
two pathways:
N02 + HO- > HN03 (4-16)
or,
019VP/D 4-11 5/2/84
-------
N02 + 03 > N03 + 02 (4-17)
N02 + N03 »- N205 (4-18)
N20S + H20 > 2HN03 (4-19)
The first pathway, which is dominant during the daytime, requires the presence
of HO- radicals that are produced during the formation of 0_, as well as from
other sources. The second pathway to HN03 is dominant at night and directly
involves reaction with 0~.
Nitric acid (HNO-), once it has been produced in the gas phase, is suffi-
ciently volatile to remain in the atmosphere as a vapor. The available labora-
tory and ambient air data indicate, however, that HNO- vapor reacts with
ammonia to form NH4N03 (Appel et al., 1980; Doyle et al., 1979; Stelson et al.,
1979), which, because of its low vapor pressure, will form nitrate aerosol
particles. Evidence also indicates that HNO, vapor will react with NaCl
aerosol in the following way:
HN03 + NH3 > NH4N03 (4-20)
HN03 + NaCl > NaN03 + HC1 (4-21)
This second reaction (equation 4-21) may account for the fact that much of the
observed particulate nitrate in Los Angeles is found in the coarse mode (Farber
et al., 1982). Obviously, the importance of this mechanism for nitrate aerosol
formation is determined by the availability of sea salt particles.
Sulfate and nitrate aerosols are present at significant levels in the
atmosphere in the form of just a few compounds. In contrast, secondary organic
aerosols are composed of a large number of species, but there is no clear
consensus concerning which ones contribute most to the mass concentration.
For all the species that are found in the secondary organic aerosol, however,
the fundamental formation mechanism is the same (chapter 5). The vapor-phase
precursor undergoes some reaction that results in formation of a product
having an equilibrium vapor pressure sufficiently low that condensation,
nucleation, or both are possible at the gaseous concentration achieved. From
the available data, it seems clear that the more highly oxygenated, larger-
019VP/D 4-12 5/2/84
-------
carbon-number species generally are those precursors likely to form secondary
aerosols in the atmosphere.
For an earlier but thorough review of the formation of secondary aerosol,
the reader is referred to the 1977 monograph on ozone and other photochemical
oxidants by the National Academy of Sciences (1977). Two recent criteria
documents prepared by the U.S. Environmental Protection Agency (1982a; 1982b)
contain thorough discussions of the contributions of ozone and of hydrogen
peroxide, also an oxidant, to the oxidation of S0? and NO^, as well as of the
respective roles of SO^ and N02 as precursors to visibility impairment and
acidic deposition.
The reactions of manmade volatile organic compounds that produce aerosols
were thoroughly reviewed and documented in the National Academy of Sciences
monograph cited above. Biogenic as well as manmade volatile organic compounds,
however, can participate in aerosol formation (Altshuller and Bufalini, 1971;
Arnts and Gay, 1979). Direct experimental evidence of aerosol formation, how-
ever, along with product analysis is available for only a limited number of
natural compounds, crpinene and p-pinene (National Academy of Sciences, 1977;
Hull, 1981; Schwartz, 1974), mainly because the analysis and characterization
of these kinds of products at ambient concentrations is extremely difficult.
Hull has conducted experiments with these two compounds at high concentrations
in a small tube reactor. Analysis of the products showed, on a weight basis,
that almost all of the reacted crpinene carbon was found in the condensed materials
extracted from the walls. Although the products he identified from these
experiments were either in the condensed phase or on the walls, Hull suggested
that at the a-pinene levels found in ambient air these products have a high
enough vapor pressure to exist both in the gas phase and in aerosols (Hull,
1981). In his recent review of the role of biogenic volatile organic compounds,
Altshuller (1983) also discusses at length the contribution of these compounds
to ambient air aerosols.
4.3 METEOROLOGICAL AND CLIMATOLOGICAL PROCESSES
As discussed in the previous section, ozone and oxidants are formed by
the action of sunlight on the precursors, N02 and hydrocarbons. The accumula-
tion of the products to form an appreciable concentration is also dependent,
however, on the prevailing meteorology in the vicinity of the precursor emissions.
019VP/D 4-13 5/2/84
-------
To understand the details of the effects of meteorology on air quality requires
a thorough knowledge of meteorology and climatology, but an appreciation of the
general factors important in the formation of elevated concentrations of oxidants
is relatively easy to acquire. Following is a brief presentation of some fea-
tures of atmospheric mixing and transport that will provide a basic understanding
of the meteorological factors that affect the concentrations of ozone and other
oxidants in urban and rural areas.
4.3.1 Atmospheric Mixing
The concentration of an air pollutant depends significantly on the degree
of mixing that occurs between the time a pollutant or its precursors are
emitted and the arrival of the pollutant at the receptor. Since, to a first
approximation, the diurnal cycle of weekday urban emission patterns for ozone
and oxidant precursor pollutants is generally uniform, it is reasonable to
ascribe a significant proportion of the large day-to-day changes in pollutant
concentrations to changes in meteorological mixing processes. The rate at
which atmospheric mixing processes occur and the extent of the final dilution
of the pollutants depends on the amount of turbulent mixing and on wind speed
and wind direction. Moreover, the transport of pollutants and precursors from
a source region to a distant receptor is also dependent on wind speed and wind
direction.
The degree of turbulent mixing can be characterized by atmospheric sta-
bility. In an atmospheric layer with relatively low turbulence, pollutants do
not spread as rapidly as they do in an unstable layer. Also, because a stable
layer has a relatively low rate of mixing, pollutants in a lower layer will not
mix through it to higher altitudes. The stable layer can act as a trap for air
pollutants lying beneath it. Hence, an elevated inversion is often referred
to as a "trapping" inversion. Also, if pollutants are emitted into a stable
layer aloft, such as from a stack, the lack of turbulence will keep the efflu-
ents from reaching the ground while the inversion persists.
It is common in air pollution considerations to equate a stable atmospheric
layer or situation with a temperature inversion, which is a layer of the
atmosphere in which the temperature increases with increasing altitude, because
inversions are common and also represent the most stable atmospheric situations.
The lowest part of an inversion layer is called the base and is defined as the
019VP/D 4-14 5/2/84
-------
altitude at which the temperature begins to increase. The top of the inversion
is the point at which the temperature begins to decrease with increasing
altitude. The distance between the base and top of the inversion layer is the
"depth" or "thickness" of the inversion. Inversion layers may begin at the
ground surface (i.e., the altitude of the base is zero), or the entire inver-
sion layer may be above the surface. The former is known as a "surface inver-
sion" and the latter as an "elevated inversion." The two types are usually
caused by different sets of weather conditions, but it is not unusual for both
types of inversions to be present at a given location at the same time. In
the United States, surface inversions are characteristic of nighttime and
early morning hours except when heavy cloud cover or windy and stormy condi-
tions prevail.
Surface and elevated inversion layers are both important in determining
pollutant concentration patterns since, as noted above, mixing and dilution
processes proceed at a relatively slow rate in such layers. Thus, if pollu-
tants are emitted into an inversion layer, relatively high concentrations can
persist for a considerable period of time or over a considerable distance of
wind travel from the source. For example, a surface inversion in the morning
could cause automotive exhaust pollutants released at the surface during the
morning rush hours to persist with minimum dilution near the ground surface
for an extended period of time, probably for 1 or 2 hours after sunrise, until
solar radiation heats the ground and causes the inversion to disappear or
"break" (Hosier, 1961; Slade, 1968). High concentrations may occur at the
ground even when an elevated inversion is present and the layers below the
inversion are unstable and are undergoing good mixing. Such a persistent
elevated inversion layer is a major meteorological factor that contributes to
high pollutant concentrations and photochemical smog situations along the
southern California coast (Holzworth, 1964; Hosier, 1961; Robinson, 1952).
The vertical mixing profile through the lower layers of the atmosphere
follows a typical and predictable cycle on a generally clear day. In such a
situation a surface inversion would be expected to form during the early
morning and to persist until surface heating becomes significant, probably 2
or 3 hours after sunrise. Pollutants initially trapped in the surface inversion
may cause relatively high, local concentrations, but these concentrations will
decrease rapidly when the surface inversion is broken by surface heating,
usually about midmorning. The surface inversion will begin to form again
019VP/D 4-15 5/2/84
-------
during the early evening hours and pollutants from near-surface sources such
as automobiles will experience progressively less dilution as the surface
inversion develops.
Elevated inversions, when the base is above the ground, are also common
occurrences (Hosier, 1961; Holzworth, 1964). Since these conditions, however,
are identified with specific synoptic conditions, they are much less frequent
than the nighttime radiation inversion. Because it may persist throughout the
day and thus restrict vertical mixing, an elevated inversion is nevertheless a
very significant air pollution feature. Smog-plagued southern California is
adversely affected by persistent elevated inversions (Robinson, 1952). When
compared to a source near the surface and the effects of a radiation (surface)
inversion, the pollutant dispersion pattern is quite different for an elevated
source plume trapped in a layer near the base of an elevated inversion. This
plume will not be in contact with the ground surface in the early morning
hours because there is no mixing downward through the surface radiation inver-
sion. Thus, the elevated plume will not affect surface pollutant concentrations
until the mixing processes become strong enough to reach the altitude of the
plume. At this time, the plume may be mixed downward quite rapidly in a pro-
cess called "fumigation." After this initial mixing, surface concentrations
will decrease as the usual daytime mixing processes continue to develop. If
the daytime mixing becomes strong enough to break the upper inversion, the
pollutants may be mixed through an increasingly deep layer of the atmosphere.
When surface heating decreases in the late afternoon and early evening, both
the surface and elevated inversions will form again. The surface inversion
will again prevent pollutants from elevated sources from reaching the ground
and surface scavenging processes will gradually reduce the concentrations of
pollutants trapped during the formation of the surface inversion.
Geography can have a significant impact on dispersion of pollutants
(e.g., along the coast of an ocean or one of the Great Lakes). Near the coast
or shore, the temperatures of land and water masses can be different, as can
the temperature of the air above such land and water masses. When the water
is warmer than the land, there is a tendency toward reduction in the frequency
of surface inversion conditions inland over a relatively narrow coastal strip
(Hosier, 1961). This in turn tends to increase pollutant dispersion in such
areas. Such conditions may occur frequently on the Gulf Coast. The opposite
condition also occurs if the water is cooler than the land, as in summer or
fall. Cool air near the water surface will tend to increase the stability of
019VP/D 4-16 5/2/84
-------
the boundary layer in the coastal zone, and thus decrease the mixing processes
that act on pollutant emissions. These conditions occur frequently along the
New England coast (Hosier, 1961). Similarly, pollutants from the Chicago area
have been observed repeatedly to be influenced by a stable boundary layer over
Lake Michigan (Lyons and Olsson, 1972). This has been observed especially in
summer and fall when the lake surface is most likely to be cooler than the air
that is carried over it from the adjacent land.
Since the diurnal mixing conditions are such an important part of the
meteorological parameters for understanding pollutant mixing and diffusion, it
is useful to have some knowledge of the mixing cycles that prevail over the
United States. Figures 4-2 and 4-3 show the average summer morning and afternoon
mixing heights as calculated on the basis of upper air temperature data and an
estimated midmorning urban temperature. Since Holzworth (1972) attempted to
include the influence of an urban heat island in this estimated temperature,
the morning results in Figure 4-2 are probably most applicable to larger urban
areas. Rural or nonurban areas would be expected to have lower mixing heights.
Summer conditions are useful to consider because of the prevalence of
high photochemical oxidant concentrations during this season. As shown in
Figure 4-2, morning mixing heights are estimated to be greater than 300
meters except for the central part of the Great Basin, where a 200-meter
isopleth includes parts of Oregon, Idaho, Utah, Arizona, and most of Nevada.
By midafternoon (Figure 4-3), the estimated mixing height at the time of
maximum temperature has increased markedly, and only a few coastal areas have
an average afternoon maximum mixing height of less than 1000 meters. In
contrast to the morning data, the central Great Basin area becomes the area of
greatest mixing in the afternoon. This would be expected since this is a hot,
arid, desert region, and the driving force generating the surface mixing layer
is the solar heating of the ground surface.
The magnitude of the afternoon mixing height is generally an indication
of the potential for recurring urban air pollution problems. If a trapping,
elevated inversion does not rise high enough in the afternoon to release the
generated pollutants that are trapped, an accumulating episode is likely.
From the average summer afternoon data shown in Figure 4-3, where the lowest
average mixing height is 600 meters and almost all of the area has a value
greater than 1500 meters, it would appear that urban air pollution should not
be severe. On the average this is probably correct; however, there are several
019VP/D 4-17 5/2/84
-------
11
Figure 4-2. Isopleths (m x 102) of mean summer morning mixing
heights.
Source: Holzworth (1972).
18
Figure 4-3. Isopleths (m x 102) of mean summer afternoon mixing
heights.
Source: Holzworth (1972).
019VP/D
4-18
5/2/84
-------
departures from the average, which result in relatively low mixing heights and
adverse dispersion over many areas of the United States on a recurring basis.
Figure 4-4 (Holzworth and Fisher, 1979) shows the frequency of occurrence
of elevated inversions in summer having a base between 1 and 500 meters (1600
feet) at the time of the afternoon upper air temperature measurement, 6:10
p.m. EST or 3:15 p.m. PST. The California coastal conditions, in which low
inversions occur with a frequency of nearly 90 percent, are clearly evident.
The northeastern coastal area from New Jersey north to Maine, where cool ocean
water prevails, also has a relatively high percentage, above 20 percent,
compared to most of the rest of the country. Stations bordering one of the
Great Lakes—Green Bay, Sault St. Marie, and Buffalo—reflect a stabilizing
lake effect with percentages above 5 percent. Except along the Pacific Coast,
these ocean and lake coastal situations are probably limited to relatively
narrow coastal zones (Hosier, 1961). Examples are evident in Figure 4-4, in
which it may be noted that inversion frequencies of 21 to 28 percent occur in
coastal New England compared to only 2 percent at Albany in upstate New York.
A similar situation is evident in a comparison of the 3 percent inversion
frequency at Washington, D.C., with the 16 percent frequency on the Delaware
coast. A non-coastal region having summer afternoon low-level elevated inver-
sions more than 5 percent of the time is the Southeast, where an area from
Louisiana and Arkansas to the Atlantic coast shows frequency values between 5
and 10 percent. Other seasons differ in details, but the general patterns are
similar.
This means that, for most of the United States, low-level stable layers
that persist through the afternoon hours are rare events, occurring on less
than 1 day in 20. Thus, air pollution situations in areas such as Kansas or
Iowa will be related to the periods when the expected morning surface inversion
persists later in the morning than usual and when winds are not strong enough
to carry pollutants rapidly away from the local area. Along both the Pacific
Coast and the Northeast Coast, low-level afternoon inversions are frequent
enough to be a significant contributor to local and regional air pollution
episodes.
4.3.2 Wind Speed and Mixing
Another major meteorological factor in the urban pollutant dispersion
problem is low-level or surface-layer wind. As would be expected, strong
019VP/D 4-19 5/2/84
-------
30
22
Figure 4-4. Percentage of summer 2315 GMT {6:15 PM EST, 3:15 PM PST) sound-
ings with an elevated inversion base between 1 and 500 m above ground level.
Source: Adapted from Holzworth and Fisher (1979).
019VP/D
4-20
5/2/84
-------
winds across a source area will dilute pollutant concentrations even though
there is a strong, low-level inversion base. San Francisco is one example of
such a location where strong winds frequently provide good ventilation in
spite of a low inversion. Conversely, light and variable or calm wind condi-
tions over an area can lead to excessive pollutant accumulations even though
the afternoon mixing depth is quite large. Thus, it is necessary to include
wind direction and wind speed frequencies in any evaluation of air pollution
potential for a given area. It must also be recognized that both elevated
inversion conditions and surface wind patterns are governed to a major degree
by the synoptic, or large-scale, weather patterns. Both wind and inversion
factors tend to favor pollutant buildup when a deep, slow-moving high-pressure
system dominates the weather across an area (Korshover, 1967; Korshover,
1975).
Figure 4-5 shows the wind climatology across the United States in the
month of July by depicting the monthly resultant vector wind at major weather
stations (U.S. Department of Commerce, 1968). Note that the flow across the
West Coast is generally directed inland, from west to east. This contributes
to a typical situation for major California cities: significant urban pollu-
tant plumes are found east of the urban core source areas while the immediate
coastline or beach areas are relatively pollutant-free. In the Northeast
States, the average wind flow is from southwest to northeast more or less
parallel to the coastline. As a result, pollutant plumes from the major urban
areas along this coast are frequently additive along the trajectory of the
wind. Polluted air moving toward the coast from major inland urban sources
may also be a factor in this Northeast region. Along the Gulf Coast, the
average winds form southerly, onshore flow. Under some weather situations,
however, there is often an offshore flow in one area (e.g., Texas) and an
onshore flow in an adjacent area. Thus, because of this recirculation, the
onshore Gulf air masses are not always pollutant-free (Price, 1976). Before
the situation was examined carefully, the recirculating pollutants were some-
times confused with natural background concentrations.
Wind climatology provides an average wind flow pattern, but it does not
provide a complete assessment of the influences of the wind on air pollution
dispersion. Wind speed and, in particular, the frequency of weak winds are an
important aspect to be considered. Figure 4-6, adapted from Holzworth and
Fisher (1979), shows the frequency with which early morning (6:15 a.m. EST or
019VP/D 4-21 5/2/84
-------
O
I
ro
ro
CARIBOU
PORTLAND
.BOSTON
SAN FRANCISCO
^ASHV-LLE-^cfi
MPHIS " KNOXVILLE-4
^^\
ATLANTA^*00"^ CHARLE^ON
DALLAS SHREVEPORT JACKSON
MOBILE—TALLAHASSEE "djACKSONVILLE
\ X AUSTIN \ LAKI
\^/-N. \ GALVECTON
iancnn ^r I
RESULTANT WIND IS THE VECTORIAL
AVERAGE OF ALL WIND DIRECTIONS
AND SPEEDS AT A GIVEN PLACE FOR
A CERTAIN PERIOD. AS A MONTH
0 5 10 15 20
I....I....I...J....I
SCALE IN mph
NOTE. BASED ON
HOURLY OBSERVATIONS
1951 1960
en
ro
oo
Figure 4-5. Mean resultant surface wind pattern for the United States for July. Direction and
length of arrows indicate monthly resultant wind.
Source: U.S. Dept. of Commerce (1968).
-------
16
41
Figure 4-6. Percentage of summer 1115 GMT (6:15 AM EST, 3:15 AM PST) sound-
ings with an inversion base at the surface and wind speeds at the surface <2.5
m/sec.
Source: Adapted from Holzworth and Fisher (1979).
019VP/D
4-23
5/2/84
-------
3:15 a.m. PST) surface inversions occur with calm or weak surface winds; that
is, wind speeds equal to or less than 2.5 m/sec or 6 mi/hr. There is consi-
derable variation between stations because terrain and geography (e.g., coastal
locations) influence both the wind flow and inversion frequency. It is clear,
however, that over large areas of the United States, especially in heavily
industrialized inland areas east of the Mississippi River, calm amd stable
summer mornings are a frequent occurrence: 50 percent or more in many areas.
This means that there will be frequent incidents of morning pollutant accumu-
lation; but afternoon heating, as shown by Figure 4-3, will usually mix the
pollutant accumulations through a deep mixing layer and disperse them. Figure
4-7 (Holzworth and Fisher, 1979) shows the average wind speed through the
depth of the summer morning mixing layer. Note that the area east of the Rocky
Mountains, except for the Appalachians, can on the average, expect winds of 4
m/sec (about 10 mi/hr) or higher through the morning mixing layer. This
probably would provide acceptable midmorning dilution of accumulated pollu-
tants. In summer afternoons, as shown in Figure 4-8, the average wind speed
within the mixing layer increases in all areas and may even double over some
of the western mountain states. It should be noted, however, that since winds
normally increase with altitude above the ground, much of the increase in the
average afternoon mixing layer wind is probably the result of the considerable
increase in the depth of the mixed layer, as shown by the differences between
Figures 4-2 and 4-3.
In summary, atmospheric mixing parameters of stability and wind in the
pollutant transport layers can exert controlling effects on 0~ and oxidant
concentrations. The effects include the amount of dilution occurring in the
source area, as well as along the trajectory followed by an urban or source-
area plume. Regions having steady prevailing winds, such that a given air
parcel can pass over a number of significant source areas, can develop signifi-
cant levels of pollutants even in the absence of weather patterns that lead to
the stagnation type of air pollution episodes. The Northeast states are
highly susceptible to pollutant plume transport effects, although some notable
stagnation episodes have also affected this area (e.g., Lynn et al., 1964).
Along the Pacific Coast, especially along the coast of California, coastal
winds and a persistent low inversion layer contribute to major pollutant
buildups in urban source areas and downwind along the urban plume trajectory
(Robinson, 1952; Neiburger et al., 1961). In the southern Appalachians, the
019VP/D 4-24 5/2/84
-------
Figure 4-7. Isopleths (m/sec} of mean summer wind speed averaged
through the morning mixing layer.
Source: Holzworth and Fisher (1972).
Figure 4-8. Isopleths (m/sec) of mean summer wind speed averaged
through the afternoon mixing layer.
Source: Holzworth and Fisher (1972).
019VP/D
4-25
5/2/84
-------
weather favors longer-term air pollution episodes (Korshover, 1967; Korshover,
1975). Generally, low pollution potential results from the conditions that
occur in the Great Plains area and south to the Texas-Louisiana Gulf Coast;
and between the Mississippi River and the crest of the Rocky Mountains.
4.3.3 Effects of Sunlight and Temperature
The significance of sunlight in photochemistry is related to its intensity
and its spectral distribution, both of which have direct effects upon the
specific chemical reaction steps that initiate and sustain oxidant formation.
Sunlight intensity varies with season and geographical latitude but the latter
effect is strong only during the winter months. During the summer, the maximum
light intensity is fairly constant throughout the contiguous U.S. and only the
duration of the solar day varies to a small degree with latitude.
The effects of light intensity on individual photolytic reaction steps
and on the overall process of oxidant formation have been studied in the
laboratory (Peterson, 1976; Demerjian et al., 1980). All of the early studies,
however, employed constant light intensities, in contrast to the diurnally
varying intensities that occur in the ambient atmosphere. More recently,
diurnal variation of light intensity has been recognized and studied as a
factor in photochemical oxidant formation (Jeffries et al. , 1975; Jeffries
et al., 1976). Such studies showed that the effect of this factor varies with
initial reactant concentrations. Most important was the observation that
similar NMOC/NO systems showed different oxidant potential depending on
/\
whether studies of these were conducted using constant or diurnal light. This
has led to incorporation of the effects of diurnal or variable light into
photochemical models (Tilden and Seinfeld, 1982).
While the effect of sunlight intensity is direct and has been amply
demonstrated (Leighton, 1961; Winer et al., 1979), the effect of wavelength
distribution on the overall oxidant formation process is subtle. Experimental
studies have shown the photolysis of aldehydes to be strongly dependent on
radiation wavelength in the near UV region (Leighton, 1961); and there is some
indication (Bass et al., 1980) that the photolysis rates for aldehydes may be
temperature-dependent. Since aldehydes are major products in the atmospheric
photooxidation of NMOC/NO mixtures, it is inferred that the radiation wavelength
should have an effect on the overall photooxidation process. This inference was
directly verified, at least for the propylene/NO and n-butane/NO chemical
/\ ^~ }\
019VP/D 4-26 5/2/84
-------
systems, in smog chamber studies (Jaffee et al., 1974; Winer et al., 1979).
In the ambient atmosphere, some variation in the wavelength distribution of
sunlight does occur as a result of variations in time of day, stratospheric
0-, ambient aerosol (Stair, 1961), and cloud cover.
It has been observed that days on which significant ozone-oxidant con-
centrations occur are usually days with warm, above-normal temperatures (Bach,
1975). This temperature effect can be explained readily as a synoptic meteo-
rological correlation rather than as a temperature-photochemical rate constant
effect, in that periods of clear skies and warm temperatures are periods of
high air pollution potential, as discussed above. Because of the close corre-
lation between above-normal temperatures and high air pollution potential, a
maximum daily temperature forecasting procedure is often useful as a substi-
tute for a more elaborate and specific program for forecasting air pollution
potential. The correlation between temperature and, thus, synoptic weather
conditions and photochemical air pollution intensity has been observed in a
number of areas. Evaluation of photochemical air pollution in Los Angeles as
early as 1948 showed a correlation with temperature. Recent studies of 0,
patterns in St. Louis, Missouri, have also shown a correspondence between
daily maximum 0- concentration and temperature (Shreffler and Evans, 1982).
4.3.4 Transport of Ozone and Other Qxidants and Their Precursors
The 1978 air quality criteria document for ozone and other photochemical
oxidants made a convincing case for the fact that ozone and other photochemical
oxidants are transported from urban source areas to downwind regions in concen-
trations approaching 0.1 ppm or more (U.S. Environmental Protection Agency,
1978a). This was a significant advance in the understanding of photochemical
air pollution. It served to answer a number of perplexing problems that had
been identified previously in studies of ozone and other photochemical oxi-
dants in nonurban areas. These included high ozone or total oxidant concen-
trations in areas remote from identifiable sources. The 1978 criteria document
also pointed out that the evaluation of impacts on local ozone or total oxidant
concentrations resulting from transport into the area was still only a quali-
tative assessment and could not then be quantified.
There have been several recent extensive studies on oxidant transport for
model development and on field measurements of oxidants for model verifica-
tion. One such program was the Northeast Regional Oxidant Study (NEROS),
019VP/D 4-27 5/2/84
-------
which was carried out in 1979 and 1980 in the corridor from Washington, D.C.,
to Boston (Clarke et al. , 1982). Reactive pollutant modeling for urban areas,
particularly the Empirical Kinetic Modeling Approach (U.S. Environmental
Protection Agency, 1981), has progressed rapidly in recent years.
Studies of the transport of ozone and other photochemical oxidants (0~-0 )
O s\
are classified into three regimes, depending upon transport distance (U.S.
Environmental Protection Agency, 1978). In the first, urban-scale transport,
the occurrence of transport of photochemical pollutants can be detected in
most large urban areas if there is sufficient 0--0 monitoring information.
O /\
It has been identified as a significant, characteristic feature of the 0--0
«5 X
problem in the Los Angeles basin (Tiao et al., 1975), as well as in San Fran-
cisco, New York, Houston, Phoenix, and St. Louis (Altshuller, 1975; Coffey and
Stasiuk, 1975; Shreffler and Evans, 1982). Urban-scale transport patterns
result from one or more of a combination of factors. First is the simple
advection of the photochemically reacting air mass and the development of
maximum 0~-0 after 1 or 2 hours of downwind travel. Maximum concentrations
O )\
may be displaced up to 20 or so miles from the center of the major source
area. It has also been noted (U.S. Environmental Protection Agency, 1978a)
that pollutant concentrations in air parcels in the central core area of major
source areas may not be the most conducive for Q.,-0 formation because of the
0 A
tendency toward occurrence there of more effective scavenging, especially
scavenging related to NO and its reactions.
The distance of the peak 0^-0 concentrations from the urban core area is
O /\
dependent on the local wind pattern and is, in general, inversely related to
the peak 0^-0 concentration. Stronger winds will carry the air parcels
O /\
farther during the reaction period, increasingly diluting pollutant concentra-
tions along the trajectory. Weak winds and very restricted mixing heights
will tend to cause higher 0,-0 concentrations closer to the central source
O /\
area. The diurnal wind cycle will also be an important factor, since in some
situations calm conditions may prevail until late in the morning but in others
a steady wind may be present throughout the emission and reaction process.
The second, or mesoscale, kind of transport of 0*3-0 is in many respects
an extension of the urban-scale transport and is characterized by urban plume
development. A report by Bell (1960) described November 1959 03-0x incidents
in northern coastal San Diego County, California. It showed conclusively that
these were caused by the 0--0 and precursors formed and emitted, respectively,
O /\
019VP/D 4-28 5/2/84
-------
the previous day in the Los Angeles basin. The transport in these situations
was over the coastal Pacific Ocean, and the 0,-0 arrived at the San Diego
«j X
receptor site as a contaminated sea breeze after overnight travel (Bell,
1960).
In the 1978 0,,-0 criteria document, more than 30 references were cited
*J A
relating to urban plume observations and investigations. Since 1978, the
results of the 1975 New England oxidant study have been published in detail,
and results of a more comprehensive 2-year field program carried out along the
Washington, D.C.-Boston corridor in 1979 and 1980 have appeared in the litera-
ture (Clark and Clarke, 1982; Clarke et al., 1982; Vaughan et al., 1982). A
major field program supported by local industries was conducted in Houston,
Texas, although the 0--0 downwind plume phases were not as extensive as in
O f\
NEROS. Chicago and adjacent shoreline areas of Lake Michigan have also been
subject to a number of ground-level and airborne studies over distances of 150
to 300 kilometers (Lyons and Olsson, 1972; Sexton and Westberg, 1980; Sexton
et al., 1981). As described above, 0,-0 plumes from major urban areas can
O }\
extend about 100 to 200 miles with widths of tens of miles (Sexton, 1982) and
frequently up to half the length of the plume. Other field studies conclusively
demonstrating mesoscale transport over New England have been reported (Spicer
et al., 1979; Clarke et al., 1982; Cleveland et al., 1976a; Cleveland et al.,
1976b; Rubino et al., 1976; Westberg et al., 1976; and Westberg et al., 1978).
Although urban plumes are frequently thought of as a problem related only to
large source areas such as New York and other major metropolitan areas, measure-
ments in plumes from smaller urban areas have shown that these sources cannot
be ignored (Sticksel et al., 1979; Sexton, 1982; and Spicer et al. , 1982).
In the third kind of pollutant transport, synoptic-scale, the transport
of 0--0 and precursors is characterized by the general and widespread elevated
O /\
concentrations of pollutants that can occur on an air-mass scale under certain
favorable weather patterns. These weather situations are generally slow-moving,
well-developed high-pressure, or anti-cyclonic systems. This type of deep
high-pressure area was considered by Korshover (1967, 1975) as a prerequisite
for stagnating air pollution episodes. Not all high-pressure systems lead to
high air pollution potential, however. Strong surface highs are frequently
found in conjunction with well-developed low-pressure storm systems, resulting
in brisk winds and good mixing, and, thus, low air pollution potential. Another
type of high-pressure system, however, is one in which the surface high-pressure
019VP/D 4-29 5/2/84
-------
area is reinforced by a warm high-pressure circulation in the upper air. This
is frequently characterized by weak winds, stable surface layers, and high
pollution potential over regional or air-mass-sized areas. This is the meteo-
rological pattern that is involved in synoptic-scale pollutant transport
(Korshover, 1967; Korshover, 1975).
The importance of synoptic-scale or air-mass pollutant situations has
been recognized for many years, probably much longer than the importance of
major plumes has been apparent. The Donora, Pennsylvania smog episode in
1948 (Schrenk et al., 1949) involved the occurrence over a wide area of a
regional air mass having relatively high pollution levels simultaneous with
the occurrence of a stagnated warm high-pressure area over the Ohio Valley and
the northern Appalachian area. Donora was an especially adversely affected
pocket within this larger system.
The synoptic-scale high-pressure air pollution system is not charac-
terized by well-defined urban plumes. Rather, a warm, slow-moving or stagnant
anti-cyclone provides a synoptic-scale weather system that, because of weak
winds and limited vertical mixing, favors the accumulation of relatively high
concentrations of air pollutants. On a climatological basis, these systems
are most common in the summer and fall months over most of the United States,
as shown by the work of Korshover (1967; 1975). The general track Of a system
is from west or southwest to east or northeast. In many cases, an anti-cyclone
will stagnate and intensify over the Midwest or East as circulation patterns
in the upper air change and become more supportive of the surface anti-cyclonic
pattern (Schrenk et al., 1949; Lynn et al., 1964).
Along the West Coast, air pollution problems are also the result of
persistent high-pressure system influences. In this case, however, the high
is the persistent subtropical anti-cyclone of the eastern Pacific rather than
the series of transitory anti-cyclone systems characteristic of the area east
of the Rocky Mountains. The persistent or semipermanent subtropical anti-
cyclone in the Pacific is linked to the large-scale general circulation of the
atmosphere rather than to moving wave systems (Neiburger et al., 1961). The
effect is much the same, except that the area of limited mixing and more
adverse air pollutant effects is found on the eastern edge of the subtropical
anti-cyclone rather than the trailing western edge as in the transitory sys-
tems.
019VP/D 4-30 5/2/84
-------
The identification and understanding of photochemical O.-O and precursor
•3 X
transport by weather systems has provided a significant advance in under-
standing photochemical air pollution and the potential extent of its effects.
Considerable progress has been made in the development of long-range photochem-
ical modeling techniques so that the likely impact of synoptic systems can be
anticipated. Such tools are very much in the research stage because the local
impact of 0,-0 results from a complex interaction of distant and local sources,
o X
urban plumes, mixing processes, atmospheric chemical reactions, and general
meteorology.
4.3.5 Surface Scavenging in Relation to Transport
A major scavenging process for 0- in the atmospheric boundary layer is
adsorption and subsequent destruction at the ground surface. This occurs
through the process of dry deposition, in which the process of eddy diffusion
moves air parcels downward through the turbulent boundary layer to the laminar
sub-layer. Individual molecules, such as ozone, will then move by Brownian
motion through this laminar layer to the underlying surface. There reactive
molecules, such as ozone, can be removed from the layer by reactions at the
surface. These reactions can maintain a vertical concentration gradient, with
the lowest concentrations occurring at the surface of the ground because of sur-
face-scavenging reactions.
Because of this surface-scavenging process, ozone will persist in an
atmospheric parcel in the absence of ozone-forming reactions only if the
parcel is dispersed such that contact with the ground surface is minimized.
It is likely that only in those air parcels moving above the surface layer
will ozone escape the surface reactions and persist long enough to undergo
long-distance transport. Aircraft observations have documented frequently the
occurrence of relatively high ozone concentrations above lower-concentration
surface layers. This is a clear indication that ozone is essentially pre-
served in layers above the surface and can be transported over relatively long
distances when continual replenishment through precursor reactions is not a
factor, such as at night.
The fact that 0- is formed in the stratosphere, mixed downward, and
4.3.6 Stratospheric-Tropospheric Ozone Exchange
The fact that 0- is formed in the stratos;
incorporated into the troposphere, where it forms a more or less uniformly
019VP/D 4-31 5/2/84
-------
mixed background concentration, has been known in various degrees of detail
for many years (Junge, 1963). For example, one of the early questions re-
solved in studies of Los Angeles smog was to eliminate stratospheric ozone as
a cause of the elevated ozone concentrations. For other areas and different
sets of conditions, however, the importance of stratospheric ozone as a signi-
ficant component of ground-level ozone cannot be ruled out.
The mechanisms by which stratospheric air is mixed into the troposphere
have been examined by a number of authors. Danielsen conducted extensive
analyses of major synoptic weather events that injected stratospheric air into
the troposphere (Danielsen, 1968; Danielsen and Mohen, 1977; Danielsen, 1980).
Reiter has been especially active in describing the atmospheric mechanisms by
which stratospheric air injection takes place and in relating these processes
to the global circulation of the atmosphere (Reiter, 1963; Reiter and Mahlman,
1965; Reiter, 1975). As a result of such research, exchange between the
stratosphere and troposphere in the middle latitudes has been determined to
occur to a major extent in events called "tropopause folds." In a tropopause
fold (TF), the jet stream in the upper troposphere plays a major role in
directing stratospheric air and high ozone concentrations into the troposphere.
Figure 4-9 is a schematic presentation of the intrusion process as described
by Danielsen (1968). The subsidence occurs along the poleward side of the
polar jet stream in the area where the jet is associated with a cold front at
ground level. The result is downward transport in the cold air behind the cold
front.
Since 1978, a considerable amount of research on TF and ozone injection
has been done, especially by SRI-International (Johnson and Viezee, 1981;
Ludwig et al., 1977; Singh et al., 1980; and Viezee et al., 1979). Figure 4-10
from Johnson and Viezee (1981) shows one example of the probing by SRI of a TF
event in the midwestern United States. Concentrations of ozone in excess of
90 ppb were found as low as 13,000 feet (3.9 kilometers), as shown in the upper
part of Figure 4-10. These authors found that ozone intrusion was lower during
this fall study (October 5, 1978) than in a number of other spring TF events.
The dew point measurements in the second part of the figure confirm the stra-
tospheric injection. The weather pattern accompanying this TF is shown at the
bottom of Figure 4-10 by a 500 millibar (about 6 kilometer) chart; the surface
cold front is also indicated. Note that the intrusion was detected well behind
the cold front and appears to have assumed a layered formation in the altitude
range of 8,000 to 12,000 feet (2.4 to 3.6 kilometers).
019VP/D 4-32 5/2/84
-------
.•?*5^'JrF--'--v"^:\r*V' LI* • TV.;.;-:
Figure 4-9. Schematic cross section, looking downwind along the jet
stream, of a tropopause folding event as modeled by Danielsen (1968).
Source: Johnson and Viezee (1981).
019VP/D
4-33
5/2/84
-------
-------
From their analysis of measurement flights in a number of TF situations,
Johnson and Viezee (1981) concluded that the ozone-rich intrusion sloped
downward toward the south. In terms of dimensions, the average crosswind
width (north to south) at an altitude of 5.5 kilometers (18,000 feet) for six
spring intrusions averaged 226 kilometers (746,000 feet), and for four fall TF
systems, 129 kilometers (426,000 feet). Ozone concentrations at 5.5 kilometers
(18,000 feet) averaged 108 ppb in the fall systemsand 83 ppb in the spring
systems. Previously it had been assumed that only a few fairly intense systems
would produce a TF event and trans-tropopause mixing. From their data, however,
Johnson and Viezee (1981) drew the very important conclusion that all low-
pressure trough systems, such as illustrated in Figure 4-10, may induce a TF
event and cause the trans-tropopause movement of ozone-rich air into the tro-
posphere.
On the basis of their field studies and the earlier models and work of
Danielsen (1968), Johnson and Viezee (1981) proposed a set of model mechanisms
or types of TF injection, which are illustrated in Figure 4-11 and described
in the following general manner:
1- Type 1. The intrusion is broken up and dispersed by mixing and
diffusion in the middle or free troposphere.
2. Type 2. The intrusion persists down to the planetary boundary
layer or the top of the mixed layer, where the lower
part of the intrusion may be incorporated into the mixed
layer and may subsequently reach the ground.
3. Type 3. The intrusion occurs close behind the cold front, where
the air parcels are caught by the downdrafts behind the
cold front; and is brought to the ground by direct
circulations associated with the front.
4. Type 4. The ozone-rich parcels are incorporated into convective
cells and brought to the ground in association with
rain-showers and thunderstorm downdrafts; similar to
Type 3.
Johnson and Viezee (1981) summarize the possible impacts of these four
types of TF events by noting that Types 1 and 2 should produce "relatively
moderate effects" at the ground in comparison to those to be expected from
Types 3 and 4. The latter two could cause "substantial" effects in terms of
high surface ozone concentrations. The action of Types 3 and 4 are supported
019VP/D 4-35 5/2/84
-------
-o
I
CO
CTi
o" ® •"*
§• 3 i
M W <
EH
Q. 3-W
(D N
< r+ W
I I 3
» 8^
"T Ti (D
(O 3" JJ-
2 2 o
, (D -t«
(D
en
ro
oo
O
O
(D
-------
by meteorological theory (Bjerknes, 1951) and by observations of surface ozone
such as those made by Dam"el sen and Mohnen (1977), Lamb (1977), and Davis and
Jensen (1976).
4.3.7 Stratospheric Ozone at Ground Level
After a detailed review of background tropospheric ozone, Viezee and
Singh (1982) came to a number of important conclusions. First, as also con-
cluded from earlier information, the stratosphere is a major but not the sole
source of background ozone in the unpolluted troposphere. This stratospheric
ozone is brought to the surface mixed layer by vertical mixing processes in
the atmosphere that have been known for many years. In the northern hemisphere,
between 30°N and 50°N, background surface ozone concentrations reach a maximum
of 55 to 65 ppb. In the tropics, lower, 15 to 30 ppb, concentrations prevail.
Viezee and Singh (1982) also noted that the stratospheric ozone reservoir has
a strong seasonal variation, with a maximum in the spring and a minimum in
fall and winter months, especially at middle latitudes. This seasonal cycle
is reflected at ground-level background observation stations, where the average
spring background ozone is generally in the range of 50 to 80 ppb and the
average fall value is between 20 and 40 ppb. In the troposphere, concentra-
tions generally increase gradually to the tropopause, but the seasonal pattern
is the same.
Viezee and Singh (1982) concluded that relatively high ozone concentra-
tions can occur for short periods of time, minutes to a few hours, over local
areas as a result of stratospheric intrusions. They were able to document
from published literature ten situations of probable intrusion of stratospheric
ozone. These instances are shown in Table 4-1, reproduced from Viezee and
Singh (1982). Note that all of the short-term situations in which peak concen-
trations exceeded 80 ppb occurred in winter and spring months and not in the
photochemically active summer season. Of the three summer instances that were
reported, two at Whiteface Mountain, New York, and one at Pierre, South Dakota,
the highest reported concentration was 56 ppb for a 1-hr average.
A number of of stratospheric 0- intrusions into the troposphere in Great
Britain have been reported by Derwent et al. (1978). The results are noted as
Case 4 in Table 4-1. The British measurements, made at Harwell, produced
results similar to the United States TF data, indicating a general tropospheric
03 background of 20 to 50 ppb based on hourly average concentrations, with
019VP/D 4-37 5/2/84
-------
TABLE 4-1. DOCUMENTED EPISODES OF TRANSPORT OF STRATOSPHERIC OZONE TO GROUND LEVEL
co
Case
no.
1
2
3
4
5
9
4
10
6
7
8
Date
3 March 1964
26 February 1971
19 November 1972
6 March 1974
8,9 January 1975
11, 12 July 1975
19 March 1977
24, 25, 28 June and
1 July 1977
4 March 1978
July 1978
15 March 1978
Geographic location
Quincy, Florida (near Tallahassee)
Observatory Hohenpeissenberg
(1000 m MSL), SW of Munich, Germany
Santa Rosa, California
Harwell, Oxfordshire, England
Zugspitze Mountain, near Garmisch-
Partenkirchen, Germany (3000 m MSL)
Whiteface Mountain, New York
(1150 m MSL)
Sibton, Suffolk, England
Whiteface Mountain, New York
Denver, Colorado
Pierre, South Dakota
Kisatchie National Forest, Louisiana
Ground-level 03
concentration, ppb
100 to 300
415
250
200 to 230
110 to 115
160 to 193
< 37
100 to 110
< 47
82
< 56
< 46
100 to 105
Duration of
observed event
3 hr
10 min
50 min
1 hr
2 hr
4 hr
24-hr average
2 hr
24- hr average
1 hr
1 hr
24-hr average
2 hr
Length of data
record examined
July 1963 through
July 1973
December 1970 through
March 1971
November 1972
4 to 5 yr discontinuous
August 1973 through
February 1976
July 1975
4 to 5 yr discontinuous
June and July 1977
1975 to 1978
July through September
1978
Spring 1978
Source
Davis and Jensen (1976)
Atmannspacher and
Hartmannsgruber (1973)
Lamb (1977)
Derwent et al. (1978)
Singh et al. (1980)
. Husain et al. (1977)
Derwent et al . (1978)
Dutkiewicz and Husain
(1979)
Haagenson et al. (1981)
Kelly et al. (1981)
Viezee et al. (1982)
Source: Viezee and Singh, 1982.
-------
occasional short-period peaks of 100 ppb or so. Derwent et al. (1978) also
made the statement, "These sporadic episodes appear to be observable
only in rural areas and make no contribution to the exposure levels in urban
areas" (Derwent et al., 1978). This comment, however, is contrary to the con-
clusions drawn by a number of investigators in the United States about the
impact of ozone from natural sources, including TF events or general stra-
tospheric injection, on ozone concentrations in urban areas. Some of the
researchers who have come to such a conclusion include Coffey and Stasiuk
(1975a), Coffey and Westberg (1977), and Coffey et al. (1977). On the other
hand, Prior et al. (1981), using 9 a.m. ozone concentrations in St. Louis as
input for a daily maximum ozone forecasting scheme, concluded that the concen-
trations they found in the morning may have represented transported rather
than local ozone.
The downward transfer of air parcels and ozone from the
stratosphere into the troposphere has been described above. There is, of
course, a compensating transfer of tropospheric air upward into the lower
stratosphere. Reiter (1975) has examined various mechanisms that contribute
to this transfer. Air parcels moving out of the troposphere will carry with
them the background concentrations of ozone that they had in the troposphere,
and, as the air parcels mix in the stratosphere, these ozone molecules will
become part of the stratospheric background ozone. Since the ozone concentra-
tions are very much lower in the troposphere compared to the stratosphere,
however, this exchange of tropospheric and stratospheric air parcels will not
result in a net upward transport of ozone.
4.4 SUMMARY
The photochemistry of the polluted atmosphere is exceedingly complex, but
an understanding of the basic phenomena is not difficult to acquire. Three
processes occur: the emission of precursors to ozone, from (predominantly)
manmade sources; photochemical reactions that take place during the dispersion
and transport of these precursors; and scavenging processes that reduce the
concentrations of both 03 and precursors along the trajectory. Because trans-
port and dispersion of the precursors determine the ambient concentrations
ozone may finally reach, an understanding of certain meteorological phenomena,
in addition to photochemical reactions, is also necessary. These latter are
discussed first, followed by a presentation of important meteorological factors.
019VP/D 4-39 5/2/84
-------
In the troposphere, 03 is formed indirectly through the action of sunlight
on nitrogen dioxide (N02). In the absence of competing reactions, a steady-
state or equilibrium concentration of 0- is soon established between the 0,,
N02, and NO (nitric oxide). The injection of organic compounds (hydrocarbons)
into the atmosphere upsets the equilibrium and allows the ozone to accumulate
at much higher than steady-state concentrations. Recent work on the photo-
chemistry of smog has demonstrated fairly conclusively that the hydroxyl radical,
HO-, is the key species in causing organic compounds to play a major role in
smog reactions.
The length of the induction period before the accumulation of 0- begins
depends heavily on the initial concentration of HO radicals. There is evidence
that nitrous acid (MONO), which is a good source of HO radicals, occurs in the
atmosphere, but at very low concentrations. The most important source of HO',
however, appears to be aldehydes, which are constituents of automobile exhaust,
as well as decomposition products of most atmospheric photochemical reactions
involving hydrocarbons.
The occurrence of organic compounds and sunlight does not mean that the
photochemical reactions will continue indefinitely. Terminating reactions
gradually remove NOp from the reaction mixtures, such that the photochemical
cycles would slowly come to an end unless fresh NO and N0? emissions were
injected into the atmosphere. Besides ozone, other oxidants that contain
nitrogen, such as peroxyacetyl nitrate (PAN), nitric acid (HNO-), and pero-
xynitric acid (HNO.), as well as organic nitrates and inorganic nitrates, are
some of the terminating compounds.
The maximum concentration that 0- can reach in polluted atmospheres appears
to depend on the hydrocarbon-nitrogen oxides ratio. At a low ratio (1:1 to 2:1),
insufficient HO- radicals are available from the hydrocarbon species to effect
the conversion of NO to NOp, a necessary first step. At high ratios (greater
than 12:1 to 15:1), conversion of NO to NOp occurs rapidly, but the termina-
ting reactions remove NOp from the reaction cycles and 0- cannot build up to
high concentrations. Only at intermediate ratios (4:1 to 10:1) are conditions
favorable to the formation of appreciable concentrations of 03-
Recent studies on the fundamental photochemistry of organic compounds
have been reasonably successful. The reactions of paraffinic compounds are
fairly well understood, as are those of olefinic compounds. Photochemical
reactions of the aromatic compounds, however, are poorly understood.
019VP/D 4-40 5/2/84
-------
Natural hydrocarbons (i.e., those organic compounds emitted from vegetation)
as well as hydrocarbons from manmade sources can react photochemically with
nitrogen oxides to yield 0~, although natural hydrocarbons seem to be mainly
scavengers of 0- rather than producers of 0~.
Besides direct adverse effects on human health and on vegetation, 0- con-
tributes to visibility degradation and to acidic deposition. Through its
photolysis by sunlight, with subsequent generation of HO radicals, ozone par-
ticipates only indirectly, but not insignificantly, in the formation of both
sulfate and nitrate aerosols, which cause reduced visibility. These sulfate
and nitrate species, on further reaction, result in acidic precipitation.
Meteorological processes are quite important in determining the extent to
which 0, precursors can accumulate, and thereby the concentration of 0~ which
can result. Atmospheric mixing depends principally on the amount of turbulent
mixing, wind speed and direction, or both. Geography can have a significant
impact, also, particularly at land-sea interfaces.
The degree of turbulent mixing can be characterized by atmospheric stabi-
lity. Pollutants do not spread rapidly in stable layers, nor do they mix
upwards rapidly through stable layers to higher altitudes. Rather, stable
layers are usually characterized by temperature inversions, in which tempera-
ture increases with increasing altitude. Since pollutants emitted below or
into an inversion layer will not readily mix across the inversion layer, they
may persist for a considerable time and distance until the inversion is broken,
usually by surface heating resulting from sunlight.
The extent to which surface heating can cause mixing heights to increase
(and to cause dilution of 0, and its precursors) is highly dependent on geo-
graphy. Along both the Pacific Coast and the Northeast Coast, as well as near
the Great Lakes, low-level inversions (i.e., the mixing height is not great)
frequently persist through the afternoon, making these areas prone to local
and regional air pollution episodes.
Wind speed and direction determine the extent to which pollutants can be
increased by passing over successive sources, or can be diluted by being
rapidly removed from the source area. The plumes of precursors and resulting
0- from large metropolitan areas have been shown to persist for hundreds of
miles. Three kinds of transport of ozone and other pollutants have been
described, in terms of transport distance. In urban-scale transport, maximum
concentrations of 0, are produced about 20 miles or so (and about 2 to 3
019VP/D 4-41 5/2/84
-------
hours) downwind from the major pollutant source areas. In mesoscale transport,
0- has been observed up to 200 miles downwind from the sources of its precur-
sors. Synoptic-scale transport is associated with large-scale, high-pressure
air masses that may extend over and persist for many hundreds of miles.
The significance of sunlight in photochemistry is related to its intensity
and its spectral distribution, both of which have direct effects on the speci-
fic chemical reaction steps that initiate and sustain oxidant formation. Days
on which significant ozone-oxidant concentrations occur are usually days with
warm, above-normal temperatures. These are also characteristic of high pres-
sure systems with inversions and low winds. The photolysis of aldehydes is
affected by the spectral distribution of light, since it is strongly dependent
on wavelength in the near ultraviolet region.
Ozone formed in the stratosphere can be brought downwards to the earth's
surface by events called "tropopause folds." These events are most commonly
observed in the mid-latitudes during spring and early summer. Relatively high
concentrations of 0~ can occur for short periods of time, minutes to a few
hours, over local areas.
019VP/D 4-42 5/2/84
-------
4.5 REFERENCES
Altshuller, A. P. (1983) Review: Natural volatile organic substances and
their effect on air quality in the United States. Atmos. Environ.
17(11):2131-2165.
Altshuller, A. P. (1975) Evaluation of oxidant results at CAMP sites in the
United States. J. Air Pollut. Control Assoc. 25:19-24.
Altshuller, A. P. and Bufalini, J. J. (1971) Photochemical aspects of air
pollution: A review. Environ. Sci. Technol. 5:39-64.
Appel, B. R.; Kothny, E. L.; Hoffer, E. M.; Wesolowski, J. J. (1980) Sulfate
and nitrate data from the California Aerosol Characterization Experiment
(ACHEX). Adv. in Environ. Sci. and Technol. Vol. 9. G. M. Hidy; P. K.
Mueller; D. Grosjean; B. R. Appel; J. J. Wesolowski, eds. pp. 315-335.
Arnts, R. R. and Gay, B. W. (1979) Photochemistry of some naturally emitted
hydrocarbons. EPA-600/3-79-081. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina.
Arnts, R. R. ; Gay, B. W.; and Bufalim", J. J. (1981) Photochemical oxidant
potential of the biogenic hydrocarbons. "Atmospheric Biogenic Hydrocar-
bons." Bufalini, J. J.; Arnts, R. R., eds. Vol II. Ann Arbor Science
Publishers, Ann Arbor, Michigan, p. 117.
Atkinson, R. ; Aschmann, S. M. ; Carter, W. P. L.; Winer, A. M.; and Pitts, J.
N., Jr. (1982) Kinetics of the reactions of OH radicals with n-alkanes at
299+2K. Int. J. Chem. Kinetics. 14:781-788.
Atkinson, R. ; Carter, W. P. L.; Darnall, K. R.; Winer, A. M.; Pitts, J. N.,
Jr. (1980) A smog chamber and modeling study of the gas phase NO -air
photooxidation of toluene and the cresols. Int. J. Chem. Kinei.
12:779-836.
Atkinson, R. ; Darnall, K. R. ; Lloyd, A. C.; Winer, A. M.; and Pitts, J. N. ,
Jr. (1979) Kinetics and mechanisms of the reactions of the hydroxyl
radical with organic compounds in the gas phase. Adv. Photochem. 11:375-
488.
Atkinson, R. ; Lloyd, A. C. (1984) Evaluation of kinetic and mechanistic data
for modeling of photochemical smog. J. Phys. Chem. Ref. Data. [In
Press].
Attmanspacher, W.; Hartmannsgruber, R. (1973) On extremely high values of ozone
near the ground. Pure Appl. Geophys. pp. 106-108, 1091-1096.
Bach, W. D., Jr. (1975) Investigation of ozone and ozone precursor concentra-
tions in nonurban locations in the eastern United States. Phase II,
Meteorological Analysis. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina. Publication No. EPA-450/3-74-034a.
019LEX/B 4-43 5/7/84
-------
Bass, A. M. et al. (1980) Temperature-dependent absorption cross sections for
formaldehyde (CH20): The effect of formaldehyde on stratospheric
chlorine chemistry. Accepted by Planetary and Space Science.
Bell, G. B. (1960) Meteorological conditions during oxidant episodes in coastal
San Diego County in October and November 1959. State of California,
Department of Public Health, Sacramento, California.
Bjerknes, J. (1951) Extratropical cyclones. In: Compendium of Meteorology.
Boston; American Meteorological Society, pp. 557-598.
Blumenthal, D. L.; Richards, L. W.; Macias, E. S.; Bergstrom, R. W.; Wilson,
W. E.; Bhardwaja, P. S. (1981) Effects of a coal-fired power plant and
other sources on southwestern visibility. Atmos. Environ. 15:1955-1969.
Calvert, J. G. and Mohnen, V. (1983) The chemistry of acid formation. j.n:
National Academy of Sciences. Acid Deposition: Atmospheric Processes
in Eastern North America. Appendix A. Washington, D.C.; National
Academy Press.
Calvert, J. G. and Stockwell, W. R. (1983) The mechanism and rates of the
gas-phase oxidations of sulfur dioxide and nitrogen oxides in the atmos-
phere. In: Acid Precipitation: S02, NO, and N02 Oxidation Mechanisms;
Atmospheric Considerations. Ann Arbor, MI; Ann Arbor Science Publishers.
Carter, W. P.; Lloyd, A. C.; Sprung, J. L.; Pitts, J. N., Jr. (1979) Computer
modeling of smog chamber data: progress in validating detailed mechanisms
for the photooxidation of propene and rrbutane in photochemical smog.
Int. J. Chem. Kinetics 11:45-101.
Carter, W. P.; Winer, A. M.; Pitts, J. N., Jr. (1981) Major atmospheric sink
for phenol and the cresols. Reaction with the nitrate radical. Environ.
Sc. Techno!. 15:829-831.
Clark, T. L.; Clarke, J. F. (1982) Boundary layer transport of NO and 03 from
Baltimore, Maryland. A case study. Paper 8.5, Third Conference on Air
Pollution Meteorology, American Meteorological Society, January.
Clarke, J. F.; Ching, J. K. S.; Brown, R. M.; Westberg, J.; White, J. H.
(1982) Regional transport of ozone. Paper 1.1, Third conference on Air
Pollution Meteorology, American Meteorological Society, January.
Cleveland, W. S.; Guarino, R.; Kleiner, B.; McRae, J. E.; Warner, J. L. (1976a)
The analysis of the ozone problem in the northeast United States. In:
Specialty Conference on Ozone/Oxidants-Interactions with the Total
Environment. Pittsburgh, Air Pollut. Control Assoc. pp. 109-120.
Cleveland, W. S.; Kleiner, B.; McRae, J. E.; Warner, J. L. (1976b) Photochemical
air pollution: Transport from New York City area into Connecticut and
Massachusetts. Science 191:179-181.
Coffey, P. E.; Stasiuk, W. N. (1975) Evidence of atmospheric transport of ozone
into urban areas. Environ. Sci. Techno!. 9:59-62.
019LEX/B 4-44 5/7/84
-------
Coffey, P. E. ; Stasiuk, W. N.; Mohnen, V. (1977) Ozone in rural and urban
areas of New York State. In: International Conference on Photochemical
Oxidant Pollution and Its Control. Proceedings. Volume I. B. Dimitriades,
ed. U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina. January. Publication No. EPA-600/3-77-001a. pp. 89-96.
Coffey, P. E.; Westberg, H. (1977) The issue of natural organic emissions.
In: International Conference on Oxidants, 1976-Analysis of Evidence and
Viewpoints. Part IV. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina. October. Publication No. EPA-600/3-77-116.
Cox, R. A. ; and Roffey, M. J. (1977) Thermal decomposition of peroxyacetyl
nitrate in the presence of nitric oxide. Environ. Sci. Techno!. 11(9):900-
906.
Danielsen, E. F. (1980) Stratospheric source for unexpectedly large values of
ozone measured over the Pacific during Gametag, August 1977. J. Geophys.
Res. 85:401-412.
Danielsen, E. F. (1968) Stratospheric-tropospheric exchange based on radioac-
tivity, ozone and potential vorticity. J. Atmos. Sci. 25:502-518.
Danielsen, E. F.; Mohnen, V. A. (1977) Project duststonn report: Ozone trans-
port jji situ measurements, and meteorological analysis of tropopause
folding. J. Geophys. Res. 82:5867-5877.
Darnall, K. R.; Atkinson, R.; and Pitts, J. N., Jr. (1979) Observation of
biacetyl from the reaction of OH radicals with o-xylene. Evidence for
ring cleavage. J. Phys. Chem. 83(15):1943.
Davis, D. D. ; Heaps, W.; Philen, D.; and McGee, T. (1979) Boundary layer
measurements of the OH radical in the vicinity of an isolated power plant
plume: S02 and N02 chemical conversion times. Atmos. Environ.
13:1197-1203.
Davis, D. R.; Jensen, R. E. (1976) Low-level ozone and weather systems. In:
Specialty Conference on Ozone/Oxidants--Interactions with the Total
Environment; Dallas, TX. Pittsburgh, PA: Air Pollution Control Assoc.
pp. 242-251.
Demerjian, K. L. ; Kerr, J. A.; and Calvert, J. G. (1974) The mechanisms of
photochemical smog formation. Adv. Environ. Sci. Technol. 4:1-264.
Demerjian, K. L.; Schere, K. L.; Peterson, J. T. (1980) Theoretical estimates
of actinic (spherically integrated) flux and photolytic rate constants
of atmospheric species in the lower troposphere. Adv. Environ. Sci.
Technol. 10:369-459.
Derwent, R. G.; Eggleton, A. E. J.; Williams, M. L.; Bell, C. A. (1978)
Elevated ozone levels from natural sources. Atmos. Environ. 12:2173-2177.
Dodge, M. C.; Arnts, R. R. (1978) A new mechanism for the reaction of ozone
with olefins. Int. J. Chem. Kinetics XI:399.
019LEX/B 4-45 5/7/34
-------
Doyle, G. J. ; Tuazon, E. C.; Graham, R. A.; Mischke, T. M.; Winer, A. M.; and
Pitts, J. N., Jr. (1979) Simultaneous concentrations of ammonia and
nitric acid in a polluted atmosphere and their equilibrium relationship
to particulate ammonium nitrate. Environ. Sci. Techno!. 13:1416-1419.
Dutkiewicz, V. A.; Husain, L. (1979) Determination of stratospheric ozone at
ground level using 7Be/Ozone ratios. Geophys. Res. Lett. 6:171.
Farber, R. L.; Huang, A. A.; Bregman, L. D.; Mahoney, R. L.; Eatough, D. J. ;
Hansen, L. D.; Blumenthal, D. L. ; Keiffer, W. S.; and Allard, D. W.
(1982) The third dimension in the Los Angeles basin. Sci. Total Environ.
23:345-360.
Gillani, N. V.; Kohli, S.; Wilson, W. E. (1981) Gas to particle conversion of
sulfur in power plant plumes. I. Parameterization of the conversion
rate for dry, moderately polluted ambient conditions. Atmos. Environ.
15:2293-2313.
Graedel, T. E. (1978) Chemical Compounds in the Atmosphere. New York, Academic
Press.
Haagenson, P. L. ; Shapiro, M. A.; Middleton, P. (1981) A case study relating
high ground-level ozone to enhanced photochemistry and isentropic trans-
port from the stratosphere. J. Geophys. Res. 86(C6):5231-5237.
Hampson, R. F. , Jr.; Garvin, D. (1978) Reaction rate and photochemical data
for atmospheric chemistry - 1977. National Bureau of Standards, Washington,
DC. National Bureau of Standards Special Publication 513.
Harris, G. W.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr.; Platt, U.;
Perner, D. (1982) Observations of nitrous acid in the Los Angeles atmos-
phere and implications for predictions of ozone-precursor relationships.
Environ. Sci. Technol. 16:414-419.
Hegg, D. A.; Hobbs, P. V. (1980) Measurements of the gas-to-particle conver-
sion in the plumes from five coal-fired electric power plants. Atmos.
Environ. 14:99-116.
Hendry, D. G.; Kenley, R. A. (1977) Generation of peroxy radicals from
peroxy nitrate (R02 N02) decomposition of peroxyacyl nitrates. J. Amer.
Chem. Soc. 99:3198.
Hendry, D. G. ; Baldwin, A. C. ; Barker, J. R. ; Golden, D. M. (1978) Computer
modeling of simulated photochemical smog. EPA-600/3-78-059. June.
Holzworth, G. C. (1964) Estimates of mean maximum mixing depths in the contig-
uous United States. Monthly Weather Rev. 92:235-242.
Holzworth, G. C. (1972) Mixing heights, wind speeds, and potential for urban
air pollution throughout the contiguous United States. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina. Publication
No. AP-101.
019LEX/B 4-46 5/7/84
-------
Holzworth, G. C.; Fisher, R. W. (1979) Climatological summaries of the lower
few kilometers of rawinsonde observations. U.S. Environmental Protection
Agency. Research Triangle Park, North Carolina. Publication No. EPA-
600/4-79-026.
Hosier, C. R. (1961) Low-level inversion frequency in the contiguous United
States. Monthly Weather Rev. 89:319-339.
Hull, L. A. (1981) Terpene ozonolysis products. In: Atmospheric Biogenic
Hydrocarbons. J. J. Bufalini and R. R. Arnts, eds. Vol. II. Ann Arbor,
MI; Ann Arbor Science Publishers, p. 161.
Husain, L. ; Coffey, P. E.; Meyers, R. E.; Cederwal, R. T. (1977) Ozone trans-
port from the stratosphere to troposphere. Geophys. Res. Lett. 4:363-365.
Jaffee, R. J.; Smith, F. C., Jr.; West, K. W. (1974) Study of factors affecting
reactions in environmental chambers. Phase III. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina. Publication
No. EPA-650/3-74-004b.
Jeffries, H.; Fox, D.; Kamens, R. (1975) Outdoor smog chamber studies. Effect
of hydrocarbon reduction on nitrogen dioxide. U.S. Environmental Protec-
tion Agency, Research Triangle Park, North Carolina. Publication
No. EPA-650/3-75-001.
Jeffries, H. E. ; Fox, D. L.; Kamens, R. M. (1976) Outdoor smog chamber studies-
light effects relative to indoor chambers. Environ. Sci. Techno!. 10(10):
1006-1011.
Johnson, W. B.; Viezee, W. (1981) Stratospheric ozone in the lower troposphere.
I. Presentation and interpretation of aircraft measurements. Atmos.
Environ. 15:1309-1323.
Junge, C. E. (1963) Air chemistry and radioactivity. New York, NY: Academic
Press.
Kamens, R. M.; Jeffries, H. E.; Sexton, K. G. (1981) Smog chamber experiments
to test oxidant-related control strategy issues. Final report. Grant
No. 805726. Bruce Gay, Jr., project officer. U.S. Environmental Protec-
tion Agency, Research Triangle Park, North Carolina.
Kelly, N. A.; Wolff, G. T.; Ferman, M. A. (1981) Background pollutant measure-
ments in air masses affecting the eastern half of the United States. I.
Air masses arriving from the northwest. Atmos. Environ. 16:1077-1088.
Killus, J. F. ; Whitten, G. I. (1982) A mechanism describing the photochemi-
cal oxidation of toluene in smog. Atmos. Environ. 16:1973.
Korshover, J. (1967) Climatology of stagnating anticyclones east of the Rocky
Mountains, 1936-1966. U.S. Public Health Service, National Air Pollution
Control Administration, Cincinnati, Ohio. Publication No. 999-AP-34.
019LEX/B 4-47 5/7/84
-------
Korshover, J. (1975) Climatology of stagnating anticyclones east of the Rocky
Mountains, 1936-1975. NOAA Technical Memorandum ERL ARL-55. U.S. Depart-
ment of Commerce, National Oceanic and Atmospheric Administration,
Washington, DC. July.
Lamb, R. G. (1977) A case study of stratospheric ozone affecting ground level
oxidant concentrations. J. Appl. Meteorol. 16:780-794.
Leighton, P. A. (1961) Photochemistry of air pollution. Academic Press, Inc.;
New York.
Ludwig, R. L. ; Reiter, E.; Shelar, E.; Johnson, W. B. (1977) The relation of
oxidant levels to precursor emissions and meteorological features. Vol.
I. Analysis and Findings. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina. Publication No. EPA-450/3-77-022a.
Lynn, D. A.; Steigerwald, B. J.; Ludwig, J. H. (1964) The November-December
1962 air pollution episode in the eastern United States. U.S. Public
Health Service, National Air Pollution Control Administration, Cincinnati,
Ohio. Publication No. 999-AP-77.
Lyons, W. A.; Olsson, L. E. (1972) Mesoscale air pollution transport in the
Chicago lake breeze. J. Air Poll. Contr. Assoc. 22:876-881.
National Academy of Sciences (1977). Ozone and Other Photochemical Oxidants.
Washington, DC; National Academy of Sciences, pp. 239-279.
Neiburger, M. ; Johnson, D. S. ; Chen, C-W. (1961) Studies of the structure of
the atmosphere over the Eastern Pacific Ocean in the summer - I. The
inversion over the Eastern North Pacific Ocean. University of California.
Publications in Meteorology, Vol. 1, No. 1, pp. 1-94. University of
California Press, Berkeley, California.
Niki, H.; Daby, E. E.; Weinstock, B. (1972) Mechanisms of smog reactions. Adv.
Chem. Series 113:16-57.
Niki, H.; Maker, P. D.; Savage, C. M. ; Breitenbach, L. P. (1978) Mechanism for
hydroxyl radical initiated oxidation of olefin-nitric oxide mixtures in
parts per millon concentrations. J. Phys. Chem. 82(2):135-137.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1981) An FT-IR
study of a transitory product in the gas-phase ozone-ethylene reaction.
J. Phys. Chem. 85:1024-1027.
Pate, C. T.; Atkinson, R.; Pitts, J. N., Jr. (1976) The gas phase reactions of
03 with a series of aromatic hydrocarbons. J. Environ. Sci. Health -
Environ. Sci. Eng. All: I.
Perner, D.; Platt, U. (1979) Detection of nitrous acid in the atmosphere by
differential optical absorption. Geophys. Res. Lett. 6:917-920.
Peterson, J. T. (1976) Calculated actinic fluxes (290-700 nm) for air pollution
photochemistry applications. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina. Publication No. EPA-600/4-76-025.
019LEX/B 4-48 5/7/84
-------
Pitts, J. N., Jr.; Winer, A. M. ; Darnall, K. R.; Lloyd, A. C.; Doyle, G. J.
(1977) Hydrocarbon reactivity and the role of hydrocarbons, oxides of
nitrogen, and aged smog in the production of photochemical oxidants. In:
Proceedings of the International Conference on Photochemical Oxidant
Pollution and Its Control, vol. II. EPA-600/3-77-016, January.
Pitts, J. N., Jr.; Winer, A. M.; Harris, G. W.; Carter, W. P. L. ; Tuazon,
E. C. (1983) Trace nitrogenous species in urban atmospheres. Environ.
Health Perspect. 52:153-157.
Platt, U.; Perner, D. ; Winer, A. M. ; Harris, G. W.; Pitts, J. N. , Jr. (1980)
Detection of N03 in the polluted troposphere by differential optical
absorption. Geophys. Res. Lett. 7:89-92.
Price, J. H. (1976) A study of factors associated with high urban ozone concen-
trations in Texas. In: Specialty Conference on Ozone/Oxidants-Interactions
with the Total Environment. Pittsburgh; Air Poll. Contr. Assoc. pp. 282-
292.
Prior, E. J.; Schiess, J. R.; McDougal, D. S. (1981) Approach to forecasting
daily maximum ozone levels in St. Louis. Environ. Sci. Technol. 15:430-436.
Reiter, E. R. (1963) A case study of radioactive fallout. J. Appl. Meteorol.
2:691-705.
Reiter, E. R. (1975) Stratospheric-tropospheric exchange processes. Rev.
Geophys. Space Phys. 13:459-473.
Reiter, E. R.; Mahlman, J. D. (1965) Heavy radioactive fallout over the
southern United States. November 1962. J. Geophys. Res. 70:4501-4519.
Roberts, J. M.; Fehsenfeld, F. C.; Albritton, D. L.; Sievers, R. E. (1983)
Measurement of monoterpene hydrocarbons at Niwot Ridge, Colorado. J.
Geophys. Res. 88:106-167.
Robinson, E. (1952) Some air pollution aspects of the Los Angeles temperature
inversion. Bull. Amer. Meteorol. Soc. 33:247-250.
Rubino, R. A.; Bruckman, L.; Magyar, J. (1976) Ozone transport. J. Air Poll.
Contr. Assoc. 26:972-975.
Schrenk, H. H.; Heimann, H.; Clayton, G. D.; Gafafer, W. M.; Wexler, H. (1949)
Air pollution in Donora, PA - Epidemiology of the unusual smog episode of
October 1948. Preliminary report, Public Health Bull. No. 306, Public
Health Service, Washington, DC.
Schwartz, W. (1974) Chemical characterization of model aerosols. U.S. Environ-
mental Protection Agency, Research Triangle Park, NC. Publication No.
EPA-650/3-74-011.
Sexton, K. (1982) Evidence of an additive effect for small city plumes.
Presentation at the 75th Annual Meeting of the Air Pollution Control
Association, New Orleans, LA, June 20-25, 1982. Paper No. 82-31.4.
019LEX/B 4-49 5/7/84
-------
Sexton, K. ; Westberg, H. (1980) Elevated ozone concentrations measured
downwind of the Chicago-Gary urban complex. J. Air Poll. Contr. Assoc.
30:911-914.
Sexton, K.; Westberg, H.; Robinson, E. (1981) Transport of pollutants along
the western shore of Lake Michigan. J. Air Poll. Contr. Assoc. 31:385-388.
Shreffler, J. H.; Evans, R. B. (1982) The surface ozone record from the
regional air pollution study, 1975-1976. Atmos. Environ. 16:1311-1321.
Singh, H. B.; Viezee, W.; Johnson, W. B.; Ludwig, F. L. (1980) The impact of
stratospheric ozone on tropospheric air quality. J. Air Poll. Contr.
Assoc. 30:1009-1017.
Slade, D. (1968) Meteorology and atomic energy. U.S. Atomic Energy Commission,
Washington, D.C.
Spicer, C. W. ; Joseph, D. W. ; Sticksel, P. R. (1982) An investigation of the
ozone plume from a small city. J. Air Poll. Contr. Assoc. 32:278-281.
Spicer, C. W. ; Joseph, D. W.; Sticksel, R. R. ; Ward, G. F. (1979) Ozone
sources and transport in the Northeastern United States. Environ. Sci.
Techno!. 13:975-985.
Stair, R. (1961) Proceedings of Third National Air Pollution Symposium, Pasadena,
California, 1955; p. 48. Cited in: P. A. Leighton, Photochemistry of
Air Pollution. New York; Academic Press, Inc.
Stelson, A. W.; Friedlander, S. K.; Seinfeld, J. H. (1979) A note on the equi-
librium relationship between ammonia and nitric acid and particulate
ammonium nitrate. Atmos. Environ. 13:369-371.
Sticksel, P. R.; Spicer, C. W.; Westberg, H. (1979) Ozone plumes from small cities
and airborne observations of ozone in a moving high-pressure system. Pre-
sentation at Annual Meeting, Air Pollution Control Association; June. Paper
No. 79-58.6
Stockwell, W. R. ; Calvert, J. G. (1983) The mechanism of the H0-S02 reaction.
Atmos. Environ. 17(11)-.2231-2235.
Su, F. ; Calvert, J. G.; Shaw, J. H. (1980) An FTIR spectroscopic study of the
ozone-ethane reaction mechanism in 02-rich mixtures. J. Phys. Chem.
84:239-246.
Tiao, G. C. ; Box, G. E. P.; Hamming, W. J. (1975) Analysis of Los Angeles
photochemical smog data: A statistical overview. J. Air Poll. Contr.
Assoc. 25:260-268.
Tilden, J. W.; Seinfeld, J. H. (1982) Sensitivity analysis of a mathematical
model for photochemical air pollution. Atmos. Environ. 16(6):1357-1364.
Tuazon, E. C. ; Winer, A. M. ; and Pitts, J. N. , Jr. (1981) Trace pollutant
concentrations in a multiday smog episode in the California South Coast
Air Basin by long pathlength Fourier transform infrared spectroscopy.
Environ. Sci. Technol. 15:1232-1237.
019LEX/B 4-50 5/7/84
-------
U.S. Department of Commerce (1968). Climatic Atlas of the United States.
Asheville, NC; National Climatic Center. (Reprinted 1977.)
U.S. Environmental Protection Agency (1978). Air quality criteria for ozone and
other photochemical oxidants. Research Triangle Park, NC. Publication No.
EPA-600/8-78-004.
U.S. Environmental Protection Agency (1982a) Air quality criteria for particulate
matter and sulfur oxides. Publication No. EPA-600/8-82-029. Research
Triangle Park, NC.
U.S. Environmental Protection Agency (1982b). Air quality criteria for oxides
of nitrogen. Research Triangle Park, North Carolina. Publication No.
EPA-600/8-82-026.
U.S. Environmental Protection Agency (1981). Comparative application of the
EKMA in the Los Angeles area. U.S. Environmental Protection Agency,
Research Triangle Park, NC. Publication No. EPA-450/4-81-030d.
Vaughan, W. M.; Chan, M.; Cantrell, B.; Pooler, F. (1982) A study of persistent
elevated pollution episodes in the Northeastern United States. Bull. Amer.
Meteorol. Soc. 63:258-266.
Viezee, W.; Johnson, W. B.; Singh, H. B. (1979) Airborne measurements of
stratospheric ozone intrusions into the troposphere over the United
States. Final Report, SRI Project 6690 for Coordinating Research Council,
Atlanta, Georgia.
Viezee, W. ; Singh, H. B. (April 1982) Contribution of stratospheric ozone
to ground level ozone concentrations - A scientific review of existing
evidence. Final Report (revised draft). EPA Grant CR 809330010, SRI
International, Menlo Park, California.
Viezee, W.; Singh, H. B.; Shigeishi, H. (1982) The impact of stratospheric
ozone on tropospheric air quality-Implications from an analysis of existing
field data. Final Report. Prepared for Coordinating Research Council,
Atlanta, GA, under CRC Contract No. CAPA-15-76(1-80), by SRI International,
Menlo Park, CA.
Westberg, H.; Allwine, K. J. ; Elias, D. (1976). Vertical ozone distribution
above several urban and adjacent rural areas across the United States.
In: Specialty Conference on Ozone/Oxidants - Interactions with the Total
Environment. Pittsburgh: Air Poll. Contr. Assoc. pp. 84-95.
Westberg, H. ; Allwine, K. J. ; Robinson, E. (1978). The transport of oxidant
beyond urban areas. Light hydrocarbon and oxidant data, New England
Study, 1975. U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina. Publication No. EPA-600/3-78-006.
Wilson, W. E. (1978) Sulfates in the atmosphere: A progress report on projected
MISTT. Atmos. Environ. 12:537-548.
Winer, A. M.; Breuer, G. M.; Carter, W. P. L.; Darnall, K. R.; Pitts, J. N. , Jr.
(1979) Effects of ultraviolet spectral distribution on the photochemistry
of simulated polluted atmospheres. Atmos. Environ. 13:989-998.
019LEX/B 4-51 5/7/84
-------
5. PROPERTIES, CHEMISTRY, AND MEASUREMENT
OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
5.1 INTRODUCTION
The previous chapter presented information on the atmospheric chemistry
of those compounds that serve as precursors to ozone and related photochemical
oxidants in ambient air. That chapter included a discussion of the atmospheric
processes that result in ozone and oxidant formation and also described the
transport of those oxidants once formed. The present chapter deals with the
physical and chemical properties and typical reactions ("type" reactions) of
ozone and other oxidants, especially those properties and type reactions in
ambient air and in solution-phase systems that are pertinent to understanding
the direct effects of ozone and related photochemical oxidants on biological
and nonbiological receptors. In addition, it presents detailed information on
methods for measuring ozone and the two other most abundant photochemical
oxidants (other than nitrogen dioxide) in ambient air, hydrogen peroxide and
peroxyacetyl nitrate (PAN), along with its higher homologues. The information
presented should prove to be an aid to state and local air pollution control
agencies and to researchers investigating health and welfare effects. The
chief reasons for presenting such information, however, are to provide relevant
information (1) for understanding the general basis of the effects of ozone
and other oxidants in biological systems; (2) for assessing the accuracy of
aerometric data on these pollutants; and (3) for determining the impact of
measurement and calibration biases on existing data on the health and welfare
effects of ozone, total oxidants, and individual other oxidants.
5.2 PROPERTIES OF OZONE, PEROXYACETYL NITRATE, AND HYDROGEN PEROXIDE
5.2.1 Ozone
Ozone (0,) is a triangularly shaped molecule consisting of three oxygen
*5
atoms arranged in four basic resonance structures:
/+^ / \
(I) (ID (HI) (IV)
019AA/A 5-1 6/15/84
-------
The first and fourth structures, which predominate, are characterized by the
presence of a terminal oxygen atom having only six electrons. The resonance
forms depicted above have no unshared electrons. Thus, unlike other common
oxidants in ambient air—oxygen (Op) and nitrogen dioxide (NO,,)—ozone is not
paramagnetic. Paramagnetism would impart free-radical properties to ozone.
Ozone, then, does not itself behave as a free radical, though it is thought to
give rise to free radicals in some biological and other aqueous systems (chap-
ters 7, 9, and 10).
As the result of the presence of only six electrons on one of the oxygen
atoms in ozone, the chemical reactions of ozone are electrophilic; that is,
ozone removes electrons from or shares electrons with other molecules or ions.
The terms "oxidant" and "oxidizing agent" characterize an ion, atom, or molecule
that is capable of removing one or more electrons from another ion, atom, or
molecule, a process called "oxidation." A "reducing agent" adds one or more
electrons to another ion, atom, or molecule, a process called "reduction."
Oxidation and reduction reactions occur in pairs and the coupled reactions are
known as "redox reactions." In redox reactions, the oxidizing agent is reduced
and the reducing agent is oxidized. The two components of such redox reactions
are known as "redox pairs." The significance of redox reactions involving
ozone is discussed in chapters 7 and 10. The capability of a chemical species
for oxidizing or reducing is termed "redox potential" (positive or negative
standard potential) and is expressed in volts. Ozone is a powerful oxidizing
agent having a standard potential of +2.07 volts in aqueous systems (Weast,
1977).
Physical properties of ozone are given in Table 5-1 (U.S. Department of
Health, Education, and Welfare, 1970, modified).
5.2.2 Peroxyacetyl Nitrate
Peroxyacetyl nitrate (PAN) has been observed as a constituent of photochem-
ical smog in many localities, though its concentrations and its ratio to ozone
differ from place to place (chapter 6). Peroxyacetyl nitrate, which has the
formula CH3C002N02, can exist in equilibrium with its decomposition products,
N0? and acetylperoxy radicals, for long periods of time in the presence of
sunlight, depending upon both temperature and the N02/N0 ratio (Cox and Roffey,
1977).
The chief property of interest regarding PAN is its oxidizing ability.
While no standard potential is available in the literature, PAN is known to be
a more potent phytotoxicant, on a concentration basis, than ozone (chapter 7).
019AA/A 5-2 6/15/84
-------
TABLE 5-1. PHYSICAL PROPERTIES OF OZONE
Physical state
Chemical formula
Molecular weight
Melting point
Boiling point
Specific gravity relative to air
Vapor density
At 0°C, 760 mm Hg
At 25°C, 760 mm Hg
Solubility at 0°C
(Indicated volume of ozone at
0°C, 760 mm Hg)
Henry's Law constant,
37°C and pH = 7
Conversion factors
At 0°C, 760 mm Hg
At 25°C, 760 mm Hg
Colorless gas
03
48.0
-192.7 ± 0.2°C
-111.9 ± 0.3°C
1.658
2.14 g/liter
1.96 g/liter
0.494 ml/100 ml water
8666 atm/mole fractionc
1 ppm = 2141 ug/m3 4
1 ug/m3 = 4.670 x 10"
1 ppm = 1962 ug/m3 _4
1 ug/m3 = 5.097 x 10~ ppm
Calculated by formula of Roth and Sullivan (1981).
Source: U.S. Department of Health, Education, and Welfare (1970), modified.
No evidence exists, however, to suggest that it is a comparably potent toxicant
in animals or man (chapters 10 and 11).
A second property of PAN of interest it its thermal instability. In the
laboratory, this thermal instability necessitates that precautions by taken in
synthesizing, handling, and storing PAN, since improper handling and storage
have resulted in explosions. The ready thermal decomposition of PAN results
in a notable temperature dependence of PAN in ambient air (chapter 4).
Partly because of the thermal instability of PAN, its properties have not
been as well characterized as those of 0~ or H-O^. Recent work on the physical
properties of PAN, however, has confirmed data reported earlier, and results
of the earlier and more recent work are shown in Tables 5-2 and 5-3 (Stephens,
019AA/A 5-3 6/15/84
-------
TABLE 5-2. PHYSICAL PROPERTIES OF PEROXYACETYL NITRATE
Physical state, @25°C
Chemical formula
Molecular weight
Boiling point, °C
Triple point, °C
Vapor pressure,
@room temperature
Vapor pressure curve
Hydrolysis
In alkaline solution
In acidic solution
@22°C, pH 5.6
@25°C, pH 5.6
Colorless liquid
121
a
106 ±2
103.9°
-50 ± 0.5'
-48 ± 0.5C
~15 mm Hg
In p = -4587/T + 18.76*' C
In p = -4585/T + 18.79°
Rate not available; products include
nitrite ion and molecular oxygen
4 x 10~4 sec"1
6.8 x 10"4 sec'1
Henry's Law constant,
@1Q°C
Conversion factors
@0°C, 760 mm Hg
@25°C, 760 mm Hg
-1
5 ± 1 M atm
1 ppm = 5398 ug/m3; .
1 ug/m3 = 1.852 x 10~ ppm
ppm
1 ppm = 4945
= 2.022 x
Sources:
aBruckmann and Wilner (1983).
°Kacmarek et al. (1978).
cTemperature, T, in °K; pressure, p, in torr.
Stephens (1967); Nicksic et al. (1967).
eLee et al. (1983).
fHoldren et al. (1984).
Source of remainder of data: U.S. Department of Health, Education, and
Welfare (1970).
019AA/A
5-4
6/19/84
-------
TABLE 5-3. INFRARED ABSORPTIVITIES OF1PEROXYACETYL NITRATE
AT APPROXIMATE RESOLUTION OF 1.2 cm"1 (RELATED TO 295°K
AND 973 mba) (ppm -1m-1 x 104)
Frequency, cm
Reference
Bruckmann and Wi liner, 1983
Stephens, 1969° >d
1842
12.4
10.0
1741
32.6
23.6
1302
13.6
11.2
1162.5
15.8
14.3
791.5
13.4
10.1
aStephens (1979), personal communication, cited in Bruckmann and Winner (1983),
bPure PAN.
CPAN in air.
Stephens (1969).
1969; Bruckmann and Willner, 1983; Holdren et al., 1984; Kacmarek et al., 1978;
U.S. Health, Education, and Welfare, 1970).
The infrared (IR) spectrum of PAN is important since most researchers
rely on it for establishing concentrations of PAN for calibration. Bruckmann
and Willner (1983) reported the IR spectrum of pure PAN and the Raman spectrum
of liquid PAN at -40°C in an argon matrix (using an Ar ion laser as the light
source). The IR and Raman spectra found in their work are shown in Figure 5-1.
The recent work by Bruckmann and Willner (1983) also confirmed effects
that correlate with the ultraviolet (UV) spectrum published earlier by Stephens
(1969); that is, PAN was shown to be stable at \>300 nm but was efficiently
photolyzed at \<300 nm (Bruckmann and Willner, 1983). Wavelengths pertinent
to tropospheric pollutants are greater than 300 nm.
Though PAN is a strong phytotoxicant, it is important to mention here
that PAN does not respond, or responds only negligibly, in measurements made by
the Mast meter (section 5.5). It is a positive interference in ozone measure-
ments made by the NBKI method (section 5.5). Since many of the early field
studies on the effects of oxidants on vegetation utilized Mast meter measure-
ments, this property of PAN relative to measurement methods is important.
5.2.3 Hydrogen Peroxide
Hydrogen peroxide (H-Op) is an oxidant that occurs in ambient air as part
of the photochemical smog complex. It is thought to be formed through the
019AA/A 5-5 6/15/84
-------
I I I I I I I I I
—-3500—3000—2500—2000 —18001—1600 <—1400—1200—1000—800—600—400—200-
I i I I I I I I I I I I I I I I I I I I
Figure 5-1. Top: IR gas spectrum of pure PAN (optical path length 10 cm);
(A) 1.5 torr, (B) 10.0 torr. Bottom: Raman spectrum of PAN at -40°C (li-
quid); excitation light, 514.5 nm (100 mW).
Source: Bruckmann and Willner (1983).
5-6
-------
recombination of two hydroperoxy radicals (H0?-) in the presence of a third,
energy-absorbing molecule (chapter 4; section 5.5.8).
In aqueous media, HpOp is an inorganic acid that has a dissociation
constant of 2.4 x 10"12 and a pK of 11.62 (at 25°C) (Weast, 1977). In the
redox pair H-O^H-O, hydrogen peroxide has a standard potential of ±1.776
volts (Weast, 1977), compared with the comparable standard potential for ozone
of +2.07 volts.
Pertinent physical properties of H202 are given in Table 5-4 (Weast,
1977).
TABLE 5-4. PHYSICAL PROPERTIES OF HYDROGEN PEROXIDE
Physical state, @25°C
Chemical formula
Molecular weight
Melting point, °C
Boiling point, °C, @760 mm Hg
Density, @25°C, 760 mm Hg
Vapor pressure, @16.3°C
Conversion factors
@0°C, 760 mm Hg
@25°C, 760 mm Hg
Colorless liquid
H202
34.01
-0.41
150.2
1.4422
~1 mm Hg
1 ppm = 1520 ug/m3;
1 ug/m3 = 6.594 x 10-4 ppm
1 ppm = 1390 |jg/m3;
1 ug/m3 = 7.195 x 10-4 ppm
Source: Weast (1977).
Additional properties should be noted here that are of interest relative
to whether effects of H?0~ in biological receptors are of significance.
First, HpOp, though classed as a reasonably strong oxidant on the basis of its
standard potential for the redox system H^Op/H^O, has been reported to be a
positive interference in measurements of total oxidants made by the Mast meter
but to give a very slow response (slow color development) in the NBKI method
for total oxidants (section 5.5). This difference should be borne in mind
when effects attributed to oxidants, as opposed to ozone, are evaluated.
019AA/A
5-7
6/15/84
-------
Second, H202 occurs normally as a substrate in biological systems and is
involved in several redox pairs of biological importance, as shown in Table
5-5. It should also be noted that enzymes are present, at least in mammalian
systems, that catalyze the breakdown of H?0?.
TABLE 5-5. NORMAL ELECTRODE POTENTIALS OF SOME HYDROGEN PEROXIDE-CONTAINING
OXIDATION-REDUCTION SYSTEMS OF BIOLOGICAL IMPORTANCE
System Potential (E'Q), volts
H02/H2°2 +1-12
H202/H0' + H20 +0.38
35°2/H2°2 +0-30
Source: West et al. (1966).
5.3 ATMOSPHERIC REACTIONS OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
5.3.1 Introduction
The atmospheric reactions of ozone and other photochemical oxidants such
as peroxyacetyl nitrate (PAN) and hydrogen peroxide (H^) are complex and
diverse. The reactions of these species result in products and processes that
may have significant environmental implications, including effects on biological
systems, nonbiological materials, and such phenomena as visibility degradation
and acidification of cloud and rain water.
Ozone, for example, is highly reactive toward certain classes of organic
compounds (e.g., alkenes) and certain reactions of ozone with alkenes lead to
the formation of secondary organic aerosols. Photolysis of ozone leads to the
formation of 0( D) atoms and, by subsequent reactions, the production of OH
radicals. Ozone may also play a role in the oxidation of SO,, to H-SO., both
indirectly in the gas phase (via formation of OH radicals and Criegee biradi-
cals) and directly in aqueous droplets. Evidence is also accumulating that
hydrogen peroxide, like ozone, is involved in both gas-phase photochemistry
and aqueous-phase oxidations. For example, studies of the rates of oxidation
of S02 by H^O- in solution suggest that this reaction is sufficiently fast
that it could be the major aqueous-phase oxidation route for S0_, under at
least some atmospheric conditions. In addition, the importance of oxidants
such as PAN to various aspects of atmospheric chemistry, such as long-range
transport of NO and multi-day air pollution episodes, is now being recognized.
019AA/A 5-8 6/15/84
-------
In the following sections, the present state of knowledge concerning the
atmospheric reactions of 03, PAN, H202, and formic acid (HCOOH) are summarized
in some detail, including the mechanisms of certain of these reactions.
Emphasis is placed, whenever possible, on those reactions that lead to products
or processes suspected or known to have effects on biological or other important
receptors.
5.3.2 Atmospheric Reactions of Ozone with Organic Compounds
In discussing the reactions of ozone with organic compounds in the tropo-
sphere, it is important to recognize that organics undergo competing reactions
with OH radicals during daytime hours (Atkinson et al., 1979; Atkinson and
Lloyd, 1984) and, in certain cases, with N0» radicals during nighttime hours
(Japar and Niki, 1975; Carter et al., 1981a; Atkinson et al., 1984a,b,c,d), as
well as photolysis. Thus, all organics except the perhaloalkanes exhibit
room-temperature OH radical rate constants of >~5 x 10 cm molecule sec
(Atkinson et al., 1979; Jeong and Kaufman, 1982). Since the ratio of 0., to OH
radical concentrations in the unpolluted troposphere during daylight hours is
believed to be of the order of 106 (Singh et al., 1978; Crutzen, 1982), only
-21
for these organics whose 0, reaction rate constants are greater than ~10
3 -1 ~1
cm molecule sec can consumption by 0~ be considered to be atmospherically
important (Atkinson and Carter, 1984). Although these ozone reactions of
interest are summarized below, the recent reviews by Atkinson and Carter
(1984) and Atkinson and Lloyd (1984) should be consulted for a detailed and
comprehensive discussion of the kinetics and mechanisms of the atmospheric
reactions of ozone with organic compounds.
5.3.2.1 Alkenes. Ozone reacts rapidly with the acyclic mono-, di-, and tri-
alkenes and with cyclic alkenes. The rate constants for these reactions range
from ~10~18 to ~10~14 cm3 molecule"1 sec"1 (Atkinson and Carter, 1984), corre-
sponding to atmospheric lifetimes ranging from a few minutes for the more
reactive cyclic alkenes, such as the monoterpenes, to several days. In polluted
atmospheres, especially in the afternoons during photochemical oxidant episodes,
a significant portion of the consumption of the more reactive alkenes will
occur via reaction with 03, rather than with OH radicals.
It is now reasonably well-established that the initial step in the ozone-
alkene reaction involves the formation of a "molozonide," which rapidly decom-
poses (Harding and Goddard, 1978; Herron et al., 1982) to a carbonyl compound
and a biradical (which is also initially energy-rich):
019AA/A 5-9 6/15/84
-------
°3 +
El
\C-C
E2X
-------
M
CH3CHOO
(40%)
• • ±
[CHjCHOO]
0
—»
CH3 °
xc\
H 0_
*
— >• [CH3COHJ
0
> fwrnrH 1
— >. rn j- rn
CH0 + CO + OH
CH, + CO. + H
3 2
+ CH.O
(19%)
(24%) (5-3)
(5%)
(12%)
where CH.OO and CH-CHOO denote thermal!zed biradicals. These thermal!zed
biradicals have been shown (Calvert et al., 1978; Herron et al., 1982; Atkinson
and Carter, 1984) to undergo bimolecular reactions with aldehydes, SOp, CO, and
HpO; and it is believed that they will also react with NO and NO..
RCHOO + NO •> RCHO + NO.
RCHOO + NO > RCHO + NO
(5-4)
(5-5)
RCHOO
RCHO
RCHOO + HO •* RCOOH + HO
RCHOO + CO + products
. . P — ^
RCHOO 4- R'CHO H
(5-6)
(5-7)
(5-8)
(5-9)
Under atmospheric conditions, the reactions with NO, NO,,, or H^O are
expected to be the dominant loss processes of these thermalized biradicals,
with the precise major reaction pathway depending on the relative concentra-
tions. Hence, in the atmosphere, ozone-alkene reactions can lead ultimately
to the formation of aldehydes and acids, as well as to the conversion of SO,,
to H?SO.. Unlike the case for other free radicals, oxidation of S02 by these
biradicals can take place at night, since ozone can remain aloft at significant
019AA/A
5-11
6/15/84
-------
concentrations through the night, reacting with alkenes in polluted air masses
to produce such biradicals. While this appears, however, to be the only
significant homogeneous gas-phase oxidation rate for SCL at night, it is
probably a minor process in the overall oxidation of SCL during long-range
transport (Finlayson-Pitts and Pitts, 1982). Furthermore, 03 can react with
S0"2 in aqueous droplets to yield acidic species (see below).
The limited data presently available for the haloalkenes show that fluorine
and chlorine substitution (no data are available for bromine or iodine substitu-
ents) on the alkenes decreases, markedly, in most cases, the room-temperature
rate constants, compared to those for the corresponding alkenes. Hence, for
the haloalkenes studied to date (Atkinson and Carter, 1984), their reactions
with ozone are of minor importance under atmospheric conditions, compared to
reactions with OH radicals.
While the above mechanistic discussion applies mainly to the simple
acyclic monoalkenes, the initial reactions for the di- and poly-alkenes, the
cyclic alkenes, and the haloalkenes are believed to be analogous (Niki et al.,
1983; Zhang et al. , 1983). For example, for 1,3-butadiene, cyclohexene, and
vinyl chloride, the reactions are expected to proceed via:
0 + CH =CH-CH=CH
'A
? ?
, CH2-CH-CH=CH
/\
HCHO + [CH2=CH-CHOO]
CH2=CHCHO
(5-10)
(5-11)
[CH -CH-CHOO]*
0
V
o
ii • • *
-> [HC-CH CE CH CH CHOOr (5-12)
019AA/A
5-12
6/15/84
-------
'A '*
0 0
I I
CH -CHC1
J (5-13)
\
HCHO + [CHC100]* [CH200]* + HCOC1
with the initially energy-rich biradicals undergoing subsequent reactions.
The excited halogenated biradicals such as [CHC100] are expected to decompose
to form Cl atoms and other radical species (Niki et al., 1983):
->• Cl* + OH + CO
The modes of reaction under atmospheric conditions of the more complex excited
biradicals, such as [CH2=CH-CHOO] and [HCO(CH2)4CHOO] shown above, are, how-
ever, presently unknown.
5.3.2.2 Alkanes and Alkynes. At the present time, there appears to be no
convincing evidence in the literature for an elementary reaction between 0~
and the alkanes (Atkinson and Carter, 1984). The reported rate constants of
~23 -26 3 —1 —1
10 to 10 cm molecule sec are thus clearly unimportant under atmos-
pheric conditions. Similarly, although there is presently substantial uncer-
tainty (Atkinson and Carter, 1984) concerning the rate constants for the
reactions of ozone with the simple alkynes (e.g., acetylene, propyne, and
1-butyne), most of the available room temperature data for these
-------
5.3.2.3 Aromatics. As in the case of the alkanes, the aromatic hydrocarbons
react only very slowly with 0. (Pate et al., 1976; Atkinson et a!., 1982) and
these reactions are not expected to be important in the atmosphere (Atkinson
and Carter, 1984). Although the cresols are significantly more reactive than
the aromatic hydrocarbons (Atkinson et al., 1982), under atmospheric conditions
their reactions with 0- are minor compared to their reactions with OH radicals
(Atkinson et al., 1978; 1982) or N03 radicals (Carter et al., 1981a; Atkinson
et al., 1984d).
5.3.2.4 Oxygen-Containing Organics. For those oxygen-containing compounds
that do not contain unsaturated carbon-carbon bonds (e.g., formaldehyde,
acetaldehyde, glyoxal, and methylglyoxal), the reactions with ozone are very
slow, and, by analogy, this is expected to be the case for all ethers, alcohols,
aldehydes, and ketones not containing unsaturated carbon-carbon bonds. For the
carbonyls and ethers (other than ketene) that contain unsaturated carbon-carbon
bonds, however, much faster reactions are observed (Atkinson et al., 1981).
Few data are available, however, concerning the mechanisms of the reac-
tions of 03 with such oxygen-containing organics, the only published informa-
tion being that of Kamens et al. (1982). From a study of the reactions of 03
with methacrolein and methyl vinyl ketone, methylglyoxal was observed as a
product, along with other minor products (Kamens et al., 1982), as anticipated
from the reaction schemes.
0 0
I I
CH -CHCOCH
(5-15)
HCHO + [CH3COCHOO]* C^COCHO +
019AA/A 5-14 6/15/84
-------
and
CH
CH
X
CHO
HCHO +
CH COO
•J »
(5-16)
CH3COCHO
-,*
5.3.2.5 Nitrogen-Containing Organics. While the kinetics of the reactions of
0_ with a variety of nitrites, nitriles, nitramines, nitrosoamines, amines,
and hydrazines have been studied (Atkinson and Carter, 1984), only for the
hydrazines are these reactions sufficiently rapid to be of atmospheric import-
ance. Indeed, hydrazine, monomethylhydrazine, and 1,1- (or unsymmetrical-)
dimethylhydrazine react sufficiently rapidly with 0- (with rate constants of
"*17 3 ~1 —1 —I*! '•I — 1-1
~3 x 10 cm molecule sec for hydrazine and >10 cm molecule sec
for monomethylhydrazine, and 1,1-dimethy!hydrazine [Tuazon et al., 1982] that
their major atmospheric reactions are likely to be via reaction with 0_. The
O
initial reactions are not completely understood, but for hydrazine, monomethyl-
hydrazine, and 1,1-dimethylhydrazine, they appear to involve H-atom abstraction
from the weak N-H bonds to form an OH radical.
OH
(5-17)
Using FT-IR absorption spectroscopy, Tuazon et al. (1981a, 1982) have
obtained extensive product data concerning the reactions of N2H2, CHJWNH-,
and (ChL)NNH9 with 0_; and these data have allowed plausible reaction pathways
6 £. 6 ,
to be postulated. In these reaction schemes, the R..R_NNH radicals
[where R.,, R. = H or ChL for the hydrazines N.H.,
are proposed to react as follows:
!, CH^NHNH-,
and (CH3)2NNH2]
For RNHNH or RNNH£ radicals:
019AA/A
5-15
6/15/84
-------
RNHNH
RNN
H2
+ 0 * RN=NH + HO
(5-18)
RN=NH +
°
OH
RN=N
OH
H20
(5-19)
R+ N2
though additional (and more uncertain) reactions, giving rise to diazomethane
and other products, occur for R=:CH3 (Tuazon et al. , 1981a; Carter et al.,
1981b). For the (CH_)?NNH radical, this reaction sequence cannot occur, and
Tuazon et al. (1981a) have postulated other pathways leading to N-nitrosodi-
methylamine, the major observed product.
(CH3)2NNH
(CH3)2NNO
0'
H
!°=
(5-20)
or
(5-21)
The N-ni trosodimethyl ami ne observed in this system by FT-IR spectroscopy
(Tuazon et al., 1981a) is considered to be a carcinogen (Andrews et al . , 1978;
Rao, 1979). Despite this progress, significant uncertainties and inconsisten-
cies remain in the present understanding of these reactions of 03 with these
simple hydrazines
019AA/A
, CHNHNH, and (CH)NNH).
32
5-16
6/15/84
-------
The sole information concerning the gas-phase reaction of 0- with diazenes,
diazomethane, and hydrazones is that of Tuazon et al. (1982). These species
were observed to be products in the reactions of the simple hydrazines with 0_
(see above) or of hydrazine with formaldehyde; and were observed to react
further when formed in the presence of CL.
5.3.2.6 Sulfur-Containing Organics. Based upon the kinetic data available for
dimethyl sulfide, thiirane, and thiophene, it appears at the present time that
the rates of reaction of 0« with sulfur-containing organics can be considered
to be unimportant for atmospheric purposes (Atkinson and Carter, 1984).
5.3.2.7 Other Reactions
5.3.2.7.1 Organometallics. To date, rate constants have been reported only
for tetramethyl- and tetraethyl-lead (Harrison and Laxen, 1978). No mechanistic
or product data are available for these reactions; but it is possible that the
reactions proceed via initial H-atom abstraction from the alky! C-H bonds, e.g.:
•
j + OH + 02 (5-22)
-Va + OH + °2 (5-23)
5.3.2.7.2 Radical species. Because of the low concentration of 0, and radicals
in the atmosphere, and because both alkyl and most alkoxy radicals react at
significant rates with 02 (which is present at a concentration >105 higher
than 03 in ambient atmospheres), these reactions can be considered to be of
negligible importance in the atmosphere.
5.3.2.8 Atmospheric Lifetimes. Table 5-6 compares the room-temperature rate
constants and the approximate corresponding atmospheric loss rate constants
for reaction with 03 (over a 24-hour period), with OH radicals during daytime
hours, and with N0_ radicals during nighttime hours.
It can be seen that under these atmospheric conditions the reactions with
03 are important for the higher alkenes, including the monoterpenes, and for
the hydrazines. For the other organics for which kinetic data are available,
their reactions with 03 are generally of negligible or minor importance.
5.3.2.9 Aerosol Formation. The reaction of ozone with alkenes having six or
more carbon atoms and with cyclic alkenes and conjugated alkenes has been
shown to lead to the formation of secondary organic aerosols (National Academy
019AA/A 5-17 6/15/84
-------
of Sciences, 1977; Scheutzle and Rasmussen, 1978). For ambient concentrations
of ozone, the amount of aerosol formation is generally greater the higher the
carbon number, and classes of organics such as the terpenes are very efficient
at producing aerosol.
A spectrum of products have been identified as constituents of secondary
organic aerosols. These include a wide range of carboxylic and dicarboxylic
acids, aldehydes, alcohols, and other oxygenated compounds. The chemical
pathways leading to these products are exceedingly complex and are far from
well-characterized. Many of the studies, however, in which products have been
TABLE 5-6. CALCULATED LIFETIMES OF SELECTED ORGANICS
RESULTING FROM ATMOSPHERIC LOSS BY REACTION WITH
03 AND WITH OH AND N03 RADICALS
Organic lifetimes
Organic compound
Anthropogenic alkenes
Ethene
Propene
trans-2-Butene
2-Methyl -2-butene
2 , 3-Di methyl -2-butene
Naturally emitted alkenes
Isoprene
a-Pinene
foPinene
A -Carene
d-Limonene
Hydrazines
Hydrazine
Monomethyl hydrazine
03,
24- hour
2.7 days
11 hr
35 min
17 min
6 min
10 hr
1.4 hr
5.5 hr
1.0 hr
11 min
-1.9 hr
<4 min
OH,
daytime
16 hr
5.6 hr
2.0 hr
1.6 hr
1.3 hr
1.4 hr
2.3 hr
1.8 hr
1.7 hr
1.0 hr
1.4 hr
1.4 hr
a
, T
N03,
nighttime
79 days
1.1 days
33 min
1.3 min
0.2 min
22 min
2 min
5 min
1.2 min
0.9 min
-
Assuming 100 ppb of 03 (24-hr average), 2 x 106 cm-3 (0.08 ppt) of OH
during daylight hours, and 100 ppt of N03 during nighttime hours.
019AA/A
5-18
6/15/84
-------
identified, were carried out at organic precursor concentrations greatly exceed-
ing their ambient levels. Hence, it is still not totally clear whether organic-
0- reactions under ambient-atmosphere conditions lead to significant amounts
of aerosol formation.
Regardless of the chemistry involved, however, the condensation of such
products into particles which then grow into the light-scattering size range
(~0.1 to 1 urn) can contribute to visibility reduction in both urban airsheds and
in natural environments.
5.3.3 Atmospheric Reactions of Ozone with Inorganic Compounds and with Light
Ozone reacts rapidly with NO to form NO-:
0. + NO -»• N0_ + 0. (5-24)
Because this reaction is rapid, ozone concentrations in urban atmospheres can-
not rise significantly until most of the NO emitted from combustion sources
has been converted to NO-. Ozone can react with N0? to produce the nitrate
(NO-) radical and an oxygen molecule:
0 + NO •*• NO + 02 (5-25)
The NO- radical has recently been shown to be an important sink for certain
•J
classes of organic compounds, including several of the monoterpenes and dimethyl
sulfide (Winer et al., 1984).
Photolysis of ozone can be a significant pathway for formation of OH
radicals, particularly during mid-afternoon when 0_ concentrations are at a
J
maximum in urban airsheds:
+ hv (X < 319 run) •» 0(1D) + 02( A) (5-26)
0(1D) + H20 -> 2 OH (5-27)
019AA/A 5-19 6/15/84
-------
The reaction of the 0( D) atoms, formed from 0» photolysis, with water vapor
occurs with about 20 percent efficiency at ambient temperatures and about 50
percent relative humidity, with about 80 percent of the 0( D) atoms being
quenched to 0( P) atoms by N~ and 0-.
5-3.4 Reactions of Ozone in Aqueous Droplets
While the thermal oxidation of S02 by ozone in the gas phase appears to
be too slow to be important in acid deposition phenomena, the role of ozone in
oxidizing SO- dissolved in water droplets (e.g., cloud, fog, or rain) may be
of considerable significance. At 25°C, ozone has a Henry's Law constant of
-2 -i -1
10 mol L atm (Kirk-Othmer, 1981). Given ambient concentrations ranging
from 30 to about 300 ppb, 0~ would be expected to have concentrations in
-10 -1
aqueous droplets in the atmosphere of approximately 3-30 x 10 mol L . The
rate of reaction between 0« and S0«, when both are dissolved in aqueous drop-
lets, has been shown in laboratory studies to be relatively fast (Penkett et
al., 1979; Kunen et al., 1983; Martin, 1984; Hoffman et al., 1984; Schwartz,
1984; Brock and Durham, 1984), although the rate of this reaction is pH-depen-
dent and decreases as the acidity of the solution increases.
Figure 5-2 shows data reported by Schwartz (1984) for the rate of aqueous-
phase oxidation of S(IV) by 30 ppb of 03 (and also by 1 ppb of H2°2* as a
function of solution pH. The aqueous-phase oxidation rate, R, per part-per-
billion S02 partial pressure decreases with decreasing pH by roughly a factor
of 20 per pH unit. This pH dependence reflects the solubility of S(IV) as
well as a slight pH dependence of the second-order rate constant for the oxida-
tion of S(IV) by 03 (Erickson et al., 1977; Larson et al., 1978; Penkett et al.,
1979). Schwartz (1984) concluded, from consideration of these data and uptake
times for SO-, that oxidation of SO- by 0_ cannot produce solution pH values
below ~4.5. Schwartz (1984), however, has also interpreted the field data of
Hegg and Hobbs (1981) for sulfate production rates at the inflow and outflow
regions of lenticular clouds as being consistent with the aqueous-phase oxida-
tions of S(IV) by 03.
An additional aspect of the role of 03 in the chemistry of aqueous droplets
concerns its photolysis to yield OH radicals in solution (Graedel and Weschler,
1981; Chameides and Davis, 1982):
(03)aq + hv + (0(1D))aq + 02 (aq) (5-28)
0(1D)aq + (H20)aq + 2(OH)aq (5-29)
019AA/A 5-20 6/15/84
-------
10"
10"
o
05
.n
a
10-'
v,
O
V)
a.
£ 10"
10-'
10''
H2O2, 1 ppb
O3, 30 ppb
1000
100
10
«""
_r
"5.
0.1
0.01
4
PH
Figure 5-2. Rate of aqueous-phase oxidation
of S(IV) by O3 (30 ppb) and H2O2 (1 ppb), as a
function of solution pH. Gas-aqueous
equilibria are assumed for all reagents.
R/p§O2 represents aqueous reaction rate
per ppb of gas-phase SO2; p/L represents
rate of reaction referred to gas-phase SO2
partial pressure per cm3—m"3 liquid water
volume fraction (Schwartz, 1984).
5-21
-------
and its reactions with aqueous OH ions and H_0p to yield aqueous H0_ radicals
(Chameides and Davis, 1982). The OH radicals formed by this i_n situ process
can result in the oxidation of S(IV).
For discussions of possible mechanisms for the oxidation of SO- by 0~ in
aqueous systems, the primary literature should be consulted (Graedel and
Weschler, 1981; Chameides and Davis, 1982; Calvert, 1984).
5.3.5 Atmospheric Reactions of Peroxyacetyl Nitrate (PAN)
With the recognition in recent years that PAN is a ubiquitous nitrogenous
species in the troposphere (Singh and Hanst, 1981; Singh and Sal as, 1983a;
Penkett, 1983; Spicer et al., 1983; Aikin et al., 1983) and in the lower
stratosphere (Aikin et al. , 1983), there has been renewed focus on the atmos-
pheric role of this organic compound.
Smog-chamber studies have shown that, once formed, PAN can be relatively
stable under atmospheric thermal conditions (Pitts et al., 1979; Akimoto et
al., 1980). Since PAN is, however, in equilibrium with acetyl peroxy radicals
and NO-s
0 0
II II
CH COONO > CH COO + NO (5~30)
3 £ J *f
any process that removes either acetyl peroxy radicals or N0? will lead to the
decomposition of PAN. One such process is the reaction of NO with CH_C(0)0_
radicals. Since PAN has been shown to persist through the night in urban
atmospheres (Tuazon et al. , 1980; 1981b), the reaction of PAN with NO during
the morning traffic peak can lead to the formation of OH radicals via the
following mechanism (Carter et al., 1981c):
0 0
II II
CH COO -f- NO •»• CH CO. + NO (5-31)
0
II
CH3C~°" * CH3* + C°2 (5-32)
M
CV + °2 * CH300' (5-33)
CH300. + NO •»• CH30. + N02 (5-34)
019AA/A 5-22 6/15/84
-------
HCHO (5-35)
H02 + NO -»• OH + N02 (5-36)
M
OH + NO -> HONO (5-37)
M
OH + N02 + HN03 (5-38)
Thus, the reaction of PAN carried over from previous air pollution episodes
with NO will lead to enhanced smog formation on subsequent days. This enhance-
ment in reactivity results both from the fact that these reactions form radicals
that initiate the transformations occurring in photochemical smog and from the
fact that these reactions convert NO to N02, which allows earlier formation of
0., and higher levels to be attained. It should be noted that this enhancement
will result even if all of the PAN reacts with NO emitted at nighttime, since
the NO conversion does not require sunlight; and at least some of the radicals
formed will be "stored" as nitrous acid, to be released when photolysis begins
at sunrise.
These results could have important implications regarding multiday photo-
chemical pollution episodes in which significant buildup of PAN is observed.
Under such conditions, carry-over of PAN may be a significant factor in promot-
ing ozone formation on subsequent days and may, in part, contribute to the
progressively higher 0, levels often observed during such episodes (Tuazon et
al., 1980; 1981b).
A second important role of PAN is its ability to contribute to the long-
range transport of NO . In the absence of significant levels of NO (i.e., in
}\
the cleaner troposphere) and in regions of lower temperature and in the upper
troposhere, when the thermal decomposition of PAN becomes unimportant, the
atmospheric lifetime of PAN will be determined by its reaction with OH radicals.
This reaction is sufficiently slow (Singh and Hanst, 1981) that PAN will
probably be long-lived and serve, hence, as a reservoir for odd nitrogen in a
manner analogous to HNO- (Aikin et al., 1983).
019AA/A 5-23 6/19/84
-------
5.3.6 Atmospheric Reactions of Hydrogen Peroxide
Although hydrogen peroxide formed in the gas phase from the reactions of
hydroperoxyl radicals plays a role in HO chemistry in the troposphere, and
J\
especially in the stratosphere (Crutzen and Fishman, 1977; Cox and Burrows,
1979), its major importance arises from its high solubility in water. The
latter ensures that a large fraction of gaseous H_0_ will be taken up in
aqueous droplets. Over the past decade, evidence has accumulated that H202
dissolved in cloud, fog, and rainwater may play an important role, and in
acidic droplets (i.e., pH <5) even a dominant role, in the oxidation of S00 to
"" £
H2S04 (Hoffman and Edwards, 1975; Penkett et al., 1979; Dasgupta, 1980; Martin
and Damschen, 1981; Graedel and Weschler, 1981; Chameides and Davis, 1982;
Calvert and Stockwell, 1983; Brock and Durham, 1984; Hoffman and Jacob, 1984;
Schwartz, 1984). Discussion of several proposed mechanisms for previous rate
studies of the oxidation of S(IV) by H202 are beyond the scope of this document,
but have recently been reviewed by several authors (e.g., Calvert, 1984).
Hydrogen peroxide may also play a role in the oxidation of N02 dissolved in
aqueous droplets, although relevant data are limited (Halfpenny and Robinson,
1952a,b; Anbar and Taube, 1954; Gertler et al., 1984) and additional research
is required. In addition to the direct oxidation of S02 and N0? dissolved in
aqueous droplets, the photolysis of H202 to produce aqueous OH radicals
(H2°2)aq + hv * 2(OH)aq (5-39)
can lead to oxidation rates of S(IV) that can be competitive with calculated
oxidation rates of S(IV) by (H«0,) and (0,) (Chameides and Davis, 1982).
-------
5.3.7 Atmospheric Reactions of Formic Acid
As a gas-phase species, formic acid (HCOOH) cannot strictly be defined as
a photochemical oxidant. Because it can be scavenged rapidly into water
droplets, however, it can potentially function as an oxidant in cloud water
and rain water. It can also be differentiated from other acids in that it is
formed readily from the reactions of the Criegee intermediates discussed
earlier and from the reaction of hydroperoxyl radicals with formaldehyde
(Calvert and Stockwell, 1983). The formation of other acids may be orders of
magnitude slower as the result of both the apparently lower rates of reaction
of H0? radicals with the higher aldehydes and the much lower atmospheric
concentrations of the higher aldehydes (Grosjean, 1982). Thus, formic acid is
an example of a non-oxidant or weak oxidant in the gas phase, being transformed,
upon incorporation in aqueous solutions, into an effective oxidizer of S(IV).
Formic acid (as well as acetic acid) has been identified among the acidic
components of rain (Galloway et a!., 1982). Although much uncertainty remains
concerning the quantitative role of HCOOH and the higher organic acids, they
potentially play a minor but still significant role in the acidification of
rain.
5.4 TYPE REACTIONS OF OZONE AND PEROXYACETYL NITRATE IN SOLUTION
Polluted urban air contains a number of photochemical oxidants, including
ozone, hydrogen peroxide, organic peroxides, singlet oxygen, peroxyacyl ni-
trates, and the various oxides of nitrogen (nitrogen dioxide, nitric oxide,
etc.). Since these substances vary widely in their relative abundance, per-
sistence, and reactivity with biological and nonbiological organic compounds,
their importance in smog-related oxidative damage also varies. The oxides of
nitrogen are found in high concentrations in polluted air and are the subject
of a separate air quality criteria document (U.S. Environmental Protection
Agency, 1982). Of the remaining compounds, the most important in terms of
abundance and potential reactivity in biological receptors are ozone and
peroxyacetyl nitrate (PAN). The purpose of this section is to provide a brief
summary of the chemistry of the reactions in solution-phase chemistry of ozone
and PAN with several of the more important types of functional groups found in
organic molecules, particularly those functional groups that occur in molecules
of biological interest.
019AA/A 5-25 6/19/84
-------
Type reactions of hydrogen peroxide are not given in this section.
Evidence noted in section 5.3 and presented subsequently, in section 5.5,
indicates that ambient air concentrations reported in the literature are
likely to be overestimations of the true concentrations present.
5.4.1 Ozone
As noted in the previous section, ozone is a very powerful oxidizing
agent and is capable of oxidizing most organic compounds, even alkanes, at
room temperatures (Pryor et a!., 1982). Despite its power as an oxidizing
agent, ozone exhibits considerable selectivity in its reactions; that is, some
functional groups are much more likely to be attacked and oxidized than are
others. As an example, the relative reactivity in solution of four types of
organic molecules with ozone has been reported by Pryor et al. (1982) to be:
alkene, 400,000; sec-alcohol, 7.0; benzene (aromatic), 0.1; alkane(s), 0.01 to
1.0. The most reactive groups of interest are discussed below.
5.4.1.1 Alkenes. Alkenes are especially susceptible to oxidation by ozone.
In fact, ozonolysis provides the basis for a standard test for identifying and
locating the position of carbon-carbon double bonds in molecules. Double
bonds occur in polyunsaturated fatty acids (PUFA) of the type that are found
in biological lipids in cell membranes; and the reaction of ozone with PUFA
has been studied in detail, since it is thought by some investigators that
this process is responsible for a large part of the cellular damage caused by
ozone (Pryor et al., 1976; 1981; 1982) (see chapter 10). The mechanism of the
reaction of ozone with alkenes is very complex and not entirely understood.
This area is the subject of a comprehensive review (Bailey, 1978).
The major reaction of ozone with non-hindered olefins is by a non-radical
process first described by Criegee in the early 1950s. According to this
scheme, ozone first reacts with the olefin by a 1,3-dipolar cycloaddition
reaction to form an unstable 1,2,3-trioxolane (the primary ozonide in equation
5-40) that undergoes rapid decomposition to give the carbonyl oxide and an
aldehyde or ketone (equation 5-41).
R,C = CR» + O, —»• R,C — CR, (5-40)
019AA/A 5-26 6/19/84
-------
o-o-o
T 1 +
R»C CR, »*R2C—O—O- + R,C =0 (5-41)
The reactions of the carbonyl oxide dictate the nature of the subsequent
reactions. If the carbonyl compound formed in equation 5-41 is a reactive
aldehyde, the carbonyl oxide and the aldehyde react to form the relatively
stable 1,2,4-trioxolane (the ozonide in equation 5-42). If the carbonyl
compound is a ketone, the carbonyl oxide dimerizes or polymerizes to yield a
range of types of peroxidic compounds. The carbonyl oxide also reacts with
active hydrogen compounds such as alcohols or acids to yield peroxidic products
(equation 5-43) or with water to give acids or aldehydes as the ultimate
products, as illustrated by equation 5-44. The reactions involving water also
yield hydrogen peroxide, a species that can cause additional oxidative damage.
f \
R,C—0—O~ + R'CHO »*R,C CHR' (5-42)
R'OHor
R'COiH
O
R— C
L
H
— OH
—OR'
0 — OH
or R-^C^C— R* (5-43)
I II
H O
—O—H
or RCOiH + H,O (5-44)
RCHO + H,OS
Most non-radical reactions of alkenes with ozone proceed through reactions
similar to the ones described above and generally result in the ultimate
cleavage of the double bond. If the double bond, however, is substituted with
bulky substituents and is sterically hindered, these cleavage products become
019AA/A 5-27 6/19/84
-------
less dominant and the formation of epoxides or other partial cleavage products
becomes more dominant (equations 5-45 and 5-46). The molecular oxygen released
in equations 5-45 and 5-46 is in the form of singlet oxygen, which can itself
cause further oxidative damage.
R,C = CR, + O, —•- R,C— CR, *-RsC— CR, + 'O, (5-45;
R,C « CR, + O, ^R,C—C—R ——»-R—-C—C—R + 'O, (5-46)
R R
For most alkenes, non-radical pathways of ozone oxidation dominate.
Ozone also reacts, however, with alkenes to produce free radicals. Even
though free-radical-generating pathways of reaction are minor when compared to
Criegee ozonolysis, these reactions may result in significant damage since
they can initiate autoxidation, which is a chain reaction. A suggested mechan-
ism (Pryor et al., 1982) for this type of reaction involves hydride abstraction
of an allylic hydrogen with ion recombination in the cage to give a hydrotrio-
xide, which then decomposes to form radicals (equation 5-47).
OOOH
R,C « CH - CR, -fr O, > [R,C = C-CR, OOOH] »-R,C » CH - CR,
0»
R,C = CH-CR, »• R,C = CH-CR, + HOO- (5-47)
5.4.1.2 Amines. Amines are, in general, close to alkenes in their reactivity
toward ozone, although protection is afforded when the ami no group exists as
an amide or salt, both of which are less susceptible to oxidation by ozone.
Whereas ozone acts as a 1,3-dipole in its reaction with alkenes, its attack on
amines is as an electrophile (equation 5-48). In general, ozone attack on
amines is by three competing reactions: (1) side chain oxidation (equation
019AA/A 5-28 6/19/84
-------
5-49); (2) amine oxide formation and transformations to nitro compounds (equa-
tion 5-50); and (3) dissociation to the nitrogen cation radical and ozonate
anion radical (equation 5-51) and the subsequent reaction of these radical
species (Bailey, 1982). Reaction 5-50 is important for primary and tertiary
amines, although in the case of primary amines the amine oxide formed is
unstable and undergoes further reactions. Reaction 5-51 is important mainly
for primary amines, while reaction 5-49 occurs with all amines bearing primary
alkyl groups. Reaction 5-52, in which radicals are produced, is important for
secondary amines. Reaction 5-52 also results in the formation of superoxide.
R,N: + 0=0-0- - ^ R.IM — O — O — O' (5-48)
r°->-
RiN ^CHtR' -r—»- R,N = CHR' «- R,N — CHR: (5-49)
+ V_^
R,N—0~O-r? ^R,N—O -I- 'Oa (5-50)
RjN—O—'O— O- -SB* R,N» + »O»O»O" (5-51)
R,NH + R2N—O—O—O v R2N-O» + O2~ + R»NH2 (5-52)
5.4.1.3 Sulfur Compounds. Like amines, sulfur compounds also undergo an
initial electrophilic attack by ozone (Bailey, 1982). Ozone reacts with
sulfides such as methionine to produce both sulfoxides and sulfones. The
ractivity of the sulfides decreases as the electron-withdrawing power of the
groups attached to the nucleophilic sulfur center increases. The mechanism
for the oxidation of a sulfide to the sulfoxide involves electrophilic ozone
attack on the sulfur, followed by loss of oxygen to give the sulfoxide (equa-
tion 5-53). A slower similar reaction then converts the sulfoxide to the
sulfone. A number of researchers have reported that less than 2:1 mole ratios
of ozone to sulfide are required for the sulfide-to-sulfone reaction and that
the more reactive the sul fide, the less ozone is required. One possible
019AA/A 5-29 6/19/84
-------
explanation for these facts is that the oxygen evolved in reaction 5-53 is
singlet oxygen, which could oxidize additional sulfide molecules.
= O + 'O2 (5-53)
While the main reaction of ozone with sulfides is apparently the attack
on sulfur with loss of oxygen to yield a sulfoxide, side-chain oxidationand
sissociation to sulfur cation radicals and the ozone anion radical also can
occur, although to a lesser extent than the analogous amine-ozone reactions.
Dissulfides undergo reaction with ozone to yield, in aqueous solution,
sulfonic acids, although the reactivity of disulfides is 40 to 50 times less
than that of thioethers. The proposed mechanism for the reaction of disulfides
is shown in equation 5-54. In agreement with this mechanism, Previero et al.
(1964) have reported the oxidation of cystine to cysteic acid, a reaction of
potential importance in the inactivation of enzymes by ozone.
-o—or
RS —SR + O3 »* RS —SR a=* RS — SR
?-o-o7
— 0
o
1!
— SR -*
II
O
o
II
Os— *-RS-
ii
O
•o
II
»o— ss
II
0
ko—oj
RS—O—SR'
o o
ii n ,
RS—O— SR + H2O 2RSO3 + 2H (5-54)
O O
Thiol groups are the most reactive of the sulfur functional groups,
having reactivities similar to those of isolated olefinic double bonds. The
reaction products for the attack of ozone on sulfhydryls are generally sulfonic
acids, although oxidation to the disulfide also has been reported (Mudd et
al., 1969). Since many enzymes rely on active cysteine residues for their
catalytic activity, sulfhydryl oxidation is a major mechanism for enzyme
inactivation by ozone.
5.4.1.4 Aromatics. The benzene ring is much less reactive toward ozone than
is the olefinic double bond. In solutions containing both olefinic and aromatic
unsaturation, ozone absorption is usually rapid until the olefinic bonds have
019AA/A 5-30 6/19/84
-------
entirely reacted, after which absorption becomes much slower. In compounds
containing both aromatic and olefinic portions, such as the indole ring of
tryptophan, the initial ozone attack occurs exclusively at the olefinic part
of the molecule. Benzenes and substituted benzenes do, however, react with
ozone, although slowly, with glyoxals and simple carboxylic acids reported as
the major products. Phenol has also been reported as a minor product from
benzene oxidation. The mechanisms for benzene ozonation are not known and are
difficult to study since the initial products are primarily olefinic and
undergo further ozonation reactions quite rapidly.
Phenol is somewhat more reactive toward ozone than benzene is and there
has been more interest in the ozonation of phenols than in any other benzene
derivatives because of the need to purify wastewater containing these compounds.
The initial attack of ozone on phenol probably involves both hydroxylation to
yield catechol (equation 5-55) and bond cleavage to give muconic acid (equation
5-56) (Yamamoto et al., 1979). Both of these products undergo further ozonations,
producing formic acid and carbon dioxide as the ultimate products from
reaction of phenol.
(5-55)
+ 03
CO,H
+ H2O
COtH
(5-56)
5.4-1.5 Aldehydes and Ketones. Aldehydes can react with ozone without the
involvement of oxygen (equation 5-57a), or ozone can initiate aldehyde autoxida-
tion (equations 5-57b, and 5-58 through 5-60). In either case, the initial
reaction produces acyl hydrotrioxides followed by decomposition to peroxides
and carboxylic acids.
RCHO + O,
'O,
HOO
(5-57a)
(5-57b)
019AA/A
5-31
6/15/84
-------
RCO,» + RCHO, » RCO,H + RCO (5-58)
0
• II
RCO + O, »• RC—OO« (5-59)
O O
RC—"OO« + RCHO "-RC —OOH + RCO (5-60)
Simple ketones react only very slowly, if at all, with ozone. Ketones
that do react give carboxylic acids as the main products.
Some of the possible reactions of ozone with important functional groups
have been described above. It should be remembered, however, that ozone is
capable of reacting with most organic molecules, even alkanes, although many
of these reactions are too slow to be important in the biosphere. Under some
conditions, however, ozone is converted into other oxidizing species. For
example, at pH greater than 9 in aqueous solution, ozone is rapidly converted
to hydroxyl radicals (HO ) that are less selective than ozone and react more
rapidly with organic substances in many cases (National Academy of Sciences,
1977). The conversion of ozone to superoxide (0?O also has been reported.
5.4.2 Peroxyacetyl Nitrate
Peroxyacyl nitrates make up another group of oxidizing substances present
in polluted air. Peroxyacetyl nitrate (PAN), CKLCOOpNO,,, is the most abundant
of these compounds and is present in smog at levels typically 5 to 20 percent
of the ozone level (chapter 6). Since at the higher levels PAN may cause crop
damage and eye irritation (chapter 1 and chapter 11), toxicological studies on
this compound have been fairly extensive. The chemistry of PAN in the gas
phase has been examined in a number of smog modeling studies; however, the
literature on the chemistry of PAN in solution, particularly in aqueous solu-
tion, is not extensive. Most of the chemical studies to be described are
included in a review published in 1976 (Mudd, 1976).
While PAN is relatively stable in the gas phase, in alkaline aqueous
solution it undergoes ready decomposition with quantitative formation of
nitrite ion (equation 5-61). Although the mechanistic details of this decompo-
sition are not fully understood, it appears to be a nucleophilic substitution
reaction that does not involve free radicals. At least some of the oxygen
produced is singlet oxygen, which may be responsible for some of the oxidative
reactions of PAN. In acidic aqueous solutions, PAN is somewhat soluble (Henry's
019AA/A 5-32 6/19/84
-------
Law coefficient of 5 at 10°C) and decomposes to nitrate, nitrite, and organic
fragments at a temperature-dependent rate (Holdren et al., 1984). The oxidation
of a number of molecular species by PAN has been examined and some of these
reactions are summarized below.
O
ii
CH3C—OONO2 + 2HO~ +* CH,COr+ NOr + O2 + H2O (5-61)
5.4.2.1 Alkenes. Alkenes are oxidized by PAN to epoxides with the production
of methyl nitrite and nitromethane by a reaction pathway suggested to involve
free radicals (equations 5-62 through 5-64).
o ,ov o
II / \ II
CH,C — OONOj + R2C = CR2 —+• R2C — CR2 + CH3C — O + NO2 (5-62)
O
il •
wriaw \J """^•^^ ^pf . -|- ^U» , — — v
(5-63)
NO2 + CH, —** CH3NO2 and CH3ONO (5-64)
5.4.2.2 Amines. Primary amines have been reported (Wendschuh et al., 1973)
to react rapidly with PAN to yield acetamides and nitrous acid (equation
5-65). The reaction of PAN with tertiary amines yields an unknown product
that produces chemiluminescence at K = 660 nm.
o o
II II
CH,C — OONO2 + RNH2 ' * CH.C — NHR + Oi + HNO2 (5-65)
5.4.2.3 Sulfur Compounds. Methionine is oxidized by PAN to methionine sulfo-
xide but with production of only negligible amounts of methionine sulfone.
The mechanism for this reaction has not been determined. The formation,
however, of free radicals has been suggested in the oxidation of dimethyl
sulfide to dimethyl sulfoxide in nonaqueous solution.
Disulfides react with PAN, although, in general, rather slowly, with large
excesses of PAN required for the reactions. The products from cystine oxidation
consist of cysteic acid along with trace amounts of cystine-S,S-dioxide.
Thiols are very reactive toward PAN. In the case of cysteine, two moles
019AA/A 5-33 6/19/84
-------
react per mole of PAN to produce disulfide. The reactivity is pH-dependent,
increasing greatly as the pH is raised, which suggests that RS is the reactive
species rather than RSH. The formation of S-acetyl products from the reaction
of PAN with thiols has also been reported.
5.4.2.4 Aldehydes. Wendschuh et al. (1973a) have reported that the addition of
an aldehyde to organic solutions of PAN results in the oxidation of the aldehyde
to the corresponding acid in approximately an 85 percent yield (equation 5-66).
9
CH,C — OOIMO2 + RCHO —»*
RCOZH + CH3HO2H + CH3NO» + CH.OIMO
+ CO2 + MINOR UNIDENTIFIED PRODUCTS (5-66)
The reaction occurs at an increased rate for aldehydes with electron-donating
R groups. The reaction is first order in PAN but the stoichiometry of reaction
is not simple, with one to three moles of aldehyde consumed for each mole of
PAN reacted. The aiuthors suggest a free-radical pathway initiated by PAN as
the most likely reaction mechanism.
Knowledge concerning the colution chemistry of peroxyacyl nitrates is
quite limited. It is known, however, that PAN reacts with many biochemically
important functional groups. The half-life of PAN in water is fairly short,
about 4.4 minutes at pH 7.2; therefore, PAN itself may not be able to react
with susceptible biological molecules before decomposition occurs, except
perhaps locally at the site of initial deposition or impact. The decomposition
products of PAN include nitrite ion and singlet oxygen, however, both of
which can cause oxidative damage. Thus, some of the toxicological effects of
PAN should possibly be attributed to the products of the decomposition of PAN
in an aqueous environment.
019AA/A 5-34 6/19/84
-------
5.5 SAMPLING AND MEASUREMENT OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
5.5.1 Introduction
Since the publication at the beginning of this decade of the first air
quality criteria document on ozone and other photochemical oxidants (1970),
there have been significant changes in the technology associated with measure-
ment of these pollutants in ambient air.
The chemiluminescent reaction of ozone (0_) with ethylene has been used
success-fully as the basic working principle in instrumentation whose response
is both specific for 0- and is linear with 0- concentrations over the range
usually found in ambient air. In general, chemiluminescence analyzers have
significantly improved both the time resolution and the ease with which moni-
toring for ozone can be routinely carried out.
Advances in electronics technology have allowed development of ultraviolet
absorption photometers of adequate precision for determining atmospheric
concentrations of 0~. The principle of the ultraviolet photometry has also
been applied to a new standard calibration procedure for 0- instruments. The
possible effect on the measurement of ambient 03 concentrations of this recent
change in calibration procedure is discussed in this section. Improved under-
standing of photochemical systems has resulted in an interest in the fate of
organic nitrogen species and of hydrogen peroxide in ambient atmospheres, such
that researchers have undertaken the development of refined methods for the
measurement of peroxyacyl nitrates and of hydrogen peroxide.
This section will describe analytical techniques for measurement of ozone
and other photochemical oxidants. Primary emphasis will be given to those
techniques presently considered most satisfactory for routine monitoring.
Since the original criteria document for photochemical oxidants emphasized
continuous methods for measuring "total oxidants," these techniques will also
be discussed briefly in order to place past measurements in perspective. For
the same reason, the relationship between measurements of ozone and total
oxidants will also be discussed.
In addition to developments in analytical techniques, EPA has codified
and instituted a formal nationwide program of quality assurance in the routine
operation of monitoring pollutants in ambient air. Some examples of these
procedures will be documented in this section as they apply to actual operation
of the analytical instrumentation. A detailed description of analytical proce-
dures, quality assurance procedures, and reporting requirements are contained
019QQ/A 5-35 6/19/84
-------
in Quality Assurance Handbook for Air Pollution Measurement Systems (U.S.
Environmental Protection Agency, 1977). Pertinent rules and regulations are
contained in the Federal Register (U.S. Environmental Protection Agency,
1979a; 1979b; 1979c).
An appreciation of some of the errors involved in present monitoring
techniques is important in evaluating the quality of ambient pollution data.
Three types of errors are discussed in this section: interferences, systema-
tic errors, and random errors.
The measurement, of individual pollutants in ambient air is complicated by
the presence of other airborne chemicals that may produce responses in the
measuring apparatus generally indistinguishable from those produced by the
pollutant being monitored. These spurious responses are known as "inter-
ferences." Extensive tests are conducted by the U.S. Environmental Protection
Agency and other laboratories, or both, on potential interferences in proposed
measurement techniques before they are considered suitable for routine monitor-
ing. In addition, researchers engaged in methods development or application
investigate interferences before reporting such methods in the literature.
This section describes reported interferences for the routine methods listed.
It should be noted, however, that not all potential interferences have equal
significance. Their magnitude will, in general, depend on the ambient concen-
trations of the interfering species, the inherent sensitivity of a given
procedure to spurious responses, and, in some cases, on details of the measuring
apparatus that may vary from instrument to instrument. An analytic technique
sensitive to interference may still be useful if the interfering species
occurs only in low concentrations in ambient air or may otherwise be accounted
for.
In addition to errors introduced by interferences, a given analytic
technique may be subject to systematic over- or underestimation of the pollut-
ant concentration, which affects the accuracy with which these concentrations
are known. Such errors are known as "biases." The assessment of the magni-
tude of such biases for a given analytical method generally requires extensive
testing, often by a number of laboratories sampling the same pollutant concen-
tration in ambient air (collaborative testing),
Random errors introduced by unknown factors such as variability in detailed
procedures used by different operators or sensitivity of the method to small
019QQ/A 5~36 6/19/84
-------
uncontrollable variations in operational parameters are generally known collec-
tively as the imprecision of the method. A measure of this type of error
often used is the standard deviation of a set of measurements. The precision
of a method is also often assessed in collaborative testing procedures. The
results of testing for these two error types are described in this section
where they are available.
5.5.2 Quality Assurance in Ambient Air Monitoring for Ozone
Quality assurance as defined by EPA rules and regulations consists of two
distinct functions. One is the assessment of the quality of monitoring data
by estimating their precision and accuracy. The other is the control and
possible improvement, depending on the results of the first function, of the
quality of the ambient air data by implementation of quality control policies,
procedures, and corrective actions.
Each quality control program, developed by the individual States and
approved by the EPA Regional Administrator, must include operational procedures
for each of the following activities:
1. Selection of methods, analyzers, or samplers (prescribed refer-
ence and equivalent methods for ambient air monitoring are
described elsewhere in this chapter);
2. Installation of equipment;
3. Calibration—Test concentrations for ozone must be obtained by
means of the ultraviolet (UV) photometric calibration procedure
described elsewhere in this chapter or by means of a certified
ozone transfer standard; permeation tubes for N02 must be
working standards that can be compared to Standard Reference
Material (SRM) from the National Bureau of Standards.
4. Zero/span checks and adjustments of automated analyzers;
5. Control checks and their frequency;
6. Control limits for zero, span, and other control checks, and
respective corrective actions when such limits are surpassed;
7. Calibration and zero/span checks for multiple range analyzers;
8. Preventive and remedial maintenance;
9. Quality control procedures for air pollution episode monitoring;
10. Recording and validation of data;
11. Documentation of quality control information.
019QQ/A 5-37 6/19/84
-------
A one-point precision check must be carried out at least every 2 weeks on
each automated analyzer used for ozone, using a precision test gas of known
concentration. Each calendar quarter, at least 25 percent of the analyzers
used by the State and Local Air Monitoring Stations (SLAMS) for monitoring
ozone must be formally audited by an independent operator by challenging with
at least one audit gas of known concentration in each of the four concentration
ranges. Similar requirements are set forth for monitoring networks designed
to assess Prevention of Significant Deterioration (PSD) requirements.
In addition to requirements and recommendations associated with the
selection, installation, and maintenance of monitoring equipment, the above-
cited Federal Register publications discuss certain design criteria for moni-
toring networks (SLAMS and the National Aerometric Monitoring Stations, NAMS:
see chapter 6). Included are requirements on siting of monitors in order to
obtain ozone concentrations that are representative of regions of varying
dimensions. For example, a "middle scale" monitor would represent conditions
close to sources of NO such that local ozone scavenging effects might be of
J\
significance. A "neighborhood scale" monitor, on the other hand, would be
located somewhere in a reasonably homogeneous urban subregion having dimensions
of a few kilometers. Other "scales" applicable to siting of ozone monitors
include urban scale, which would be used to estimate concentrations character-
istic of an area having dimensions between several and 50 kilometers or to
measure high concentrations downwind of an area with high precursor emissions;
and regional scale, used to typify concentrations over portions of a major
metropolitan complex up to dimensions of hundreds of kilometers. For ozone
SLAMS stations, applicable scales are middle, neighborhood, urban, and regional.
Requirements for NAMS stations for ozone are neighborhood and urban scale.
Two ozone NAMS stations are expected to be sufficient for each urban area:
one for specific transport conditions leading to high ozone; and the other for
monitoring peak concentrations relative to population exposure.
5.5.3 Sampling Factors in Ambient Air Monitoring for_0zone
Sampling factors may have a crucial effect on the quality and utility of
measurements both in ambient air and in controlled laboratory situations.
Sampling techniques and strategies must preserve the integrity of a representa-
tive fraction of ambient air and must be consistent with the specific purpose
of the measurement. In this section, the significance of some sampling factors
will be discussed briefly. For more detailed discussions of this subject, the
019QQ/A 5-38 6/19/84
-------
reader is referred to Ott (1977) and to reports prepared for EPA by the National
Academy of Sciences (Standing Air Monitoring Work Group, 1977; (National
Academy of Sciences, 1977a; National Academy of Sciences, 1977b).
5.5.3.1 Sampling Strategies and Air Monitoring Needs. Air monitoring data
relevant to assessing ambient 0_ or oxidant levels are collected for a variety
of specific needs, including:
1. Data to be used in trend analysis as indicators of the state of
attainment of ambient air quality standards.
2. Data to be used in development of 0- control strategies and
evaluation of their effectiveness.
3. Data to be used in the development and validation of air quality
simulation models capable of application to the 03 problem.
4. Data to be used in investigation of causes of the ozone problem
both in general and in specific localities.
5. Data to be used in special research studies such as the effects
of ambient air pollution on human health and welfare.
Each specific purpose or need requires special considerations in designing
a suitable air sampling strategy. For example, several years of 0- data might
be required for the adequate assessment of trends that resulted from the
application of a particular control strategy rather than trends that resulted
from chance local meteorological conditions. In contrast, the validation of
an air quality simulation model might require only a few carefully chosen days
of very detailed measurements of 0-, hydrocarbons, and NO , as well as detailed
»3 X
meteorological data and time-varying emissions along the trajectory of the air
parcel in question.
5.5.3.2 Air Monitoring Site Selection. Ozone in the lower troposphere is a
product of photochemical reactions that involve sunlight, hydrocarbons, and
oxides of nitrogen. In typical urban atmospheres, ozone precursors react to
produce ozone at such a rate that the 0- reaches its daily peak level in the
middle of the day at locations downwind from the source"intensive center-city
area. Thus, if peak 03 concentrations are to be measured, monitoring stations
should, in general, be located downwind from city centers. This downwind
distance may be on the order of 15 to 30 kilometers (9 to 19 miles), depending
on predominant wind patterns in the area (Standing Air Monitoring Work Group,
1977). It should be emphasized, however, that this distance may be highly
019QQ/A 5-39 6/19/84
-------
area-specific. A study (Wolff and Lidy, 1978) of emissions originating in the
Houston area suggested that, in this case, the distance may be considerably
greater than the figures quoted above.
Once a station is located, additional sampling considerations arise
because of the chemical reactivity and instability of the 0, molecule. Ozone
reacts extremely rapidly with NO and with some hydrocarbon compounds, including
most of those emitted by vegetation. Also, 03 decomposes readily on contact
with the surface of many materials. Consideration of these effects led to the
development of specific criteria for locating an 0_ monitoring station (Stand-
ing Air Monitoring Work Group, 1977; National Academy of Sciences, 1977b).
Briefly, the inlet of the sampling probe of the ozone analyzer should be
positioned 3 to 15 meters (10 to 49 feet) above ground, at least 4 meters
(13 feet) from large trees, and 120 meters (349 feet) from heavy automobile
traffic. Sampling probes should be designed so as to minimize 0~ destruction
by surface reaction or by reaction with NO.
Another consideration that has significance for the selection of sites
for air monitoring stations is the fact that ambient monitoring data, as
routinely obtained, have limited validity as absolute measures of air quality.
This limitation arises from the fact that, at ground level, the ambient atmos-
phere is inhomogeneous as a result of a continuous influx of fresh emissions,
incomplete mixing, and destruction of 0~ by fresh and unreacted emissions and
destruction on surfaces. In view of such inhomogeneity, monitoring data from
a fixed network provide measures of air quality at a discrete number of loca-
tions but may not detect temporal and spatial variations in ozone concentra-
tions of a localized nature. This problem can be alleviated by use of a
greater density of monitoring stations or by use of a validated air quality
model. Such models are capable of helping quantify the emission, dispersion,
and chemical reaction processes. Their outputs can provide data on the distri-
bution of air quality concentrations between widely spaced ambient monitors.
The emphasis in this section has been on a brief discussion of sampling
strategies. The word sampling is also widely considered to mean those tech-
niques that are required to obtain a parcel of air that is representative of
the polluted atmosphere, and to maintain its integrity until a measurement of
concentration has been carried out. Considerations relating to this meaning
of sampling are discussed as appropriate in the following sections on measure-
ment techniques.
019QQ/A 5-40 6/19/84
-------
5.5.4 Measurement Methods for Total Oxidants and Ozone
5.5.4.1 Total Oxidants. Although ozone was first unambiguously identified in
polluted atmospheres by spectroscopic techniques (Stephens et al., 1956), the
earliest procedures for routinely monitoring 0- and other oxidizing species in
the atmosphere were based on iodometry. lodometric techniques are inherently
non-specific in that a variety of oxidizing species in addition to 0- may be
positive interferences, producing iodine in solution; whereas reducing agents
are negative interferences. Thus, the name "total oxidants" was coined because
the technique responded not only to 03 but to other oxidants such as peroxides,
peroxyacetyl nitrate (PAN), and nitrogen dioxide (N02). Total oxidants are
then actually defined by the particular iodometric procedure used, since the
response to the various oxidizing species present will depend on the details
of the procedure. This will be more evident when interferences are discussed
below. The use of the word "total" is in itself a misnomer. The measurement
does not reflect a sum of the oxidizing species present because the various
oxidants present in the atmosphere react to produce iodine at different stoi-
chiometries and different rates. In spite of these difficulties, the measure-
ment of total oxidants was a useful method for characterization of the atmos-
phere because of its correlation with the principal oxidant, 0,; and, conse-
quently, there is a large oxidant data base available. For these reasons, the
two principal methods used for monitoring total oxidants are discussed in more
detail below.
The bulk of the total oxidants data base was obtained by the use of two
types of continuous monitoring instruments. In both types, an air sample is
continuously scrubbed by an aqueous reagent containing potassium iodide (KI).
In colorimetric oxidant instruments, the iodine is measured photometrically by
ultraviolet absorption. In the other common type instrument, the iodine pro-
duced is measured by electrochemical means. Both of these instruments are
discussed and compared in more detail below. Many other chemical techniques
for oxidants have been proposed and in some cases applied, but for these
reference is made to the original literature (Hodgeson, 1972a; Katz, 1976).
The first colorimetric analyzers were patterned after the instrument de-
scribed by Littman and Benoliel (1953). In this and the commercial versions,
the air sample flow and the liquid reagent flow were mixed countercurrently in
a contacting column. The reagent contained 20 percent KI (later 10 percent) and
Q19QQ/A 5-41 6/19/84
-------
was maintained at pH = 6.8 with a phosphate buffer. Ozone and other oxidants
produce iodine according to the reaction:
03 + 3I~ + H20 = I3~ + 02 + 20H (5-67)
The tri-iodide ion exhibits an intense absorption maximum at 352 nm,
which is used to monitor continuously the I3 concentration. These instruments
operate over a range of about 0.01 to 1 ppm and have a 90 pecent full-scale
response time of a few minutes. Interferences and compensation for them are
discussed below.
The electrochemical oxidant sensors are more correctly called amperometric
sensors in that an electrochemical current is measured rather than an absolute
transfer of charge. Two types of electrochemical cells have been used, an
electrolytic cell (Brewer and Mil ford, 1960; Mast and Saunders, 1962) and a
galvanic cell (Hersch and Deuringer, 1963). Of these, the Brewer cell has been
by far the most frequently used in the commercial Mast Meter version. In
the Brewer cell, sample air and reagent (2 percent KI, 5 percent KBr buffered
at pH = 6.8 with phosphate buffer) flow concurrently over a wire helix cathode.
A polarization voltage of 0.24 volt applied between the cathode and anode pro-
duces a thin layer of hydrogen at the cathode and a small but constant polari-
zation current that represents zero 03 concentration. Ozone absorbed in solu-
tion produces the tri-iodide ion, which reacts with and removes hydrogen, tem-
porarily depolarizing the cathode. A current to repolarize the cathode then
flows through the external circuit. The magnitude of this differential current
is proportional to the hydrogen removed and is recorded as a function of 03
concentration. If contact efficiences were 100 percent and reaction stoichio-
metries well-established, the absolute coulometric current should be equivalent
to the concentration of 03 absorbed. In practice, it is necessary to calibrate
these analyzers with standard 03 samples (see below). The coulometric yield of
the Brewer cell has been reported to be about 75 percent (Wartburg et al., 1964).
The normal operating range for the Brewer cell in atmospheric monitoring is 0.01
to 1 ppm, but these sensors will also work at higher concentrations.
The interferences for both colorimetric and amperometric 03 analyzers are
other oxidizing and reducing species in the atmosphere. The major oxidant in
ambient air by far is 03 (chapter 6); and the other oxidants present, except
019QQ/A 5-42 6/19/84
-------
N02, are considered part of the total oxidants measured rather than inter-
ferences. The dominant reducing interference is S0?. Thus, the major inter-
ferences present in ambient air and considered below are N02 and S02.
The magnitude of the N02 interference is variable and depends upon con-
tact design and KI concentration (Tokiwa, 1972; Intersociety Committee, 1970).
For the Brewer amperometric cell, the interference from N0? is only 6 percent
of an equivalent concentration of 03 (Tokiwa, 1972), presumably because of the
low KI concentration and short contact time. For the colorimetric oxidant
analyzer, NG2 interference equivalents are about 20 percent for 10 percent KI
solution and vary from 20 to 32 percent for 20 percent KI depending on 0,
•J
concentration (Tokiwa, 1972). The method used for compensating for N0? in-
terference is simultaneous measurement of N0? and correction of the correspond-
ing oxidant reading. For this reason, the terms "corrected" or "adjusted"
oxidant are often used.
The interference from S02 is quantitative for both colorimetric and elec-
trochemical oxidant measurements, with one mole of S0? consuming one mole of
tri-iodide ion. If the S02 concentration is less than that of total oxidant
and S02 is simultaneously measured, the "adjusted" oxidant reading may also
contain a correction for S0?. This was the procedure previously applied in
the older aerometric data for California, where S02 levels were inherently low
because of low-sulfur fuel requirements. For many areas of the East Coast and
Midwest, such a correction was not possible and preferential SO- scrubbers
were used. The most common of these consisted of chromium trioxide impregnated
on glass fiber filters (Saltzman and Wartburg, 1965) or an inert chromatography
support (Mueller et al., 1973). These scrubbers may be effective in the hands
of skilled operators but their use is not without problems. Among these pro-
blems are partial oxidation of NO to N02 and of H2S to S0?, and partial removal
of 0. when the scrubber is wet or contaminated (Hodgeson, 1972a).
5.5.4.2 Ozone
5.5.4.2.1 Gas-phase chemi'luminescence. Many of the 0, oxidation reactions
are sufficiently energetic that they produce electronically excited products,
intermediates, or reactants, which in turn may chemiluminesce (Zocher and
Kautsky, 1923; Bowman and Alexander, 1966). Although well known for many
years, such reactions were not applied to chemical analysis until the 1960s.
In 1965, Nederbragt reported a detector that employed chemiluminescence from
the reaction of 03 with ethylene for measurement of 0, in the vicinity of
large accelerators (Nederbragt et al., 1965; Warren and Babcock, 1970).
019QQ/A 5-43 6/19/84
-------
Applications to atmospheric analysis were a natural consequence (Stevens and
Hodgeson, 1973). When it promulgated the first standards for the criteria
pollutants in 1971 (U.S. Environmental Protection Agency, 1971a), EPA also
published reference methods for measurement of these pollutants. The reference
method for (L to be used by EPA and the states for determining compliance was
the 03-ethylene chemiluminescence method. Appendix D of 40 CFR, Part 50,
describes the principle of the method, including a method of calibration (U.S.
Environmental Protection Agency, 1971b). Since then, the measurement principle
has remained the same but calibration procedures have undergone extensive
revision as discussed below (section 5.5.5). It is also noteworthy that the
reference method is specific for 03, whereas the data used for establishing
the standard were based on measurement of total oxidants. This issue is
addressed in section 5..
A flow of sample air (1 to 5 L/min) containing 03 and a small flow of
pure ethylene are mixed at atmospheric pressure in a small reaction chamber
closely coupled to the photocathode of a photomultiplier tube. The reaction
between 0, and ethylene produces a small fraction of electronically excited
O
formaldehyde. Chemiluminescence from this excited state results in a broad
emission band centered at 430 nm (Finlayson et a!., 1974). The emission
intensity that is monitored is a linear function of 0- concentration fran
0.001 to greater than I ppm. The relation between intensity and concentration;
i.e., instrument calibration, must be determined for each instrument with
standard concentrations of 0, in air. The minimum detection limit and the
o
response time are functions of detector design. Detection limits of 0.005 ppm
and response times of less than 30 seconds are readily attained, however, with
modest design features. For example, cooling the photomultiplier improves the
sensitivity but is not normally required. There are no known interferences
among the common atmospheric pollutants. There have been reports of a positive
interference when 0, is measured in the presence of water vapor; i.e., a
signal enhancement of 3 to 12 percent in high humidity as opposed to measure-
ment of the same concentration of 03 in dry air (California Air Resources
Board, 1976). Where this may be a real problem, it can be minimized by per-
forming calibrations with humidified air. Finally, in order to obtain accept-
able measurement precision and constant span, analyzers must contain means for
maintaining constant air and ethylene flow rates.
Under Title 40, Code of Federal Regulations, Part 53, EPA has published
ambient air monitoring reference and equivalent methods (U.S. Environmental
019QQ/A 5-44 6/19/84
-------
Protection Agency, 1975). This regulation prescribes methods of testing and
performance specifications that commercial analyzers must meet in order to be
designated as a reference method or as an equivalent method. An analyzer may
be designated as a reference method if it is based on the same principle as
the reference chemiluminescence method and meets performance specifications.
An automated equivalent method must meet the prescribed performance specifi-
cations and show a consistent relationship with a reference method. These
specifications for 0. analyzers are listed in Table 5-7. Commercial analyzers
that have been designated as reference or equivalent methods are listed in
Table 5-8. Information concerning the applications supporting the designation
of analyzers as reference or equivalent methods may be obtained by writing the
U.S. Environmental Protection Agency, Environmental Monitoring and Support
Laboratory, Research Triangle Park, NC 27711.
5.5.4.2.2 Gas-solid chemiluminescence. The first chemiluminescence technique
for 0, was developed by Regener for stratospheric measurements (Regener, 1960)
•3
and later for measurements in the troposphere (Regener, 1964). The chemi-
luminescence was obtained from the reaction of 0» with Rhodamine-B adsorbed on
activated silica gel. The emission is in the red region of the visible and is
characteristic of the fluorescence spectrum of Rhodamine-B. The intensity is
a linear function of 0, concentration, the minimum detection limit can be
lower than 0.001 ppm, and no atmospheric interferences have been observed
(Hodgeson et al., 1970). The technique is, in fact, more sensitive than the
gas-phase Nederbragt method and does not require critical control of flow
rate. It had the disadvantage in the original analyzer built, however, that
frequent and periodic internal calibration cycles were required to compensate
for changes and decaying sensitivity of the surface of the detector (Regener,
1964; Hodgeson, 1970).
Improvement was made in the stability of the surface response in a modi-
fication added by Bersis and Vassiliou (1966), in which gallic acid is also
adsorbed on the surface in excess. The 0- apparently reacts with and consumes
the gallic acid rather than Rhodamine-B. An energy transfer step to Rhodamine-
B subsequent to the initial reaction results in the same chemiluminescence
from the dye compound, which is now no longer consumed. A commercial analyzer,
Phillips Model PW9771, is based on this principle and has been designated as
an equivalent method under EPA regulations.
5.5.4.2.3 Ultraviolet photometry. Ozone has a moderately strong absorption
band in the ultraviolet (UV), with a maximum very near the mercury 254 nm
019QQ/A 5-45 6/19/84
-------
TABLE 5-7. PERFORMANCE SPECIFICATIONS FOR AUTOMATED METHODS
Performance parameter
Range
Noise
Lower detectable limit
Interference equivalent
Each interferant
Total interferant
Units
ppm
ppm
ppm
ppm
ppm
Specification
0 to 0.5
0.005
0.01
±0.02
0.06
Zero drift, 12 and 24 hour ppm ±0.02
Span drift, 24 hour
20% of upper range limit
80% of upper range limit
Lag time
Rise time
Fall time
Precision
20% of upper range limit
80% of upper range limit
percent
percent
minutes
minutes
minutes
ppm
ppm
±20.0
±5.0
20
15
15
0.01
0.01
Source: U.S. EPA, 40 CFR, Part 53.
emission line. This band is essentially a continuum near 250 nm. The molar
absorption coefficient at the mercury line has been measured by several investi-
gators with good agreement and has an accepted value of 134 M cm (base 10)
at 0°C and 1 atm (Hampson et al., 1973). The UV absorption at 254 nm has long
been used as a method of measuring gas-phase 0™ in fundamental chemical and
physical studies. Some of the very first atmospheric 0_ measurements were, in
fact, made by UV photometry; e.g., the Kruger Photometer. These early instru-
ments and the problems with their use are described more completely in the
first criteria document for photochemical oxidants (U.S. Department of Health,
Education, and Welfare, 1970). The major problem with the older photometric
instruments was the large imprecision involved in measuring the very small
absorbance values obtained. For example, an absorbance value of 0.005 is a
typical minimum for most conventional photometric measurements. At this
019QQ/A 5-46 6/19/84
-------
TABLE 5-8. LIST OF DESIGNATED REFERENCE AND EQUIVALENT METHODS
Identification and source
Fed. Register notice
Vol.PageDate
Designation
(E=equivalent,
Preference)
Beckman Model 950A Ozone
Analyzer
Beckman Instruments
2500 Harbor Boulevard
Fullerton, CA 92634
Bendix Model 8002 Ozone
Analyzer
The Bendix Corporation
Post Office Drawer 831
Lewisburg, WV 24901
Columbia Scientific Industries
Model 2000 Ozone Meter
11950 Jollyville Road
Austin, TX 78759
Dasibi Model 1008-AH Ozone
Analyzer
Dasibi Model 1003-AH
1003-PC or 1003-RS Ozone
Analyzers
Dasibi Environmental Corp.
616 E. Colorado Street
Glendale, CA 91205
MEC Model 1100-1 Ozone Meter,
MEC Model 1100-2 Ozone Meter,
or MEC Model 1100-3 Ozone Meter
Columbia Scientific Industries
11950 Jollyville Road
P.O. Box 9908
Austin, TX 78766
Meloy Model OA 325-2R Ozone
Analyzer
Meloy Model OA 350-2R Ozone
Analyzer
Columbia Scientific Industries
11950 Jollyville Road
Austin, TX 78759
Monitor Labs Model 8810
Photometric Ozone Analyzer
Monitor Labs, Incorporated
10180 Scripps Ranch Boulevard
San Diego, CA 92131
42
28571 6/3/77
R
41 5145 2/4/76
45 18474 3/21/80
44 10429 2/20/79
48 10126 3/10/83
42 28571 6/3/77
E
E
41 46647 10/22/76
42 30235 6/13/77
40 54856 11/26/75
40 54856 11/26/75
R
R
46 52224 10/26/81
019QQ/A
5-47
6/19/84
-------
TABLE 5-8. LIST OF DESIGNATED REFERENCE AND EQUIVALENT METHODS (continued)
Identification and source
Fed. Register notice
Vol. Page Date
Designation
(E=equivalent,
R=reference)
Monitor Labs Model 8410E 41
Ozone Analyzer
Monitor Labs, Incorporated
4002 Sorrento Valley Boulevard
San Diego, CA 92121
PCI Ozone Corporation Model 47
LC-12 Ozone Analyzer
PCI Ozone Corporation
One Fairfield Crescent
West Caldwell, NJ 07006
Philips PW9771 0. Analyzer 42
Philips Electronic Instruments, 42
Incorporated
85 McKee Drive
Mahwah, NJ 07430
Thermo Electron Model 49
UV Photometric Ambient 0- 45
Analyzer
Thermo Electron Corporation
Environmental Instruments Division
108 South Street
Hopkinton, MA 01748
53684 12/8/76
13572 3/31/82
38931 8/1/77
57156 11/1/77
57168 8/27/80
value, a photometer pathlength of almost 1 km would be required to measure a
concentration of 0.01 ppm, the minimum value specified for acceptable automated
methods (Table 5-7).
This problem of adequate sensitivity with moderate pathlengths has been
overcome by modern digital techniques for measuring small absorbancies. The
first instrument of this new generation of photometers was marketed by Dasibi
of Glendale, California, in the early 1970s. The details of this instrument
have been described by Bowman and Horak (1972). Several other commercial in-
struments have since been marketed and, along with the Dasibi, have been
designated as equivalent methods by EPA (Table 5-8). All of these instruments
operate effectively as double-beam digital photometers. A transmission signal
is averaged over a finite period of time with 03 present and is compared to a
similar transmission signal obtained through an otherwise identical reference
019QQ/A
5-48
6/19/84
-------
air stream from which the CL has been preferentially scrubbed. The electronic
comparison of the two signals can be converted directly into a digital display
of 00 concentration. In the Dasibi instrument, the 0, is removed during the
3 »*
reference cycle by a manganese dioxide scrubber, through which other species
that absorb at 254 nm should pass unaffected.
The UV photometric technique has the advantages, like gas-solid chemilum-
inescence, that a reagent gas flow is not required and that sample air flow
control is not critical. In addition, the measurement is in principle an
absolute one, in that the concentration can be computed directly from the
measured absorbance since the absorption coefficient and the pathlength are
known. This capability is used extensively for the purpose of Og calibration
as discussed in section 5.5.5. Commercial UV photometers for 03 can serve a
dual function as a secondary standard for 0- calibration, if they are in turn
calibrated against a primary UV standard such as those provided by the National
Bureau of Standards (NBS) (Bass et at., 1977). In practice, UV photometric
analyzers that are used for monitoring 0_ concentrations in the atmosphere are
calibrated with standard 03 samples in order to compensate for possible 03
losses in the sampling and inlet systems. A UV photometric analyzer has the
potential disadvantage that any molecular species that absorbs at 254 nm
(e.g., SO,,, benzene, mercury vapor) and that may also be removed along with 0-
during the reference cycle can interfere. Documentation of such interference
during atmospheric monitoring is lacking at present.
5.5.5 Generation and Calibration Methods for Ozone
Unlike the other criteria pollutants, 0^ is a thermally unstable species
that must be generated in situ during the calibration of analyzers used for
atmospheric monitoring. This creates special requirements not encountered
with other pollutants and thus this section deals with means for generating
dynamic air streams containing stable 0- concentrations and chemical and
physical means for absolute measurement of these concentrations.
5.5.5.1 Generation. Ozonized samples of air can be produced by a number of
means, including photolysis (Brown and Milford, I960), electrical discharge
(Toyami and Kobayashi, 1966), and radiochemical methods (Steinberg and Dietz,
1969). Electrical discharges are useful for producing high concentrations of
0, in air for other applications; e.g., 0, chemistry. Radiochemical methods
would be ideal except for their cost and required safety features. By far the
019QQ/A 5-49 6/19/84
-------
most common method, however, for generating low concentrations of 0- in air in-
w
volves the photolysis of molecular oxygen.
02 + hY(A<200 nm) * 20 (5-68)
0 + 02 + M (M = N2 or 02) = 03 + M (5-69)
One of the most common photolytic generators uses a mercury vapor 6- or 8-inch
PenRay photolysis lamp positioned parallel to a quartz tube through which a
controlled rate of air flows. The 03 concentration is simply varied by means
of an adjustable and calibrated mechanical sleeve placed over the lamp envelope
(Hodgeson et a!., 1972b) or by varying the voltage or current supplied to the
lamp.
5.5.5.2 Calibration
5.5.5.2.1 KI procedures: original EPA reference method. The output of
photolytic 03 generators can provide air samples containing stable CL concentra-
tions over a considerable period of time with careful control of flow rate,
lamp voltage, temperature, and pressure. It is necessary to calibrate these
generators periodically with an absolute reference method. Prior to 1975,
there were as many as seven different calibration methods for 03 employed to
varying extents in this country (National Academy of Sciences, 1977). A good
part of the variability in older data may result from biases in calibration
procedures. In an attempt to standardize the methodology, EPA published a
reference calibration procedure with the reference method in 1971 when the
oxidant (as 03) standards were promulgated (U.S. Environmental Protection
Agency, 1971a). This method was the 1 percent neutral buffered potassium
iodide (NBKI) procedure, a technique that had been used by EPA and other
agencies for some time.
During the early 1970s, it became evident that there were serious defi-
ciencies with the NBKI reference method. Several problems with the NBKI pro-
cedure, summarized by a joint EPA-NBS workshop in 1974 (Clements, 1975), in-
cluded the gradual continued release of iodine after sampling, variable results
obtained with different types of impingers, reagent impurities, and a positive
bias when compared to other 0_ measurement methods. In 1973, a significant
bias was observed between calibration results obtained with the 1 percent un-
buffered KI method used by the Los Angeles Air Pollution Control District
(LAAPCD) and the 2 percent NBKI procedure used by the California Air Resources
019QQ/A 5-50 6/19/84
-------
Board (GARB). (The latter technique is very similar to the original 1 percent
NBKI EPA reference method). As a result, an interagency collaborative study
was undertaken to intercompare the LAAPCD, CARB, and EPA methods, using UV
photometric 0, measurements as the reference. The results of this study
«5
(Demore et al., 1976) demonstrated the positive bias of the NBKI methods. The
results obtained by the EPA and CARB NBKI methods were higher by 15 to 25
percent than those obtained by UV photometry; whereas the results obtained
with the unbuffered KI method were 4 percent lower and showed considerable
scatter. Concurrent with and after these earlier reports, a large number of
individual studies ensued on the evaluation of KI methods and their intercom-
pan' son with UV photometric 0- measurements and CL measurements by gas-phase
titration (GPT) with standard nitric oxide (NO) samples. The history of these
studies will not be reviewed here since they were presented in the previous
criteria document for ozone and other photochemical oxidants (U.S. Environmental
Protection Agency, 1978) and have been reviewed by Burton et al. (1976). The
major conclusions from these studies are presented below.
1. Results obtained by NBKI procedures are higher than those obtained
by UV photometry or gas-phase titration by 5 to 25 percent,
depending on details of the procedure.
2. When 0- is measured in the presence of humidified air, NBKI
results tend to be even higher by another 5 to 10 percent (e.g.,
California Air Resources Board, 1975). The reason for this
apparent moisture effect is not known.
3. In general, NBKI techniques are subject to large imprecision
because of procedural variation.
Because of these difficulties, EPA published a notice in the October 6,
1976, Federal Register of its intent to evaluate alternative calibration pro-
cedures and to replace the NBKI procedure with one of four alternative pro-
cedures (U.S. Environmental Protection Agency, 1976). These four alternate
procedures were based on (1) UV photometry, (2) GPT with excess NO, (3) GPT
with excess 0,, and (4) a KI technique that uses boric acid as buffer (BAKI).
In subsequent studies (Rehme et al., 1981), UV photometry was considered to
give superior results in terms of accuracy, precision, and simplicity of use;
and in 1979 Appendix D of 40 CFR, Part 50, was amended to designate UV photo-
metry as the calibration procedure for 0., reference methods (U.S. Environmental
Protection Agency, 1979).
019QQ/A 5-51 6/19/84
-------
Although NBKI methods are no longer used in this country for the purpose
of calibration, there is a considerable data base available on health and wel-
fare effects, as well as atmospheric chemistry and monitoring, that is based
on these methods as standards. Therefore, it is important to consider how
these data may be evaluated and compared to newer effects and aerometric data
based on the new UV calibration standard. Since a systematic bias is known to
exist, between calibrations by KI methods and UV photometric methods, it should
be possible, in principle, to apply correction factors to convert from a KI
reference to a UV photometric reference. There are several problems inherent
in attempting such corrections, however. A fairly wide range of variations has
been reported in the literature on the comparison of KI and UV photometric
measurements. In addition, some of the studies report a significant intercept
in the linear correlation of KI and UV photometric data. For example, in the
interagency collaborative study (Demore et a!., 1976), the relation between
EPA NBKI data and UV photometric data fH the following linear equation,
C°3]EPA = 1"24 [03]UV " °'035 (5"70)
(Standard Error = ±0.013)
when the intercept is expressed in units of ppm. As a result, the ratio of
KI/UV measurements varied from approximately 1.0 for the lowest concentration
measured (0.1 ppm) to 1.20 for the highest concentration measured (0.8 ppm).
It is inappropriate, however, to apply a general correction factor for the
intercept because the presence and magnitude of such an intercept will be
strictly dependent upon procedural variations during calibration; e.g., KI
reagent purity (Clements, 1975; Beard et a"!., 1977).
It should be possible to apply a correction factor related to the slope
of the equation, since the slope determines the absolute relation between
simultaneous measurements in the absence of effects leading to non-zero inter-
cepts. Even the magnitude of the slope, however, can depend to some extent on
procedural variables. As discussed previously, the presence of moisture in
the calibration air increases the magnitude of the bias and the slope. Fortu-
nately, both the CARB and the LAAPCD procedures called for the consistent use
of humidified air, whereas the EPA reference method prescribed the use of dry
air. In addition, the elapsed time between sample collection and color meas-
urement will also affect the magnitude of the slope because of the slow libera-
tion of iodine after sampling (Clements, 1975; Beard et a!., 1977; Hodgeson,
019QQ/A 5-52 6/19/84
-------
1976). Other unknown experimental factors may also influence the slope; e.g.,
impinger design (Clements, 1976; Beard et al. , 1977).
In view of these difficulties, it is tempting to ignore the problem and
simply state that it is impossible or without meaning to attempt to apply
correction factors. Nevertheless, there are considerable effects and aero-
metric data bases available that should have a positive calibration bias
compared to newer data based on the UV photometry standard. Therefore, an
assessment has been made of the previous KI versus UV intercomparisons and
recommendations are given in Table 5-9 for correction factors to apply to
calibration data for conversion from UV to a KI reference or vice versa. It
should be emphasized that these factors could validly be applied to correct
for a calibration bias only and can not be applied for comparison of data
where other effects are present; e.g., the comparison of oxidants versus 0»
data where the effects of oxidizing or reducing interferences must be consid-
ered. In this assessment, consideration was given only to those studies in
which the KI procedure was compared directly to UV photometry. Several studies
have compared KI measurements to GPT measurements, but there have been some
differences observed in the intercomparison of GPT and UV measurements. The
recommended value for data based on the CARB method assumes the use of humidi-
fied air. The value recommended for the EPA method assumes that dry air was
used and that color measurement was made immediately after sample collection.
The uncertainties assigned reflect the fact that a range of values has
been reported for the ratios in previous studies. No uncertainty value is
reported for the LAAPCD method because only one intercomparison (Demore et
al., 1976) has been reported and no correction should be attempted for this
method. Finally, whenever any attempt is made to convert from one data base
TABLE 5-9. FACTORS FOR INTERCOMPARISON OF DATA CALIBRATED BY
UV PHOTOMETRY VERSUS KI COLORIMETRY
Calibration method Ratio, KI/UV
EPA, 1% NBKI 1.12 ± 0.05
CARB, 2% NBKI 1.20 ± 0.05
LAAPCD, 1% UKI 0.96a
Correction for this method not recommended; only one intercomparison has been
reported.
019QQ/A 5-53 6/19/84
-------
to another, these uncertainties must be added by conventional error propagation
techniques to the uncertainty inherent with the original measurement.
5.5.5.2.2 Ultraviolet photometry method, A major reason UV photometry was
designated as the calibration procedure was the excellent precision of the
photometric measurement. In the collaborative study by Rehme et al. (1981),
measurement with ten individual UV photometers gave only a 3.4 percent varia-
bility when compared to a reference measurement system. Other significant
factors in the selection of UV photometry were the inherent simplicity of UV
photometric measurements and the ready availability of commercial instruments
that can also serve well as transfer standards between laboratory photometers
and field 03 analyzers. (See National Academy of Sciences, 1977, and McElroy,
1979, for a discussion of transfer standards.)
It was also presumed that UV photometry gives more accurate results,
since the accuracy is determined primarily by the 03 absorption coefficient,
which is well known (Hampson et al., 1973; Oemore and Patapoff, 1976). The
lack of any significant bias between the ten UV photometers and the reference
system in the collaborative study was to be expected since the reference
system was itself calibrated against a standard photometer. Although there is
little doubt that the accuracy of 0- measurements has been significantly
improved by conversion to the UV basis, some question still exists regarding
the absolute relation between 0, measurements by UV photometry and 0_ measure-
ments by GPT measurements based on either an NBS standard reference material
(SRM) nitric oxide (NO) gas cylinder or an N02 SRM permeation tube. These
intercomparisons have been made by several investigators over the past 10
years and have been summarized by Burton et al. (1976) and Paur et al. (1979).
The agreement between GPT and UV measurements was generally close to 1:1,
although in some cases 0~ measurements by GPT have shown a small positive bias
with respect to UV measurements. In the EPA collaborative study (Rehme et
al., 1981), a number of independent GPT measurement systems were compared to
the reference measurement system and gave CL data that had a mean positive
bias of 7 percent with respect to the UV reference. Demore and Patapoff
(1976) reported a 1:1 agreement between simultaneous measurements of 0« by GPT
and UV with a 5 percent uncertainty on the ratio of these measurements. In a
recent detailed study conducted at the National Bureau of Standards (NBS)
(Fried and Hodgeson, 1982), 0- measurements made with an NBS standard photometer
(Bass et al., 1977) were compared to GPT measurements of £>3 that were standard-
ized against both NO cylinders (NBS SRM) and N02 permeation tubes (NBS SRM).
019QQ/A 5-54 6/19/84
-------
Since the measurement of flow rates is a critical GPT variable and has been
considered as a major source of error in GPT measurements (Demore and Patapoff,
1976), NBS facilities were used for making absolute flow measurements by both
gravimetric and volumetric means. The results of this study were that values
of 0- measured by GPT based on NO- or NO SRMs agreed to within less than 1
percent, but that values of CL measured by UV were lower by a small but signi-
ficant 3 percent. For a consideration of possible error sources, reference is
made to the original article by Fried and Hodgeson (1982). In summary, the UV
photometric CL standard agrees quite closely with the NO and N02 measurement
standards by GPT, as it should in principle. The resolution of any small
biases that remain seems an appropriate matter for consideration by EPA and
NBS.
The measurement principle for the absolute measurement of CL by UV photo-
metry is the same as that used by instruments for monitoring atmospheric 0-, as
described in section 5.5.4.2.3 (Bowman and Horak, 1972; Demore et al., 1976;
Bass et al., 1977). Ozone is measured in a dynamic flow system by measuring
the transmission, I/Io, of ozonized clean air in an absorption cell of path-
length, £. When the concentration is to be expressed in units of ppm, meas-
urement of temperature and pressure is also required. The 03 concentration
may then be calculated directly from the Beer-Lambert equation:
where a = 0_ absorption coefficient at 254 nm, 1 atm, and 0°C,
-1 -1
= 308 ± 4 atm cm (log base e),
and
T = temperature, °K;
P = pressure, torr.
Laboratory photometers used for primary 0- calibrations have pathlengths of 1
to 5 meters and sophisticated digital electronic means for measuring small
absorbancies (Bass et al., 1977; Bowman and Horak, 1972).
A major difference between a photometer for calibration and one for
atmospheric monitoring is that the calibrator uses clean air during the refer-
ence cycle rather than chemically scrubbed ambient air. The conversion of a
019QQ/A 5-55 6/19/84
-------
commercial 03 photometric monitor to a photometer for use as a transfer standard
for calibration has been described by Paur and McElroy (1979). Definitions
are in order here. A primary standard UV photometer is one that meets the
requirements and specifications given in the 1979 revision of the 0, measurement
O
and calibration procedures (U.S. Environmental Protection Agency, 1979). A
transfer standard as used by EPA is a device or a method that can be cali-
brated against a primary photometer and transferred to another location for
calibration of 03 analyzers. Commercial 0. photometers have served well in
this regard, but other devices have been used as well; e.g., calibrated genera-
tors and GPT apparatus. Guidelines on transfer standards for 07 have been
•j
published by EPA (McElroy, 1979), and reference has already been made to the
NAS discussion on transfer standards (National Academy of Sciences, 1977).
Recently a laboratory photometer has been developed by Paur and Bass (1983)
for use in the quality assurance program at EPA on 03 measurements. Some
unique features of this instrument include the mechanism for making the absorb-
ance measurement; internal temperature and pressure transducers; and a mini-
computer for controlling the measurement cycle, computing CL concentrations,
and labeling, storing, and printing calibration data.
The use of UV photometry is unique in air pollution measurements in that
it is based on a physical measurement principle rather than a chemical standard.
It is then worthwhile to trace how the measurement chain works from a primary
standard to field measurements. The primary standard is referenced to the
accepted 0_ absorption coefficient. Transfer standards are then calibrated
with primary photometers maintained at EPA, NBS, and elsewhere. The use of
commercial photometers in this regard has been described by several investiga-
tors (Demore et al., 1976; Hodgeson et al., 1977). These and other kinds of
transfer standards are then used to calibrate 03 analyzers used for field
measurements.
5.5.5.2.3 Other procedures. Although UV photometry has been specified as the
reference calibration procedure, other procedures are available that can give
equivalent results. These include the BAKI method, which was allowed as an
interim alternative method for the calibration of CL monitors when the UV
method was designated in 1979. Other KI methods that have been used success-
fully in Europe are also briefly discussed here. Finally, the GPT method is
reviewed since it has been used extensively in this country and was discussed
above with regard to the cross-check of method accuracies.
019QQ/A 5-56 6/19/84
-------
In the California collaborative study already cited (Demore et a!.,
1976), the 1 percent unbuffered KI (UKI) procedure showed no significant bias
compared to the reference UV procedure. Because it used a titration procedure
for measuring the iodine produced, the technique suffered from large impreci-
sion and its application was not pursued. Thus, the key reason that the
method was not biased was apparently overlooked at the time; that is, the
method did not employ the phosphate buffer. A major problem with NBKI tech-
niques is the slow release of iodine and continued color development after
sampling. Flamm (1977) evaluated the rate of this iodine production and found
that it was the same as the rate with which hydrogen peroxide (HpOp) releases
iodine from the same solution. Based on this observation and a consideration
of other possible species that might be responsible, Flamm concluded that
certain buffer anions, including phosphate, catalyze the formation of H?02 and
yield stoichiometries for iodine production greater than 1. Measurements made
with a 1 percent KI reagent containing 0.1 M boric acid (BAKI), pH=5, did not
exhibit this phenomenon and gave answers that agreed closely with measurements
by UV photometry. These results have been confirmed in other studies (Hodgeson,
1976; Rehme et al., 1981).
The BAKI method was evaluated as one of four alternative techniques in
the collaborative study conducted by EPA (Rehme et al., 1981). No significant
bias was observed between BAKI and the reference technique based on UV photo-
metry. An analysis, however, of BAKI measurements by ten volunteers revealed
a large system-dependent variability, and thus the BAKI technique was not
recommended as an independent calibration method. It is noteworthy that the
system variability attributable to calibration was reduced somewhat if each
operator assumed a molar absorption coefficient for iodine (as I9) of 25,600
-1 -1
M cm rather than independently measuring the absorption coefficient with
standard I_ solutions as these procedures usually prescribe. The BAKI technique
was allowed by EPA as an alternative procedure for the calibration of 0_
monitors, but only for a period of 18 months following the 1979 amendment.
Measurement systems based on the BAKI procedure may still be certified as
transfer standards provided the guidelines for certification given in the EPA
technical assistance document for such standards are followed (McElroy, 1979).
Methods based on iodometry have been used in Europe for some time for the
calibration of 03 analyzers. Bergshoeff (1970) described a method for use in
the Netherlands, in which thiosulfate is added to the KI reagent (KIT method)
along with 0.1 M phosphate buffer. The iodine released is immediately reduced
019QQ/A 5-57 6/19/84
-------
by the thiosulfate and the amount of iodine consumed is determined by back-
titration of the thiosulfate. This method has the advantage that problems
associated with iodine instability in solution are eliminated. In the Federal
Republic of Germany, the standard is based on a 2 percent KI reagent with 2
percent KBr (KIBr Method) and a low concentration (0.02 to 0.03 M) of phosphate
buffer (Van de Wiel et al., 1978). These techniques have been compared to UV
and GPT measurement procedures by Van de Wiel et al. (1978). Measurements
made with the KIBr method were in essential agreement with measurements by UV
or GPT, while measurements by the KIT method were too high by 15 to 25 percent,
depending on the relative humidity of the samples. Modifications have since
been made in the KIT method by the addition of KBr and reduction of the phos-
phate concentration.
The gas-phase titration (GPT) method employs the moderately rapid bimole-
cular reaction between 03 and NO to produce N02 (Rehme et al., 1974):
NO + 03 « N02 + 02 (5-72)
This approach was, in fact, one of the early methods used to measure the
absorption coefficient of 03 (Clyne and Coxon, 1968) and yielded excellent
agreement with other absolute techniques (Demore and Patapoff, 1976). When NO
is present in excess, no side reactions occur and the stoichiometry is as
given above. This method has the distinct advantage that it gives an absolute
relation among three common pollutants. A measurement of the quantity of NO
or 0- consumed or N02 produced provides a simultaneous measurement of the
other two species and the GPT procedure has been used in all three modes.
This calibration technique is often used in the calibration of chemiluminescence
NO (NO + N09) analyzers. In order to obtain accurate concentration measure-
/N £,
ments in the procedure as normally employed, accurate flow measurements are
required; and this is the principal complexity and difficulty with this proce-
dure (Demore and Patapoff, 1976). Because of this problem and unexplained
biases between GPT measurement systems and the UV reference in the EPA collabo-
rative study, the GPT method was not recommended as an independent calibration
technique (Rehme et al., 1981). It is still allowed, however, as a transfer
standard in accordance with the EPA guidelines for these standards (McElroy,
1979).
019QQ/A 5-58 6/19/84
-------
5.5.6 Relationship between Method^ for Total Oxldants and Ozone
When the ambient air quality standards for criteria pollutants were
originally established, a numerical standard was set for photochemical oxidants
as defined by measurements based on iodometric techniques. Much of the health-
related and weIfare-related evidence used as the basis for the standards was
obtained using the total oxidant instrumentation discussed above. The reference
method specified in 1971, however, was the chemiluminescence measurement of
0_. Although not specifically stated at the time, the reasons for specifying
0"3 measurements were undoubtedly the following. First, instrumental methods
for the specific measurement of atmospheric 03 became commercially available
in 1970. These had several very practical advantages over total oxidant KI-
based instruments. These advantages were greater sensitivity, precision,
specificity—no interferences from ambient S0» and N0«—and improved reliabi-
lity in routine monitoring. Second, the data available showed that 03 was the
major contributor to total oxidant measurements, that 0- was the major contri-
butor to observed health and welfare effects, and that 03 could probably
serve, then, as the best surrogate for measurements of total oxidants and for
controlling effects of oxidants in ambient air (see reviews in Burton et al.,
1976; U.S. Environmental Protection Agency, 1978).
Notwithstanding the promulgation of standards for ozone rather than
photochemical oxidants by EPA in 1979, an examination of the temporal and
quantitative relationships between total oxidant and 0~ data remains of con-
«J
siderable interest, largely because pre-existing data and many newer data on
health and welfare effects were obtained by means of total oxidant methods.
Aside from the relative paucity of data on simultaneous measurements, there
are two distinct problems in making such comparisons. The first is the diffi-
culty in estimating the contributions to the total oxidant measurements from
other oxidizing species such as N0? and from reducing species such as S0?.
The presence of such species could cause the total oxidant measurements to be
either higher or lower than 0, concentrations. The second difficulty is in
estimating the bias created between past and present data as a result of the
change from the NBKI to the UV photometry calibration procedures. Fortunately,
these two problems can be treated separately and the latter problem vanishes
for comparison of simultaneous 0 and oxidant data obtained using the same
calibration procedure.
019QQ/A 5-59 6/19/84
-------
In the sections below, the relationship that should exist between total
oxidant and 0_ is considered from an evaluation of the response of NBKI measure-
ments to other oxidizing and reducing species. The predicted relationship is
then compared to data obtained in simultaneous field measurements of total
oxidant and 0».
5.5.6.1 Predicted Relationship. The predicted total oxidant measurements
can be expressed as the sum of the contributions from oxidizing and reducing
species that release or consume iodine in NBKI reagent:
[Total Ox] = a[03] + Z. b. [Ox]i + c[N02] (5-73)
d[S02] - I. ei [Red].
In this equation, [Ox], and [Red], represent the concentrations of other
oxidizing and reducing species in the atmosphere. The atmospheric concentra-
tions of other reducing species, such as H?S, are normally quite low compared
to S0~ concentrations (Stevens et al. , 1972b, and references therein) and
these species will not be considered further here. If the concentrations
above are true atmospheric concentrations, the constants a, b, c, and d repre-
sent the efficiencies with which the various species release or consume iodine.
For example, the value of the constant, a, for an oxidant instrument calibrated
by the CARB 2 percent NBKI method would be approximately 1.2 (section 5.5.5.2).
Since the instruments are calibrated with ozonized air, the factor, a, repre-
sents the bias of the calibration method used. If the 0^ concentration is
overestimated because of calibration bias, then so are the contributions of
the other species by the same factor; i.e., the constants b, c, and d are all
higher than their true values by the same constant, a. Therefore, it should
in principle be possible to correct total oxidant data for calibration bias by
dividing both sides of the equation above by a.
[Total Ox]1 = [Total Ox]/a (5-74)
= [03] +1. b'. [Ox]. c'[N02] + d'[S02]
As discussed previously (section 5.5.5.2), whenever such a correction is
attempted, the net uncertainty in the total oxidant data will have to be
increased by an amount equivalent to the uncertainty in the calibration bias
factor. Next, literature values reported for b^, c, and d will be discussed.
019QQ/A 5-60 6/19/84
-------
The magnitude of these factors, together with estimated atmospheric concentra-
tions, will be used to compare predicted relationships between total oxidants
and 0,.
Other atmospheric oxidants that have been identified and that may contri-
bute to the total oxidant reading are hydrogen peroxide (HLCL), small organic
peroxides (e.g., methyl and ethyl hydroperoxide), peracetic acid, peroxyacyl
nitrates (Cohen et al., 1967b), and pernitric acid (Niki et a!., 1976). An
estimation of the contribution of these species to the total oxidant measure-
ment is quite difficult because individual b.'s have not been measured and
there are few data available on atmospheric concentrations of individual
species. The magnitude of the efficiency term will depend not only on the
stoichiometry of the oxidation reaction, but also on the rate. For many of
the oxidants above, the overall stoichiometry may be equivalent to that of (L,
given sufficient time for completion of the reaction. Cohen et al. (1967b)
reported nearly equal molar absorption coefficients for iodine production by
DO, HpCL, peracetic acid, acetyl peroxide, and ethyl hydroperoxide. Only 0,
and peracetic acid gave immediate color development and the others were classi-
fied as slow oxidants because the color developed slowly. A summary of the
effects of various oxidants on NBKI reagent and the Mast oxidant meter is
given in Table 5-10 (Cohen et al., 1967b; Purcell and Cohen, 1967; Burton et
al., 1976).
In contrast to the b. terms, reaction efficiencies for NO,, and SOp are
relatively well known. The most definitive study of the effect of NO,, is by
Tokiwa et al. (1972). In that study the reaction efficiencies were 6 percent
for the Mast oxidant meter, 22 percent for a 10 percent KI colon'metric anal-
yzer, and 32 percent for a 20 percent KI colorimetric analyzer. At the
20 percent KI concentration, the reaction efficiency actually varied from
33 percent at zero 0- concentration to 19 percent at an 0- concentration of
0.6 ppm. Some of the earlier studies reported a negative interference from S0
but gave variable values for the magnitude of the effect (Cholak et al., 1956;
Deutsch, 1968). It is now well-documented that SO,, is a quantitative negative
interference with a 100 percent efficiency for reducing the oxidant reading by
an amount equivalent to the S02 concentration present (Cherniack and Bryan,
1965; Saltzman and Wartburg, 1965).
Returning to the analytical expression for total oxidant, an "adjusted"
or corrected oxidant value can be expressed as below, assuming that the same
019QQ/A 5-61 6/19/84
-------
TABLE 5-10. RESPONSE OF NBKI REAGENT AND MAST METER TO VARIOUS OXIDANTS3
Ozone
Peracetic acid
Hydrogen peroxide
Acetyl peroxide
Ethyl hydroperoxide
n-Butyl hydroperoxide
tert-Butyl hydroperoxide
Nitrogen oxides (NO )
J\
Peroxyacetyl nitrate (PAN)
Peroxypropionyl nitrate (PPN)
NBKI
Aa
A
B
B
B
B
B
D (10% as N02)
0
D
Mast
E
-
D
N
-
-
-
D (10%)
N
D
A = immediate color development; B = slow color development; D = positive
interference; E = good response; and N = no response (or negligible).
Source: Cohen et al. (1967).
calibration procedure is used for total oxidant and 0,, and assuming that no
other significant reducing interferences are present:
[Total Ox]CQrr = [Total Ox] - c[N02] + [SOg]
(5-75)
= [03]
[Ox].
Thus, a total oxidant measurement for which legitimate corrections or compen-
sations for NOp and S0« have been made should always be higher than a simul-
taneous 03 measurement by an amount that is a function of the type and con-
centrations of other oxidants present. The only major qualifications to this
prediction are that both types of measurements must be sampling the same air
mass and be calibrated with respect to the same reference; that no other
significant reducing interferences are present; and that 03 losses within the
sample inlet system are insignificant. On the other hand, total oxidant data
019QQ/A
5-62
6/19/84
-------
uncorrected for S02 and NOp interferences may be higher or lower than corre-
sponding 0- data, depending on the concentrations of these pollutants. Because
of the potential presence of these interferences, it is quite difficult or
impractical to compare oxidant and 03 measurements during evening and early
morning hours, when 0~ concentrations are quite low. As shown in chapter 2
and discussed in chapter 4, however, when the (L concentration has reached its
diurnal maximum value during late morning or early afternoon hours, the NCk
concentration is near a minimum and is low compared to the CL concentration
(Leighton, 1961; Stevens et al., 1972a). Likewise, at the 0^ maximum, the SCk
concentration also is often, but not always, small compared to the 0, concen-
tration, particularly in studies done in California (Dickinson, 1961; Ballard
et al., 1971a; Stevens et al., 1972b). Therefore, in the comparison of total
oxidant and 0, simultaneous field measurements below, emphasis is placed on
comparison of peak hourly averages. This seems especially appropriate since
the oxidant standard and compliance monitoring are based on the second-highest
hourly average (U.S. Environmental Protection Agency, 1979).
5.5.6.2 Empirical Relationship Determined from Simultaneous Measurements.
Because of the difficulties and uncertainties in predicting the relationship
between 03 and oxidant measurements, this comparison is best determined by
simultaneous measurements of the atmosphere. Nevertheless, the predicted
relation discussed above is useful in evaluating results from field measure-
ments. Several precautions should be taken in performing simultaneous measure-
ments. Both kinds of instruments must be calibrated frequently with the same
ozonized air stream that has been analyzed by a common reference method. In a
simultaneous comparison, daily calibrations should be made with an (L generator
and the generator output should be analyzed weekly. Both instruments should
sample the same air parcel. Routine maintenance should be frequent to ensure
constant gas and reagent flow rates, clean sample inlet systems, etc. Finally,
in any meaningful comparison of (k and oxidant data, simultaneous measurements
of NCL and S0? should be made. If chromium trioxide scrubbers are used to
remove S02 in the inlet to the oxidant instrument, these must be frequently
tested to ensure that (k is not also removed during continued use, particularly
under very humid conditions. These scrubbers may cause some additional bias
by oxidation of NO to NO,,. During stagnation periods in cities, NO can build
up to high concentrations in the morning and oxidation of NO by the scrubber
will give a significant response in colorimetric analyzers. Several studies
on the comparison of total oxidant and 0, measurements have been made and are
019QQ/A 5-63 6/19/84
-------
summarized below. In some cases the precautions noted above were taken and in
some cases they were not.
The earliest comparative study reported was by Renzetti and Romanovsky
(1956). This study compared a phenolphthalein total oxidant monitor, a KI
continuous oxidant monitor, a rubber-cracking apparatus, and an open-path
ultraviolet spectrometer, which monitored the UV absorption at characteristic
0- absorption wavelengths. The only meaningful measurements for consideration
here are the KI oxidant and UV 0., measurements, since these are similar to
measurement methods used later. The KI oxidant monitor was the version (Littman
and Benoliel, 1953) used for later colorimetric oxidants measurements, but
used with a 20 percent KI rather than 10 percent KI reagent. The UV 03 spec-
trometer, however, differed in a number of respects from the 0- photometers in
use today. Measurements were made across an open optical path of 325 ft of
the transmission at three wavelengths, A... = 265 nm, K^ ~ 313 nm, and \^ - 280
nm. Intensity ratios at the three wavelengths were used to minimize the
effects of other UV absorbers and of particulate scattering. Some non-O^
absorption may still have been present, and, if so, the measured values would
be higher than the true 0, values. The published absorption coefficients at
these three wavelengths were used to compute 03 concentrations (Vigroux,
1952). This should not be a serious source of inaccuracy since laboratory
measurements of these coefficients have not changed significantly (Demore et
al., 1976). Measurements were made over a 4-month period. Figures 5-3 and
5-4 are illustrative of the data obtained for a single day or a monthly average.
Peaks of total oxidant and of 03 occurred at the same time, but the 03 maximum
was usually less than the total oxidant maximum. The UV 03 data were usually
higher in the wings at low 03 concentrations. If this effect was the result
of particulate scattering or absorption by other UV absorbers, which is pos-
sible, these same interferences were likely to have been present at the peak
0- concentration also, and the concentration was then overestimated. Renzetti
•j
and Romanovsky attributed the higher total axidant reading to the presence of
"other oxidants" and estimated concentrations of other oxidants of 0.1 to 0.4
ppm, depending on 03 concentration. Since the total oxidant instrument was
calibrated by an NBKI method, this estimate is almost certainly too high and a
large portion of the difference between oxidant and 03 may have been a result
of the 20 to 25 percent positive calibration bias. Since interferences may be
present in the UV measurement and simultaneous measurements of NOg and SO^
019QQ/A 5-64 6/19/84
-------
I
a
a.
60
50
40
I 30
ui
O 20
O
U
10
I
a
a
QC
Z
LJJ
O
Z
O
U
T I I I I I I I I I I I I I
OXIDANTS BY Kl
O3 BY UV
I I I I I I I I
I I I I I I I
III!
I I I I I I I I 1
12 1 234 5678 9 10 11 12 1 234 5 6 7 8 9 10 11 12
A.M. AUGUST, 1955 P.M. P.S.T.
Figure 5-3. Ozone and oxidant concentration in the Pasadena area, August 1955.
Source: Renzetti and Romanovsky (1956).
60
50
40
30
20
10
I I
1 I I I I II I T
— OXIDANTS BY Kl
— O,BY UV
12 1 2 34 5 6 7 8 9 10 11 12 1 2345678 9 10 11 12
A.M. AUGUST 27,1955 P.M. P.S.T.
Figure 5-4. Ozone and oxidant concentration in the Los Angeles area.
Source: Renzetti and Romanovsky (1956).
5-65
-------
were not available at the time, no attempt is made to make any more quanti-
tative assessment of this study.
A later study (Cherniack and Bryan, 1965) compared a 10 percent colorime-
tric KI oxidant instrument, a Mast oxidant meter (Brewer and Mil ford, 1960), a
galvanic-cell oxidant instrument (Hersch and Deuringer, 1963), and a UV 0,
«J
photometer (Bryan and Romanovsky, 1956). This latter instrument was similar
in principle to present-day photometers. The precautions noted above were
taken. All the instruments were calibrated with respect to the 2 percent UKI
calibration procedure used by the LAAPCD. The results obtained for the cali-
bration of the UV photometer are interesting. The absolute concentrations for
the UV photometer ([033UV), based on the 03 absorption coefficient, were
related to colorimetric oxidant meter readings ([Oolnvrn), calibrated by 2
percent UKI, by the following linear equation:
[03]uv = 1.027 [03]QXID + 0.005 (5-76)
The corresponding slope in the later study by Demore et al. (1976), comparing
the LAAPCD calibration method with UV photometry, was 1.04. Simultaneous SO,,
and N0? measurements were made, but no corrections were made because the con-
centrations were reported to be quite small during the period of comparison.
Atmospheric sampling was conducted over an unspecified period of time, and the
data were expressed in terms of a linear regression of data from each instru-
ment versus the colorimetric oxidant analyzer as the reference. The linear
regression analysis of the data over the concentration range 0 to 0.6 ppm gave
the relationships shown in Table 5-11 after correction for calibration factors.
Thus, the data show a much better absolute agreement and correlation between
0^ measurements and colorimetric total oxidant than between electrochemical
total oxidant and colorimetric total oxidant. Other studies have also shown a
similar comparison between the electrochemical and colorimetric measurements
(Tokiwa, 1972; Stevens et al., 1972a; Stevens et al., 1972b). In addition,
these data do not indicate any significant contribution by "other oxidants" to
the total oxidant measurement. The only qualifications to these observations
are that corrections for N0? and S0? were not made, although they were reported
as low. The data in Table 5-11 represent the total concentration range; it
would have been informative to examine individual relationships at the oxidant
maximum.
019QQ/A 5-66 6/19/84
-------
TABLE 5-11. COMPARISON OF CORRECTED INSTRUMENT READINGS
TO COLORIMETRIC OXIDANT READINGS DURING ATMOSPHERIC SAMPLING3
Instrument
Mast meter
Galvanic cell
Ozone photometer
m
0.896
0.776
0.980
b
-0.013
+0.004
-0.005
r
0.868
0.867
0.982
(y = Instrument reading
= mx + b; x = colon"metric oxidant measurement; m = slope; b = intercept;
r = correlation coefficient)
Source: Cherniack and Bryan (1965).
During the 1970s, several studies were conducted on the intercomparison
of 0- and total oxidant instrumentation. The Research Triangle Institute
(RTI) of North Carolina, under contract to EPA, conducted extensive field
studies on 0- and total oxidant measurements in both Los Angeles and St. Louis
(Ballard et al., 1971a; Ballard et a!., 1971b; Stevens et al., 1972a; Stevens
et al., 1972b). Measurements were made for 0- by chemi luminescence and for
total oxidant by a colorimetric KI analyzer and a Mast meter. Calibrations
were carried out frequently with an 0~ generator calibrated by the 1 percent
NBKI method. Data processors were used to collect and store all monitoring
and calibration data. Simultaneous N0? and S0? measurements were also made
and the oxidant data reported were corrected for these interferences. In
another pertinent study (Clark et al., 1974), several instruments for the
specific measurement of atmospheric 0- were intercompared by monitoring in a
rural environment. These instruments were a commercial UV photometer, three
different commercial gas-phase chemiluminescence analyzers, and a gas-solid
chemiluminescence analyzer. The instruments were all calibrated by a common
reference procedure and hourly-averaged field measurements were collected over
a 1-month period in August 1972 (Clark et al., 1974). Davis and Jensen (1976)
reported intercomparisons of Mast meter total oxidant measurements and chemilu-
minescence 0- measurements. The instruments were not calibrated by the same
procedure in that study, however, nor were any corrections attempted for SO-
and N02 interferences. Okita and Inugami (1971) reported an intercomparison
of KI total oxidant measurements with chemiluminescence 0_ measurements in the
019QQ/A 5-67 6/19/84
-------
urban atmosphere of Musashino, Japan. An intercomparison of total oxidant by
KI and CL by chemiluminescence in irradiated auto exhaust was reported by
J
Carroll et al. (1971). In another extensive field study conducted at an air
monitoring station near the Houston ship channel, Severs and coworkers (Severs,
1975; Neal et al., 1976) examined the relationship between ozone and total
oxidants for this area by making simultaneous measurements with a gas-phase
chemiluminescence CL monitor and a Beckman colorimetric total oxidant analyzer.
O
Primary calibrations of the instruments were performed periodically using the
EPA 1 percent NBKI method. Hourly-averaged measurements were accumulated over
a period from May 1972 through December 1973. No corrections were attempted
for N0? or S0? interferences, but during the latter part of this study (August-
December 1973) a chromium trioxide scrubber was placed in the inlet of the
total oxidant analyzer.
All of these 1970 studies were reviewed in the previous criteria document
(U.S. Environmental Protection Agency, 1978). Only the major conclusions are
repeated here. In general, the averaged data showed fairly good qualitative
and quantitative agreement between the diurnal variations of total oxidants
and 0-. The usual trend was a slightly higher value for the total oxidants
*J
measurement at the maximum, a not unexpected result in view of the discussion
above. Comparisons of monthly-averaged data taken from studies in Los Angeles
and St. Louis are shown in Figures 5-5 and 5-6 (Stevens et al., 1972a; 1972b).
The total oxidant data shown in Figure 5-6 are uncorrected and show distinct
morning and evening peaks resulting from N02 interference (see chapter 2 for
diurnal patterns of N02). Examination of data taken from individual days
shows considerably more variation among the methods, with total oxidant measure-
ments both higher and lower than 03 measurements. Intercomparisons of only UV
photometric and chemiluminescence 0» analyzers have not shown these large
variations (Clark et al., 1974; Wendt, 1975). In all probability, these
variations result from the large imprecision and interferences in total oxidant
measurement.
Two of the studies above reported consistently lower total oxidant measure-
ments. In one of these (Davis and Jensen, 1976), the reference KI method was
used for calibrating the chemiluminescence analyzer while a factory calibra-
tion was used for the Mast meter. As pointed out above, other studies have
found low oxidant readings for the Mast meter as compared to colorimetric
analyzers (Cherniack and Bryan, 1965; Tokiwa et al. , 1972; Stevens et al.,
019QQ/A 5-68 6/19/84
-------
OZONE-CHEM
TOTAL OX-MAST
Ofc
0400
0800 1200
TIME OF DAY
1600
2000
2400
Figure 5-5. Measurements for ozone and oxidants in Los Angeles.
Source: Stevens et al., 1972b.
5-6°
-------
0.12
0.10
a
a
O 0.08
cc
z
O
u
0.06
0.04
0.02
-I I I I I I I I I I I I I I I I I I I I I I I
COLORIMETRIC
COULOMETRIC
CHEMILUMINESCENCE
hTTT i +
i i M"TT>T
02 4 6 8 10 12 14 16 18 20 22 24
TIME, hours
Figure 5-6. Measurements for ozone and oxidants in St. Louis.
Source: Stevens et al., 1972a.
5-70
-------
1972a, 1972b). The use of the factory calibration would cause the Mast readings
to be even lower because of a calibration bias (Cherniack and Bryan, 1965;
Tokiwa et al., 1972). The results reported by Severs and coworkers are more
difficult to evaluate. Chemiluminescence 0, values generally higher, and
sometimes considerably higher, than total oxidant measurements were reported,
although the measurements were referenced to the same calibration procedure.
Correlations were reported both with and without a chromium trioxide scrubber
in the oxidant inlet. It is interesting that slightly better agreement was
obtained without the chromium trioxide scrubber; but a large bias remained.
These results are inconsistent with the known responses of the instruments and
the results of other investigators, and it is probable that a serious measure-
ment problem or interference existed in that study. The data reported for one
day of high (L» but abnormally low oxidants, are shown in Figure 5-7. It is
highly improbable that the problem is with the chemiluminescence 0- measurement,
since this is typical of a normal CL diurnal variation and no other species
are known to interfere. It is far more probable that some other species of
pollutant in the highly industrialized area of the Houston ship channel re-
pressed the response of the total oxidant analyzer, which thus does not respond
to 0~, much less to any other oxidant.
The most recent comparison in the literature involved simultaneous 0- and
o
total oxidant measurements in the Los Angeles basin by the California Air
Resources Board (1978) in the years 1974, 1976, and 1978. The maximum hourly
data pairs were correlated (Chock et al., 1982) and yielded the following
regression equation for 1978 data, in which a large number (927) of data pairs
were available:
Oxidant (ppm) = 0.870 03 + 0.005 (5-77)
(correlation coefficient = 0.92)
Thus, when the 0_ levels were relatively high, they were actually slightly
higher than total oxidant. The total oxidant data were uncorrected for hKK
and S02 interferences.
In summary, specific 0, measurements agree fairly well with total oxidant
-------
0.150
i
z'
2 0.100
DC
HI
O
Z
o
o
O3 BY CHEMILUMINESCENCE
OXIDANTS BY BECKMAN ACRALYZAR (Kl)
0.050
cc
o
10
15
20
TIME, hours
Figure 5-7. Measurement of ozone and oxidants, Houston Ship Channel,
August 11, 1973.
Source: Severs (1975).
5-7?
-------
oxidizing and reducing interferences with KI measurements. As a result of
these interferences, on any given day the total oxidant data may be higher
than or lower than simultaneous CL data. The quantitative relationship between
oxidant and 03 data, such as that used by Chock et al. (1982), is probably
quite location-dependent. From a methodologic standpoint, the measurement of
03 is a more reliable indicator than total oxidant measurements of oxidant air
quality; and such difficulties and controversy as may be involved in the
intercomparison of 03 and oxidant measurements are eliminated if the air
quality standard is defined in terms of 0^.
5.5.7 Methods for Sampling and Analysis of Peroxyacetyl Nitrate and Its Homologues
5.5.7.1 Introduction. Since the discovery of the mysterious compound X
(Stephens et al., 1956), later unambiguously identified as peroxyacetyl nitrate
(PAN) (Stephens et al., 1961), much effort has been directed toward its
atmospheric measurement. Peroxyacetyl nitrate is a product of photochemical
reactions involving hydrocarbons and oxides of nitrogen (NO ) in the atmosphere
(Stephens, 1969). The significance of atmospheric PAN is twofold. It is a
potent lachrymator and phytotoxicant in the ppb concentration range (Huess and
Glasson, 1968). Because of the reversible thermal equilibrium (Hendry and Kenley,
1977),
< CH3CO(02)- + N02 (5-78)
which is sensitive to the N02/N0 ratio, PAN may serve as an important reservoir
for peroxy radicals and N02 (Singh and Sal as, 1983a, 1983b) and may play a
significant role in both the atmospheric nitrogen cycle and in tropospheric
ozone formation (Spicer et al., 1983).
Only two analytical techniques have been used to obtain significant data
on ambient PAN concentrations. These are gas chromatography with electron
capture detection (GC-ECD) and long-path Fourier-transform infrared (FTIR)
spectrometry. Atmospheric data on PAN concentrations have been obtained
predominantly by GC-ECD because of its relative simplicity and superior sensi-
tivity. These analytical techniques are described in section 5.5.7.2 along
with attendant methods of sampling. Peroxyacetyl nitrate is somewhat analogous
to 03 in that it is a thermodynamically unstable oxidant and PAN standards
must be generated and analyzed by some absolute technique for the purpose of
019QQ/A 5-73 6/19/84
-------
calibrating the GC-ECD. Generation and calibration techniques are discussed
separately in section 5.5.7.3. Finally, the analysis of PAN homologues is
discussed briefly in section 5.5.7.4.
5.5.7.2 Analytical Methods. By far the most widely used technique for the
quantitative determination of ppb concentrations of PAN is GC-ECD (Stephens,
1969). With Carbowax or SE30 as a stationary phase, PAN can be separated from
components such as air, water, and other atmospheric compounds, as well as
ethyl nitrate, methyl nitrate, and other contaminants that are present in PAN
synthetic mixtures. Electron-capture detection (using a Nickel-63 source and
a pulsed-current detector or a tritium source and a direct-current detector)
provides sensitivity for PAN in the ppb and sub-ppb ranges. A typical column
for the separation of PAN would be 3 to 5 feet in length and 1/8-inch in diameter
(i.d.), and would be run isothermally at 25° to 60°C. Under these conditions,
a peak assignable to PAN appears after 2 to 3 minutes. Table 5-12 shows
parameters used by several investigators to determine trace levels of PAN by
GC-ECD.
Sample injection into the GC is accomplished by means of a gas-sampling
valve with a gas-sampling loop of a few milliliters volume (Stephens and
Price, 1973). Sample injection may be performed manually or automatically.
Typically, manual air samples are collected in 50 to 200 ml ungreased glass
syringes, and purged through the gas-sampling valve. Samples collected from
the atmosphere should be analyzed as soon as possible because PAN undergoes
thermal decomposition in the gas phase and at the surface of containers.
Automatic sample collection and injection may be accomplished by using a small
pump to pull ambient air continuously through the sampling loop of an automatic
sampling valve, which periodically injects the sample onto the column (Stephens
and Price, 1973). Recently, Singh and Salas (1983b) have used cryogenic trap-
ping of PAN, with liquid argon, from relatively large air samples for the
purpose of measuring PAN concentrations in the sub-ppb range.
Most of the atmospheric PAN measurements have been made in polluted urban
environments, where maximum concentrations of 5 to 50 ppb may occur, with
average concentrations of a few ppb (Stephens, 1969; Lonneman et al., 1976).
For the purpose of such measurements, chromatographic detection limits of 0.1
to 1 ppb are sufficient. The recent work of Singh and Salas (1983a, 1983b) on
the measurement of PAN in the free (unpolluted) troposphere is illustrative of
current capabilities for measuring low concentrations. A 50 to 200 ml volume
019QQ/A 5-74 6/19/84
-------
TABLE 5-12. SUMMARY OF PARAMETERS USED IN DETERMINATION OF PAN BY GC-ECD
Reference
Heuss and Glasson,
1968
Grosjean, 1983
Parley et al. ,
1963
Stephens and Price,
1973
en
i
01
Lonneman et al. ,
1976
Holdren and Spicer,
1984
Peake and Sandhu,
1983
Singh and Sal as,
1983
Grosjean et al . ,
1984
Nielson et al . ,
1982
Column dimensions
and materials
4 ft x 1/8 in
Glass
6 ft x 1/8 in
Teflon
3 ft x 1/9 in
Glass
1.5 ft x 1/8 in
Teflon
3 to 4 ft x 1/8 in
Glass
5 ft x 1/8 in
Teflon
3.3 ft x 1/8 in
Glass
1.2 ft x 1/8 in
Teflon
1.7 x 1/8 in
Teflon
3.9 ft x 1/12 in
Glass
Stationary phase
SE30
(3.8%)
10% Carbowax 400
5% Carbowax 400
5% Carbowax E 400
10% Carbowax 600
60/80 Mesh
Carbowax 600
5% Carbowax 600
10% Carbowax 600
10% Carbowax 400
5% Carbowax 400
Solid support
80-100
Mesh
Diatoporte S
60/80 Mesh
Chromosorb P
100-200 Mesh
Chromosorb W
Chromosorb
G 80/100
Mesh treated
with dimethyl
dichlorosilane
Gas Chrom Z
Gas Chrom Z
Chromosorb W
80/100 Mesh
Supelcoport
60/80 Mesh
Chromosorb G
Chromosorb
W - AW - DCMS
Column
temperature,
°C
25
30
35
25
25
35
33
33
30
25
Flow rate, Carrier
(ml/mi n) gas
N.A. N.A.
40 N2
25 N2
60 N2
70 95% Ar
5% CH4
70 90% Ar
10% CH4
50 N2
30 95% Ar
5% CH4
40 N2
40 N2
Elution
time,
min
N.A.
N.A.
2.17
1.75
2.7
3.00
N.A.
H.A.
5.0
6.0
Concentration
range
ppb range
ppb range
3 to 5 ppb
37 ppb
0.1 to 100 ppb
ppb range
0.2 to 20 ppb
0.02 to 0.10 ppb
2 to 400 ppb
11 ppb
-------
of air was collected by preconcentration into an unpacked 0.15 cm o.d. stainless
tube of 1.24 ml volume, at liquid argon temperature, prior to analysis (see
Table 5-12). For measurements in humid environments, the air sample was
passed through a Nafion drier (Foulger and Simmonds, 1979) to reduce the
humidity prior to preconcentration. A minimum detection limit of 0.010 ppb
was obtained. Free tropospheric concentrations in the 0.010 to 0.100 ppb
range were always observed and indicated that PAN is a natural constituent of
the atmosphere and may constitute a significant fraction of the reactive
nitrogen.
There are conflicting reports in the literature on the effects of variable
relative humidity on PAN measurements by GC-ECD. In 1973, Stephens and Price
stated that in preparing PAN calibration samples "the diluent (gas) should be
of normal humidity so that the chromatogram will be a realistic one." The
reason(s) for this precaution were not given. Subsequently, Holdren and
Rasmussen (1976) observed a reduced response to PAN calibration samples when
the relative humidity of the sample was 30 percent or lower and a tenfold
decrease in PAN response when the relative humidity approached 0 percent.
This effect was attributed to an interaction between sample and GC column.
Nieboer and van Ham (1975) reported that "the elution gas stream was previously
humidified . . . because it appeared that the height of the PAN peak depends
on the relative humidity of ambient air if dry elution gas was used." In
contrast to these studies, Lonneman (1977) observed no effect on peak height
in PAN calibration samples in which the relative humidity varied from 10 to 50
percent.
In 1978, Watanabe and Stephens reported on a reexamination of the moisture
anomaly and investigated the effects of humidity on PAN storage flasks, columns,
and detectors. A consistent PAN loss to the walls of dry acid-washed glass
storage flasks was observed. This PAN could be recovered by the addition of
moisture to the flasks. A tritium direct-current detector showed no humidity
effect except for a small (5 to 10 percent) decrease in peak height in a few
cases at very low humidities (2 to 3 percent). With a different GC instrument
employing a Ni-63 detector, erratic responses were observed at low humidities,
with responses reduced 30 to 95 percent from that obtained at 53 percent
relative humidity. No conclusion was drawn on whether this difference reflec-
ted a humidity effect on the detector, the column, or the sampling value.
Finally, the moisture anomaly did not appear to depend on column history or
loading even after a bake-out treatment.
019QQ/A 5-76 6/19/84
-------
In the most recent study (Grosjean et a!., 1984), the humidity effect on
PAN calibration samples prepared by the dynamic and static irradiation of
ChLCHO-Clp-NO- mixtures was examined. A small decrease (<3 percent) was
observed in PAN peak height when the dry air stream was passed either over an
impinger containing water or directed through the water impinger. These
results are in contradiction to all of the above, in which the response to
humidified PAN samples is either greater than or the same as dry PAN samples.
It is noteworthy that the chromatographic systems employed by different
investigators often employ different materials of construction (e.g., glass,
Teflon, stainless steel), column loadings, and detectors. The resolution of
all the differences noted above in regard to a suspected humidity effect might
require considerable effort. For the present, if a moisture effect is suspec-
ted in a PAN analysis, the bulk of this evidence suggests that humidification
of PAN calibration samples (to a range approximating the humidity of the
samples being analyzed) would be advisable.
Conventional long-path infrared spectroscopy and Fourier-transform infrared
spectroscopy (FTIR) have been used to detect and measure atmospheric PAN.
Sensitivity is enhanced by the use of FTIR over conventional long-path infrared
spectroscopy. Accurate knowledge of the absorptivities of many IR bands
assignable to PAN makes possible the quantitative analysis of PAN without the
use of calibration standards. The most frequently used IR bands have been
assigned and the absorptivities shown in Table 5-13 have been reported. Only
the key bands are shown, but all 27 fundamentals are infrared active and
Bruckmann and Wiliner (1983) have assigned most of them. The assignment by
Adamson and Guenthard (1980) of the bands at 1435, 1300, and 990 cm"1 to an
impurity, CH-ONO-, is apparently incorrect. Bruckmann and Wi liner (1983) ob-
served these same bands in a 99 percent pure PAN sample.
The initial discovery of PAN in simulated photochemical smog was accom-
plished by long-path infrared absorption spectrometry (Stephens et al., 1956).
Some recent simultaneous measurements of PAN and other atmospheric pollutants
such as 0_, HN03, HCOOH, and HCHO have been made by long-path FTIR spectrometry
during smog episodes in the Los Angeles Basin. Tuazon et al. (1978) have
described an FTIR system operable at pathlengths up to 2 km for ambient measure-
ments of PAN and their trace constituents. This system employed an eight-mirror
multiple reflection cell with a 22.5-m base path. The spectral windows available
at pathlengths of 1 km were 760-1300, 2000-2230, and 2390-3000 cm"1. Thus, PAN
019QQ/A 5-77 6/19/84
-------
PRELIMINARY DRAFT
TABLE 5-13, PAN INFRARED ABSORPTIVITIES
en
i
co
Absortivity, ppm-1 m-1 x 104
Frequency,
cm-1
1842
1741
1302
1162.5
930
791.5
Mode
v(c=o)
v (N02)
as 2
vs(N02)
v(c-o)
v(o-o)
6(N02)
Liquid PAN
(Bruckmann and
Winner, 1983)
12.4
32.6
13.6
15.8
N.A.a
13.4
PAN in air
(Stephens, 1969)
10.0
23.6
11.2
14.3
N.A.a
10.1
Frequency,
cm-1
1830
1728
1294
1153
930
787
PAN in air
(Stephens and
Price, 1973)
10.0
23.6
11.4
13.9
1.8
10.3
PAN in octane
(Holdren and
Spicer, 1984)
9.44
26.5
9.44
9.66
N.A.3
10.1
Not available.
-------
-1 -1
could be detected by the bands at 793 and 1162 cm . The 793 cm band is
— T
characteristic of peroxynitrates, while the 1162 cm band is reportedly caused
by PAN only (Hanst et a!., 1982). Tuazon et al. (1981) reported on ambient
measurements with his systems taken during a smog episode in Claremont, CA, in
1978. Maximum PAN concentrations ranged from 6 to 37 ppb over a 5-day episode;
the report presented diurnal patterns for PAN and several other pollutants for
the 2 most severe days. The detection limit given for PAN at a 900-m pathlength
was 3 ppb.
Hanst et al. (1982) modified the FTIR system used by Tuazon et al. (1978)
by changing it from an eight-mirror to a three-mirror cell configuration and
by considerably reducing the cell volume. Measurements were made over a
1260-rn optical path folded along a 23-m base path at 0.25 cm resolution.
Measurements were reported for PAN and a variety of other pollutants for a
2~day smog episode at California State University, Los Angeles, in 1980. The
maximum PAN concentration observed was 15 ppb for this period of only moderate
smog intensity. An upper limit of 1 ppb of peroxybenzoyl nitrate (PbzN) was
placed based on observations in the vicinity of the PBzN band at 990 cm .
The reports by Tuazon et al. (1978) and Hanst et al. (1982) both refer to
earlier FTIR ambient air studies.
Sampling may constitute one of the major problems in the analysis of
trace reactive species, such as PAN, by long-path FTIR spectrometry. The
folded-path White cells have a significant internal volume (15 m for Tuazon
et al., 1979; 3 m for Hanst et al., 1982). The large internal surface area
may serve to promote the decomposition or irreversible adsorption of reactive,
trace species. To minimize these effects, both Tuazon et al. and Hanst et al.
employed high-speed blowers to pull ambient air through the cells at high
velocities. For interior cell linings, Hanst et al. employed 0.5 mm polyvinyl
chloride sheeting and Tuazon et al. used Plexiglas and FEP Teflon.
Pitts et al. (1973) proposed a chemiluminescence technique for continuous
monitoring of ambient concentrations of PAN. The reactions of both PAN and 03
with triethylamine in the gas phase produce chemiluminescence. The spectra
reported overlap somewhat with a \max of 520 nm for the 03 reaction and \max
of 650 nm for the PAN reaction. Pitts proposed a technique whereby, through
measurement of the emission intensity in the two regions by the use of optical
cut-off filters, the PAN concentration could be determined, provided simultane-
ous measurements were made of the absolute 03 concentration. Concentrations
019QQ/A 5-79 6/19/84
-------
of 6 ppb PAN were detected and a lower limit of detection of 1 ppb was estimated.
No interfering emissions were observed from methyl nitrate, ethyl nitrate, ethyl
nitrite, or NOp. No further work has been reported on the development of this
technique, and there have not been any atmospheric applications.
5.5.7.3 Generation and Calibration. Because of the thermal instability of
dilute PAN samples and the explosive nature of purified PAN, calibration samples
are not available, and each laboratory involved in making such measurements must
prepare its own standards. The PAN samples are prepared by various means at
concentrations in the ppm range and these must be analyzed by some absolute
technique. The analyzed samples must then be diluted to obtain gas-phase sam-
ples in the low ppb range for direct calibration of GC instruments. Thus, the
following section includes descriptions of various means of PAN generation,
methods of analysis, and the procedures for sample handling and storage where
applicable.
The earlier methods used for the preparation of PAN have been summarized
by Stephens (1969). These included (1) the photolysis of mixtures of nitrogen
oxides with organic compounds in air or oxygen (Stephens, 1956; Stephens
et al., 1961); (2) the photolysis of alkyl nitrite vapor in oxygen (Stephens
et al., 1965); (3) the dark reaction of aldehyde vapor with nitrogen pentoxide
(Tuesday, 1961); and (4) the nitration of peracetic acid. Of these methods,
the photolysis of alkyl nitrites was favored and used extensively by Stephens
and other investigators. As described by Stephens et al. (1965), the liquefied
crude mixture obtained at the outlet of the photolysis chamber is purified by
preparation-scale GC. [CAUTION: Both the liquid crude mixture and the purified
PAN samples are violently explosive and should be handled behind explosion
shields using plastic full-face protection, gloves, and a leather coat at all
times. These PAN samples should be handled in the frozen state whenever possi-
ble.] The pure PAN is usually diluted to about 1000 ppm in nitrogen cylinders
at 100 psig. When refrigerated at <15°C, PAN losses are less than 5 percent per
month (Stephens et al., 1965). Lonneman et al. (1976) used the photolysis pro-
ducts without purification for the calibration of GC instruments in the field
and discussed the use of Tedlar bags for the preparation and transport of cali-
bration samples.
Gay et al. (1976) have used the photolysis of C12: aldehyde: N02 mixtures
in air or oxygen for the preparation of PAN and a number of its homologues at
high yields:
019QQ/A 5-80 6/19/84
-------
C12 + hv » 2 Cl (5-79)
0 0
Cl + RC-H »• R-C' + HC1 (5-80)
0 0
RC- + 02 + M > RC-02' + M (5-81)
0 0
RC-02' + N02 » RC-0 -NO. (5-82)
This procedure has been utilized in a portable PAN generator that can be used
for the calibration of GC-ECD instruments in the field (Grosjean, 1983; Grosjean
et al., 1984). The output of this generator is a dynamic flow of PAN in air
at a concentration of about 2 to 450 ppb. Dilute concentrations of reactant
gases for the photolysis chamber are obtained by passing a controlled flow of
air over CK, NO^, and acetaldehyde permeation tubes.
The other technique for PAN preparation in current use involves the
nitration of peracetic acid. In the 1969 review (Stephens, 1969), this approach
was considered not useful for synthesis. Several investigators, however, have
recently reported on a condensed-phase synthesis of PAN with peracetic acid
that produces high yields of a pure product free of other alkyl nitrates
(Hendry and Kenley, 1977; Kravetz et al., 1980; Nielsen et al., 1982; Holdren
and Spicer, 1984). Most of these procedures call for the addition of peracetic
acid (40 percent in acetic acid) to a hydrocarbon solvent (pentane, heptane,
octane) maintained at 0°C in a dry-ice acetone bath, followed by acidification
with sulfuric acid. Nitric acid is formed jjn situ with stirring by the slow
addition of sodium nitrate. After the nitration is complete, the hydrocarbon
fraction, containing PAN concentrations of 2 to 4 mg/ml (Nielsen et al.,
1982), can be stored at -20°C for periods longer than a year (Holdren and
Spicer, 1984). After analysis, the PAN-hydrocarbon solutions can be used
directly for calibration by the evaporation of measured microliter volumes of
solution into Tedlar bags containing known volumes of clean air.
019QQ/A 5-81 6/19/84
-------
The most direct method for absolute analysis is by infrared absorption
using absorptivities given in Table 5-13. This is the technique used by
Stephens (1969; analysis of PAN in N_ cylinders), Lonneman et al. (1976;
analysis of gas-phase products from photolysis of ethyl nitrite); and Holdren
and Spicer (1984; analysis of PAN in octane solutions). Whereas long, folded-
path cells and FTIR spectrometry are required for the analysis of atmospheric
PAN, conventional IR instruments and 10-cm gas cells can analyze gas standards
with concentrations greater than 35 ppm (Stephens, 1969) and Holdren and Spicer
(1984) used 50-um liquid microcells for the analysis of PAN in octane solutions.
Another candidate technique for absolute PAN analysis is gas-phase coulometry
using a tandem electron-capture detector (Lovelock et al., 1971). Singh and
Sal as (1983) have shown, however, that this technique is unsuitable for absolute
PAN analysis because a significant fraction of the PAN is destroyed prior to
coulometric detection.
The alkaline hydrolysis of PAN to acetate ion and nitrite ion in quantita-
tive yield (Nicksic et al., 1967) provides a means independent of infrared for
the quantitative analysis of PAN. Molecular oxygen is also produced in quanti-
tative yield by the reaction (Stephens, 1967):
0 0
CH-COONO + 20H~ = CH,CO~ + 0, + NO," + H,0 (5-83)
O £,. o £.£.£.
The col orimetric determination of nitrite ion with Saltzman reagent was first
used to measure PAN quantitatively (Stephens, 1969; Kravetz et al., 1980).
Nielsen et al. (1982) analyzed the hydrolyzed products of pure PAN samples by
ion chromatography for nitrite and nitrate and found that 4 percent of the
nitrite had been oxidized to nitrate. Some gas-phase PAN calibration samples
(e.g., photolysis of Cl?: acetaldehyde: N0~) contain impurities such as N0?
that will yield nitrite and nitrate in aqueous solution. Thus, Grosjean
(1983) and Grosjean et al. (1984) performed ion chromatographic analysis of
the acetate ion to determine the PAN output of a portable generator.
An alternate calibration procedure has been proposed based on the thermal
decomposition of PAN in the presence of excess NO (Lonneman et al. , 1982;
Lonneman and Bufalini, 1982). The peroxy radical, CH3C(0)02, and its decompo-
sition products rapidly oxidize NO to N0_. In the presence of a small amount
of benzaldehyde, which is used to scavenge the hydroxyl radical and control
019QQ/A 5-82 6/19/84
-------
the stoichiometry, simulation models predict that 5 molecules of NO will be
removed per PAN molecule present. By the use of NO and PAN standard mixtures
and the chemiluminescent measurement of the NO consumed, the experimental
value was determined to be ANO/APAN = 4.7 ± 0.2. While it will not give the
precision of which the techniques above are capable, this measurement could be
performed in field stations where chemiluminescent NO analyzers are usually
available.
5.5.7.4 Methods of Analysis of Higher Homologues. The GC-ECD analyzer is
likewise used for the higher homologues of PAN (Darley et a!., 1963; Stephens,
1969; Heuss and Glasson, 1968). The higher homologues elute with longer reten-
tion times. The first observation of PPN in heavily polluted air was by Darley
et al. (1963) who also measured peroxybutyryl nitrate in synthetic mixtures by
GC-ECD. The concentrations of the higher homologues in ambient air are usually
below the detection limits of the GC-ECD technique. Heuss and Glasson (1968)
measured PBzN in irradiated auto exhaust samples by GC-ECD and reported that
this homologue was 100 times more potent than PAN as a lachrymator. The direct
analyses of PBzN by GC-ECD is reported to be complicated by interferences (Appel,
1973). Therefore, an analytical technique was developed in which the PBzN was
quantitatively hydrolyzed to methyl benzoate (MeOBz), followed by GC analysis for
MeOBz using a flame ionization detector (Appel, 1973). An upper limit of 0.07 ppb
was placed on the concentration of PBzN in the San Francisco bay area. The ana-
lysis for the higher homologues of PAN in the atmosphere by FTIR spectrometry is
not feasible because of inadequate sensitivity, although Hanst et al. (1982)
placed an upper limit for PBzN in smoggy Los Angeles air of 1 ppb based on ab-
sorption in the 990 cm region.
The higher homologues of PAN may be prepared in the same manner as PAN by
the use of a compound containing the parent alky! group. Thus, PPN and PBzN
have been prepared by the photolysis of alkyl nitrates in oxygen (Stephens,
1969) and parent aldehydes plus chlorine and N02 (Gay et al., 1976). The
study of Gay et al. (1976) confirmed that the first member of the series,
peroxyformyl nitrate [HC(0)OpNO?], is too unstable to be observed. There have
been few reports of the absolute analysis for the higher homologues. Infrared
absorption analysis of purified samples should be the preferred technique.
Infrared absorptivities of homologues have been reported by Stephens (1969)
and Gay et al. (1976).
019QQ/A 5-83 6/19/84
-------
5.5.8 Methods for Sampling and Analysis of Hydrogen Peroxide
5.5.8.1 Introduction. Hydrogen peroxide (H?0?) is expected to be mechanisti-
cally significant in photochemical smog as a chain terminator and as an index
of the hydroperoxyl radical (H0?) concentration (Bufalini, 1969; Demerjian et
al., 1974). The major reaction leading to the formation of H^O- is the recom-
bination of the hydroperoxyl radical (Graedel, 1976):
HO^ + HO^ + M = H202 + 02 + M (5-84)
Recent studies have implicated atmospheric H?CL in the aqueous-phase oxidation
-2
of SO- to SO. and in the acidification of rain (Penkett et al., 1979; Dasgupta,
1980; Martin and Damschen, 1981; Overton and Durham, 1982). (As described in
chapter 4, however, it is not the predominant mechanism, since the OH radical
appears to be the main agent for the homogeneous gas-phase oxidation of S02 to
SO-, which is subsequently oxidized to SO.).
Some controversy appears to exist concerning the time and concentration
correlation between HpO? and 0_. If this correlation could be assessed, then
the concentration of H202 could be predicted from the concentration of 03-
According to one report (Gay and Bufalini, 1972), the H202 concentration
reached a maximum of 0.18 pphm at the same time as the Og maximum of 0.65 pphm
(at 3:00 pm) in the urban atmosphere of Riverside, California. On the other
hand, Kok et al. (1978a) did not observe any definitive correlation between 03
and H_02 in the same area. This discrepancy may result in large part from the
use of different measurement methods. One of the major problems in assessing
the role of atmospheric H?0? has been a lack of adequate measurement methodol-
ogy. Suitable techniques may now be available for aqueous HJ),,, but recent
studies (Zika and Saltzman, 1982; Heikes et al., 1982) have cast doubt on the
validity of methods for atmospheric H202 that use aqueous sampling because of
interfering reactions of absorbed 0-. Techniques that have been used or
proposed for aqueous- and gas-phase H202 are discussed below, as well as
methods for sampling, generation, and standardization of H202 samples.
5.5.8.2 Sampling. Almost all of the methods used for the measurement of
atmospheric H-Op have used aqueous traps for sampling. In the method used by
Kok et al. (1978b), atmospheric samples containing H202 were collected in
aqueous solution using midget impingers. A continuous extraction process for
sampling H?0_ from the atmosphere and concentrating it in the aqueous phase
for chemiluminescence measurement has also been described by Kok et al. (1978a).
019QQ/A 5-84 6/19/84
-------
The apparatus used by Zika and Saltzman (1982) to sample H202 in ambient
air consists of two 500-ml fitted gas-washing traps, each containing 500 ml of
water, a vacuum pump, and a flow meter. Teflon tubing was used to connect the
various components and to draw air into the apparatus. The traps were filled
with distilled water and air was drawn through them at a controlled flow rate.
At various intervals, the flow was briefly interrupted, and samples were
withdrawn from the traps for H202 analyses. Prior to analysis, samples were
degassed with high-purity helium to remove any 0» that may have been present.
The efficiency of the system for extraction of H202 from the air was tested.
The first trap was found to be 99 percent efficient in removing hL02 from the
gas stream; and over a concentration range of 10"8 to lo"3 M, no Hp02 was
detected in the second trap.
The most serious problem with methods that use aqueous traps is their
potential interference from atmospheric 0_, which is present in much higher
concentrations. The recent study of Zika and Saltzman has shown that absorbed
03 leads to both the formation and destruction of hL02 (Zika and Saltzman,
1982). Details of the aqueous chemistry of 03 can also be found in other
sources (Hoigne and Bader, 1976; Kilpatrick et al., 1955; Taube and Bray,
1940). An obvious research need in H202 measurements is a clear delineation
of the nature of any 03 interferences and the development of means for their
prevention.
5.5.8.3 Measurement. A number of methods for measuring low levels of H_0?
have been reported, including the following:
1. Titanium colorimetric methods.
2. Chemiluminescence methods.
3. Enzyme-catalyzed methods.
4. Fourier-transform infrared method.
5. Electrochemical methods.
6. H202~olefin reaction.
7. Mixed-!igand complexes.
8. lodometry.
Of these, techniques I through 3 above have been used for atmospheric analysis
and will be emphasized. Methods 4 through 7 have not been used for atmos-
pheric analysis and will be only briefly summarized. lodometric techniques
are useful only for calibration and will be discussed in that section.
019QQ/A 5-85 6/19/84
-------
5.5.8.3.1 Titanium colorimetric methods. The titanium sulfate-8-quinolinol
reagent method has been used in several studies on atmospheric H202 (Bufalini
et al., 1972; Gay et al., 1972a; Gay et al. , 1972b; Kok et al. , 1978a).
Hydrogen peroxide in air is scrubbed in a coarse-fritted bubbler containing
aqueous titanium sulfate-ammonium sulfate-sulfuric acid solution at a concen-
tration of approximately 50 pg/ml of titanium (IV). After sampling, the pH of
the solution is adjusted to 4.2 ± 0.2 and the mixture is extracted with an
aliquot of 0.1 percent 8-quinolinol in chloroform. The absorbance at 450 nm of
the titanium (IV)-H202~8-quinolinol complex in chloroform is determined.
This method has the following advantages. The molar absorptivity of the
titanium(IV)-H202-8-quinolinol complex at 450 nm is 3060 M cm ; i.e., the
method is very sensitive. The time between sample preparation and absorbance
readings can be short. At pH 4.2 ± 0.2, the stoichiometry of the titanium
(IV)-H202 complex is 1 to 1 (Babko and Volkova, 1951). The major disadvantage
of the method is that the 8-quinolinol reagent is highly pH-sensitive; i.e.,
the absorption maximum occurs in a narrow pH range of 4 to 5. The linear
dynamic range of Beer's law for this method lies between 0 and 6 ug/ml. The
sensitivity of this method is 1.6 x 10 M per 0.005 absorbance unit in a 1-cm
cell.
A positive interference is expected from any compound that can liberate
H?0? via acid hydrolysis (Pobiner, 1961). Accordingly, t-butyl hydroperoxide
gives rise to the titanium(IV)-H2Q2 complex. Of the major atmospheric pollut-
ants investigated'(S02, 03, N02, NO, and hydrocarbons), only S02 at high
concentrations gave a small (0.7 percent) negative interference (Gay et al. ,
1972b). Other compounds tested by Cohen and Purcell (1967)--peracetic acid,
ethyl hydroperoxide, n-butyl hydroperoxide, acetyl peroxide, and peroxyacetyl
nitrate (PAN)--were found to give slight negative interferences at high con-
centrations.
The titanium tetrachloride method of analysis for hydrogen peroxide is
described by Pilz and Johann (1974) and Kok et al. (1978a). Samples are col-
lected in a midget impinger containing an aqueous titanium tetrachloride-hydro-
chloric acid (TiCK-HCl) solution at a concentration of approximately 8 mg/ml
of titanium(IV). A stable TiCl4-H202 complex is formed immediately, and,
after dilution to a known volume, the absorbance of the complex at 410 nm is
determined (molar absorptivity = 735 M^cnf1). For H202 concentrations less
than 100 ppb, 5-cm cells and 0.05 absorbance full-scale were used. The princi-
019QQ/A 5-86 6/19/84
-------
pal difficulty with this method is the formation of fine particles, presumably
TiOp, in solution. The presence of suspended precipitate scatters visible
radiation and creates an apparent absorption. Another problem is that the
reagent is quite acidic and hygroscopic. Kok et al. (1978a) has compared
measurements with both titanium reagents and chemiluminescence.
5.5.8.3.2 Chemi1uminescence. Hydrogen peroxide in the atmosphere may be
detected at low concentrations by the chemiluminescence obtained from Cu(II)-
catalyzed oxidation of luminol (5-amino-2,3-dihydro-1,4-phthalazinedione) by
H,0_ (Armstrong and Humphreys, 1965). The reagent is a solution containing
_c
luminol, 1 x 10 M Cu(II), and NaOH (pH = 12.8). The products of the reaction
with HpO. are 3-amino-phthalic acid, a nitrogen molecule, and a photon of
light at 450 nm. The detection limit for atmospheric samples has been given
as 0.001 ppm, and the linear dynamic range is 0.001 to 1 ppm.
A small positive interference was reported for PAN (Kok et al., 1978b).
If 0- absorption leads to the formation of H^O- (Zika and Saltzman, 1982;
Heikes et al., 1982), as reported, then 0., is a major interference. There
have also been undocumented reports of a negative interference from SO-- The
exact mechanism of the chemiluminescent oxidation of luminol is not known, but
involves the decomposition of H202 by the copper (II) catalyst (Delumyea,
1974; Burdo and Seitz, 1975).
5.5.8.3.3 Enzyme-Catalyzed Methods (Peroxidase). This general method involves
three components: a substrate that is oxidizable; the enzyme, horseradish
peroxidase (HRP); and H-Op. The production or decay of the fluorescence in-
tensity of the substrate is measured as it is oxidized by H202, catalyzed by
HRP. Some of the more widely used chromogenic substrates are scopoletin
(6-methoxy-7-hydroxy-l,2-benzopyrone) (Andreae, 1955; Perschke and Broda,
1961); 3-(pj-hydroxyphenyl)propionic acid (HPPA) (Zaitsu, 1980); and leuco
crystal violet (LCV) (Mottola et al. , 1970).
In the scopoletin method, the reagent solution is mixed with a second
solution containing the H-O^, the concentration of which must not be less than
0.33 nor more than 0.84 times the concentration of scopoletin (Perschke and
Broda, 1961). The disappearance of scopoletin fluorescence is monitored and
the fluorescence intensity can be used to obtain the concentration of H»02
from a calibration curve. The most significant advantages of this method of
analysis are the specificity for H202, the sensitivity, and the stoichiometry
of the scopoletin:H?0? reaction (1:1 mole per mole as long as scopoletin is
019QQ/A 5-87 6/19/84
-------
present in a 20-fold excess over HRP). Ascorbic acid, glutathione, and mangan-
ous ions are reported to inhibit the oxidation of scopoletin by peroxidase
(Andreae, 1955). As the pH of the mixture deviates from 10, the method produces
less reliable results. The chief disadvantage of the scopoletin method is
that the concentration of hLO. must be within a narrow range in order to
obtain an accurately measurable decrease in fluorescence. This limits the
usefulness of the technique in determining unknown H,,0? concentrations over
several orders of magnitude (Armstrong and Humphreys, 1965; Andreae, 1955).
Detection limits for this technique are quite low and are in the range of 1.5
x 10"11 M (Perschke and Broda, 1961) to 2 x 10"10 M (Andreae, 1955).
With the leuco crystal violet (LCV) substrate, intensely colored crystal
violet is formed from the reaction of H?0? with LCV, catalyzed by HRP. In the
method described by Motto!a et al. (1970), the substrate is added to an aqueous
sample containing HpOp. Upon addition of a solution of HRP, the solution
turns violet. The absorbance is measured at 596 nm, where the absorption
coefficient of crystal violet is 10 M cm , a very high value and an inherent
advantage of this method. The concentration of HJ)- is a linear function of
the concentration of crystal violet produced. The detection limit reported is
~8
about 10 M H_0? for an absorbance of 0.005 in a 5-cm cell.
The HRP catalyzes the oxidation of a wide variety of hydrogen-donating
substrates by H^CL. Zaitsu and Ohkura (1980) have tested a number of 4-hydroxy-
phenyl compounds and found that 3-(g-hydroxyphenyl) propionic acid (HPPA)
provided the most sensitive and rapid means for determining HpOp. When HPPA
reagent solution is mixed with HRP solution and a test solution containing
H?0?, a product is formed that fluoresces at 404 nm following excitation at
320 nm. The intensity of this fluorescence is monitored as a function of H_0,,
-10
concentration. The detection limit was reported to be 10 mole H?0? with a
-8
linear range extending to 10 mole H202 when a test solution of only 0.1 ml
volume was used. Presumably the molar sensitivity could be improved by the
use of larger sample volumes. No interference studies were reported.
The enzymatic methods appear to be the most promising aqueous, colori-
metric methods for Hp02 and have considerably greater sensitivity than the
methods employing titanium reagents. Studies of potential atmospheric inter-
ferences, however, have not been reported for any of these three substrates.
5.5.8.3.4 Other Methods. Hydrogen peroxide has been monitored in the gas phase
at ppm concentrations in laboratory mechanistic studies by the use of long-path
019QQ/A 5-88 6/19/84
-------
FTIR absorption using the Hfln ^anc^ in tne re9i°n> 1200-1350 cm (Su et a].,
1979; Niki et al., 1980). The band most suitable for quantitative analysis is
centered at 1267 cm . The infrared absorptivity at 1250 cm has recently
been measured and a value of 8.4 cm atm obtained at 1-cm resolution (Hanst
et al. , 1981). An attempt was made to observe H?0? absorption during an
intense smog episode in Pasadena, California, and the possible presence of
0.070 ppm H_0_ was reported (Hanst et al., 1975). Unfortunately, the minimum
detection limits in this region are severely compromised by neighboring absorp-
tion bands of FLO and ChL. Hanst et al. (1981) have discussed the problems
with FTIR measurements of H^O^ and have assigned a minimum detection limit of
0.040 ppm for a 1-km total pathlength. Thus, concentrations of H?0_ as high as
0.010 to 0.030 ppm would not be detectable by FTIR as currently used.
The redox system, Op/H^O^, lends itself to electrochemical analysis.
This system is complex because of the presence of three related redox systems:
0_/H202, H-Op/H^O, 02/H»0. Accordingly, to determine the concentration of
HpO~ in aqueous media, the indicator electrode must be adequately specific
toward the 02/H202 redox system. Pisarevskii and Polozova (1980) used a
modified graphite electrode (MGE) consisting of a thin layer of graphite
etched onto electron-conducting silicate glasses. With a reference electrode
of Ag/AgCl/KCl (saturated), H?0~ could be measured over the range of 5 x 10
M to 1 M.
Hauser and Kolar (1968) reported an interference from HJ)- in the reaction
between 0~ and l,2-di-(4-pyridyl)ethylene (DPE) and investigated this reaction
for the measurement of H«0 The reaction of DPE with 03 or H^Op produces
pyridine-4-aldehyde, which may be analyzed colorimetrically with 3-methyl-2-
benzothiazolinone hydrazone (MBTH) reagent. The detection limit reported for
—fi —4
H_0? was approximately 10 M with a linear range extending to about 5 x 10
M. No atmospheric applications have been reported. Atmospheric 0_ would be a
major interference.
Mixed-!igand complexes of the type M:L:H»Op provide sensitive means for
the determination of HpO^. In particular, vanadium(V)-xylenol orange (XO)
chelates, vanadium hydroxamic acid chelates, and uranium hydroxamic acid
chelates provide good examples of sensitive mixed ligands (Csanyi, 1981;
Meloan, 1961). A vanadium(V)-XO reagent can be used for the spectrophotometric
determination of 10 M concentrations of H_0? (Csanyi, 1981). Methods using
vanadium hydroxamic acid chelates (VHAC) (I) and uranium hydroxamic acid
019QQ/A 5-89 6/19/84
-------
chelates (UHAC) (II) measure the concentration of the chelate remaining after
reaction between the chelate and H?0p. These methods are reportedly specific
for hLCL. The following alkyl peroxides were found not to interfere in the
determination of H^O^: tert-butyl peroxide, di-tart-butyl peroxide, tert-butyl
perbenzoate, hydroxyheptyl peroxide, and lauroyl peroxide (Meloan et al.,
1961). This method can be used to detect a total concentration of 7 x 10
mole of H?0p if the uranium system is used and 1 x 10 mole if vanadium is
used. Interferences can be expected among other strong oxidizing agents that
are water-soluble and are extracted into the aqueous phase along with H^Op.
5.5.8.4 Generation and Calibration Methods
5.5.8.4.1 Generation. Standard samples containing trace concentrations of
H-Op are required for testing and calibration of various measurement methodol-
ogies. As with 0-, such standards are not available and are usually prepared
at the time of use. A number of techniques have been employed for generating
aqueous standards, but convenient methods for the generation of gas-phase
standards are noticeably lacking.
Techniques for the generation of high concentrations of HLO^ have been
discussed by Shanley (1948). Commercial solutions of 30 percent aqueous H202
are readily available. Trace levels of H^O,, in water may be generated by the
60
irradiation of water with CO yradiation (Hochanadel, 1952; Armstrong and
Humphreys, 1965) and by enzymatic means (Andreae, 1955). By far the most
convenient method for generating aqueous standards containing micromolar
concentrations of HJ*? 1S simPlv the serial dilution of commericial-grade 30
percent HJdy (Fisher Analytical Reagent). Samples prepared in this manner
must be standardized and the method usually employed is the iodometric tech-
nique discussed below. Stock standard solutions of H?0? as low in concentration
. t. C-
as 10 M have been found to be stable for many weeks if kept in the dark
(Armstrong and Humphreys, 1965).
Techniques for the convenient generation of gas-phase standards are not
available. With the increased interest in atmospheric H^O,,, there is an
obvious need for an H?02 generator comparable to the photolytic 03 generator
discussed in section 5.5.5.1. A technique that has often been used for gener-
ating ppm concentration levels of H202 in air has been described by Cohen
and Purcell (1967). Microliter quantities of 30 percent H202 solution are
injected into a metered stream of air that flows into a Teflon bag. The con-
centration of H202 in the bag is then determined by the iodometric method
discussed below.
019QQ/A 5-90 6/19/84
-------
5.5.8.4.2 Calibration. By far the most common method for standardizing low
concentrations of hLO,, is based on iodometry (Allen et al., 1952; Hochanadel,
1952; Cohen et al., 1967). Hydrogen peroxide liberates iodine from an iodide
solution quite slowly, but in the presence of a molybdate catalyst the reaction
is quite rapid. The iodine liberated may be determined by titration with
standard thiosulfate at higher concentrations or by photometric measurement of
the tri-iodide ion at low concentrations. The molar absorption coefficient of
4
the tri-iodide ion at 350 nm has been determined to be 2.44 x 10 by measuring
-5 -4
10 to 10 M H2Q2 solutions prepared from 0.2 M stock tiJ^y solution standard-
ized against primary grade arsenious oxide (Armstrong and Humphreys, 1965).
The stoichiometry is apparently 1 mole of iodine released per mole of Hp09.
Definitive studies of the stoichiometry, however, have not been performed to
the same extent as those of the stoichiometry for the iodometric determination
of 03.
5.6 SUMMARY
5.6.1 Properties
Ozone, peroxyacetyl nitrate, and hydrogen peroxide, along with other photo-
chemical oxidants occurring at very low concentrations in ambient air, are
characterized chiefly by their ability to remove electrons from or to share
electrons with other molecules or ions (i.e., oxidation). The capability of a
chemical species for oxidizing or reducing other chemical species is termed
"redox potential" (positive or negative standard potential) and is expressed in
volts. Ozone has a standard potential of +2.07 volts in aqueous systems (for
the redox pair, 0.,/H?0). Hydrogen peroxide has a standard potential of +1.776
in the redox pair, H20p/H20. No standard potential for peroxyacetyl nitrate in
neutral or buffered aqueous systems, such as those that occur in biological sys-
tems, appears in the literature. In acidic solution (pH 5 to 6), PAN hydrolyzes
fairly rapidly; in alkaline solution it decomposes with the production of nitrite
ion and molecular oxygen.
The toxic effects of oxidants are attributable to their oxidizing ability.
Their oxidizing properties also form the basis of the measurement techniques
for all three of these pollutants. The calibration of ozone and PAN measurements,
however, is achieved via their spectra in the ultraviolet and infrared, respec-
tively. The calibration of measurement methods for H202 is achieved with iodo-
metric techniques that depend on the oxidizing properties of H202-
019QQ/A 5-91 6/19/84
-------
An important property of PAN, especially in the laboratory, is its thermal
instability. Its explosiveness dictates its synthesis for calibration purposes
by experienced personnel only. All three pollutants must be generated jjn situ
for the calibration of measurement techniques. For ozone and HpO,,, generation
of calibration gases is reasonably straightforward.
5.6.2 Reactions of Ozone and Other Oxidants in Ambient Air
The atmospheric reactions of ozone and of other photochemical oxidants such
as peroxyacetyl nitrate (PAN) and hydrogen peroxide (H?0,,) are complex and diverse.
but are becoming increasingly well-character!'zed. The reactions of these species
result in products and processes that may have significant environmental and
health- and welfare-related implications, including effects on biological systems,
nonbiological materials, and such phenomena as visibility degradation and acidi-
fication of cloud and rain water. Ozone may play a role in the oxidation of SO^
to H7SO., both indirectly in the gas phase (via formation of OH radicals and
Criegee biradicals) and directly in aqueous droplets. Evidence is also accumu-
lating that hydrogen peroxide, like ozone, is involved in both gas-phase photo-
chemistry and aqueous-phase oxidations. For example, studies of the rates of
oxidation of S0? by HJ)2 in solution suggest that this reaction is sufficiently
fast that it could be the major aqueous-phase route for the oxidation of S02
under certain atmospheric conditions. In addition, the importance of oxidants
such as PAN in various aspects of atmospheric chemistry, such as long-range
transport of NO and multi-day air pollution episodes, is now being recognized.
/\
Ozone can react with organic compounds in the troposphere. It is important
to recognize, however, tht organics undergo competing reactions with OH radicals
in the daytime (Atkinson et al., 1979; Atkinson and Lloyd, 1984) and, in certain
cases, with NO., radicals during the night (Japar and Niki, 1975; Carter et al.,
1981a; Atkinson et al., 1984a,b,c,d), as well as photolysis. Only for organics
whose ozone reaction rate constants are greater than ~10 cm molecule sec
can consumption by ozone be considered to be atmospherically important (Atkinson
and Carter, 1984).
Ozone reacts rapidly with the acyclic mono-, di-, and tri-alkenes and with
"*1 ft
cyclic alkenes. The rate constants for these reactions range from ~10 to
~10~14 cm3 molecule"1 sec"1 (Atkinson and Carter, 1984), corresponding to
atmospheric lifetimes ranging from a few minutes for the more reactive cyclic
alkenes, such as the monoterpenes, to several days. In polluted atmospheres,
a significant portion of the consumption of the more reactive alkenes will
019QQ/A 5-92 6/19/84
-------
occur via reaction with ozone, rather than with OH radicals, especially in the
afternoons during photochemical oxidant episodes. Reactions between ozone and
alkenes can result in aerosol formation (National Academy of Sciences, 1977;
Schuetzle and Rasmussen, 1978), with alkenes of higher carbon numbers the chief
contributors.
Because of their respective rate constants, neither alkanes (Atkinson and
Carter, 1984) nor alkynes (Atkinson and Aschmann, 1984) are expected to react
with ozone in the atmosphere, since competing reactions with OH radicals have
higher rate constants.
The aromatics react with ozone, but quite slowly (Pate et al., 1976; Atkinson
et al., 1982), such that their reactions with ozone are expected to be unimpor-
tant in the atmosphere. Cresols are more reactive toward ozone than the aro-
matic hydrocarbons (Atkinson et al., 1982), but their reactions with OH radicals
(Atkinson et al., 1978, 1982) or N03 radicals (Carter et al., 1981a; Atkinson
et al., 1984a) predominate.
For oxygen-containing organic compounds, especially those without carbon-
carbon double bonds, reactions with ozone are slow. For carbonyls and ethers
(other than ketene) that contain unsaturated carbon-carbon bonds, however,
much faster reactions are observed (Atkinson et al., 1981).
The kinetics of the reactions of ozone with a variety of nitrites, nitriles,
nitramines, nitrosamines, and hydrazines have been studied (Atkinson and Carter,
1984), but only for the hydrazines are these reactions sufficiently rapid to be
of atmospheric importance. Chamber studies have shown that N-nitrosodimethyl-
amine can result from the reaction of ozone with simple hydrazines (Tuazon et
al., 1981a). Whether this product would ever be formed by reaction with ozone
in the atmosphere obviously depends upon the presence, and level, of the pre-
cursor hydrazines in ambient air.
Certain reactions of ozone other than its reactions with organic compounds
are important in the atmosphere. Ozone reacts rapidly with NO to form NOp, and
subsequently with NO, to produce the nitrate (NO.,) radical and an oxygen
molecule. Photolysis of ozone can be a significant pathway for the formation
of OH radicals, particularly in polluted atmospheres when ozone concentrations
are at their peak.
5-6.3 Reactions of Ozone and Peroxyacetyl Nitrate in Aqueous (Biological)
Systems
Both ozone and PAN can react directly and rapidly with many organic mole-
cules, including many types occurring in biological systems. Additionally,
019QQ/A 5-93 6/19/84
-------
active species such as singlet oxygen, hydroxyl radicals, and superoxide are
produced either as products of primary reactions or from the decomposition of
ozone or PAN in water; these species also have the potential for causing bio-
logical damage.
The reactions of ozone with biologically important functional groups have
been described in the literature, although such information remains relatively
sparse and is based on iji vitro work that is not always pertinent to reactions
that occur under the conditions of jj} vivo exposure. Among the functional
groups with which ozone reacts relatively rapidly are the carbon-carbon double
bonds (alkenic group) found in biologically important compounds such as some
of the essential fatty acids and polyunsaturated fatty acids (PUFA) of the kind
found in the lipids of cell membranes. Amines are, in general, close to alkenes
in their reactivity toward ozone, although amino groups existing as the amide
or salt are less susceptible to ozone than unprotected amino groups. Sulfur-
containing compounds, such as methionine, can also undergo electrophilic attack
by ozone, resulting in the formation of both sulfoxides and sulfones. Under
some conditions (e.g., pH > 9), ozone is rapidly converted to hydroxyl radicals,
which are less selective than ozone in reactions with organic molecules. The
conversion of ozone to superoxide (Op*) and hydroperoxy radicals (H0««) has
also been reported.
Aromatic compounds are much less reactive toward ozone than alkenic com-
pounds in aqueous systems. In compounds containing both aromatic and alkenic
groups, such as the indole ring of tryptophan, the initial ozone attack occurs
exclusively at the alkenic part of the molecule. Aldehydes react with ozone
with and without the involvement of oxygen. Either way, acyl hydrotrioxides
are formed that subsequently decompose to peroxides and carboxylic acids. Sim-
ple ketones react slowly or not at all with ozone.
Reactions between ozone and specific molecules of importance in biological
systems have been described in chapter 10.
Knowledge of the solution chemistry of PAN is limited. It is known, how-
ever, that PAN can react with alkenes, with sulfur-containing compounds, and
with aldehydes. The half-life of PAN in water (pH 7.2) is only about 4.4 min-
utes. Thus, some of the toxicological effects ascribed to PAN should possibly
be attributed to its decomposition products instead.
019QQ/A 5-94 6/19/84
-------
5.6.4 Sampling and Measurement of Ozone and Other Photochemical Oxidants
The analysis of ozone and other, related atmospheric oxidants includes
requirements for representative sampling, specific and sensitive measurement
methodologies, methods for the generation of standard samples, absolute methods
for the calibration of these standards, and procedures by which to provide
quality assurance for the whole measurement process. In the summary presented
below, recommended procedures are given for all of these analytical steps.
Sampling and quality assurance are discussed in general terms for all of the
oxidants. Methods of analysis, sampling, and generation and calibration are
discussed specifically for ozone, peroxyacetyl nitrate (PAN) and its homologues,
and hydrogen peroxide. Because of the large existing data base that employed
measurements for "total oxidants," non-specific iodometric techniques are
discussed and compared to current specific 0- measurements.
5.6.4.1 Quality Assurance and Sampling. A quality assurance program is
employed by the U.S. Environmental Protection Agency for assessing the accuracy
and precision of monitoring data and for maintaining and improving the quality
of ambient air data. Procedures and operational details have been prescribed
in each of the following areas: selection of analytical methods and instrumen-
tation (i.e., reference and equivalent methods); method specifications for
gaseous standards; methods for primary and secondary (transfer standards)
calibration; instrumental zero and span check requirements, including frequency
of checks, multiple-point calibration procedures, and preventive and remedial
maintenance requirements; procedures for air pollution episode monitoring;
methods for recording and validating data; and information on documentary
quality control (U.S. Environmental Protection Agency, 1977).
A crucial link in the measurement cycle involves sampling strategies and
techniques. Sampling strategies, which involve the design and operation of a
sampling network, must be consistent with the specific purpose of the measure-
ments. Ambient air monitoring data are collected for a variety of purposes,
each of which may have different requirements that affect sampling strategy.
For example, a sampling strategy for health effects research studies may
require a number of monitoring stations carefully situated to assess human
exposure for a given urban population over a finite period of time. In addition,
since ozone, PAN, and H«02 are all secondary pollutants formed after an initial
induction period, stations for monitoring peak concentrations should be located
downwind of the urban center of precursor emissions.
019QQ/A 5-95 6/19/84
-------
The reactivity and instability of CL and other photochemical oxidants
dictate special sampling techniques. Samples of air containing 0. cannot be
collected and stored and must be analyzed dynamically. Analysis for PAN must
likewise be performed as soon as possible after collection. Hydrogen peroxide
collected in aqueous media is fairly stable, but samples are subject to inter-
fering reactions from ozone trapped in solution (section 5.5.8). Sampling
probes should be constructed of Teflon or some similarly inert material, and
inlet residence times should be as short as possible. Other design criteria for
0_ monitoring stations have been given (Standing Air Monitoring Work Group,
«3
1977; National Academy of Sciences, 1977b). The most important of these are
that inlets of the sampling probe should be positioned 3 to 15 m (10 to 49 ft)
above ground and at least 4 m (13 ft) from large trees and 120 m (349 ft) from
heavy automobile traffic.
5.6.4.2 Measurement Methods for Total Oxidants and Ozone. Techniques for the
continuous monitoring of total oxidants and 0- in ambient air are summarized in
Table 5-13. The earliest methods used for routinely monitoring oxidants in the
atmosphere were based on iodometry. When atmospheric oxidants are absorbed in
potassium iodide (KI) reagent, the iodine produced,
03 + 3l" + H20 = I3" + 02 + 20H~ (5-85)
is measured by ultraviolet absorption in colorimetric instruments and by ampero-
metric means in electrochemical instruments. The term "total oxidants" is of
historical significance only and should not be construed to mean that such
measurements yield a sum of the concentrations of the oxidants in the atmosphere.
The various oxidants in the atmosphere react to yield iodine at different rates
and with different stoichiometries. Only ozone reacts immediately to give a
quantitative yield of iodine. Hydrogen peroxide, for example, produces iodine
at a slow rate and because of its low concentration compared to ozone (see
section 5.5.9) would be expected to have little effect upon a total oxidants
measurement. As discussed below, the total oxidants measurement correlates
fairly well with the specific measurement of ozone, except during periods when
significant N0? and S02 interferences are present. The major problem with the
total oxidants measurement was the effect of these interferences. Total oxidants
instruments have now been replaced by specific ozone monitors in all aerometric
networks and in most research laboratories. Biases among and between "total
oxidants" and ozone methods are still important, however, for evaluating existing
data on health and welfare "effects levels" where concentrations were measured
by "total oxidants" methods.
019QQ/A 5-96 6/19/84
-------
The reference method promulgated by EPA for compliance monitoring is the
chemiluminescence technique based on the gas-phase ozone-ethylene reaction (U.S.
Environmental Protection Agency, 1971). The technique is specific for ozone,
the response (luminescence intensity) is a linear function of concentration,
detection limits of 0.001 to 0.005 ppm are readily obtained, and response times
of 30 seconds or less are readily obtained. Prescribed methods of testing and
prescribed performance specifications that a commercial analyzer must meet in
order to be designated as a reference method or an equivalent method have been
published by EPA (U.S. Environmental Protection Agency, 1975). An analyzer may
be designated as a reference method if it is based on the same principle as the
reference method and meets performance specifications. Commercial analyzers that
have been designated as reference methods were listed in Table 5-8. An accepta-
ble equivalent method must meet the prescribed performance specifications and
also show a consistent relation with the reference method. Commercial analyzers
that have been designated as equivalent methods were also listed in Table 5-8.
The designated equivalent methods are based on either the gas-sol id chemilu-
minescence procedure or the ultraviolet absorption procedures (Table 5-14). The
first designated equivalent method was based on ultraviolet absorption of the mer-
cury 254 nm emission line. The absorption coefficient of ozone is accurately
known at this wavelength with an accepted value of 134 M cm . Detection limits
of 0.005 ppm are readily obtained by modern digital capabilities for making pre-
cise measurements of weak absorbancies at moderate pathlengths. Compensation
for potential interferences that also absorb at 254 nm is made by comparing an
averaged transmission signal of ozone in air to the transmission signal through
an otherwise identical air sample from which the ozone has been preferentially
scrubbed. Advantages of this UV absorption technique are that a reagent gas
is not required and control of sample air flow is not critical. In addition,
the measurement is in principle an absolute one, in that the ozone concentra-
tion can be computed from the measured transmission signal since the absorption
coefficient and pathlength are accurately known.
In the gas-solid chemiluminescence analyzer, the reaction between ozone and
Rhodamine-B adsorbed on activated silica produces chemiluminescence, the inten-
sity of which is directly proportional to ozone concentration. The sensitivity
is greater than the gas-phase chemiluminescence method and a controlled reagent
gas flow is not required. The sensitivity of the reaction surface decays grad-
ually with time, but the analyzer contains internal means to compensate for the
decay.
019QQ/A 5-97 6/19/84
-------
TABLE 5-14. SUMMARY OF OZONE MONITORING TECHNIQUES
Principle
Continuous
col or i metric
Continuous
electrochemical
f_n
vo Chemi luminescence
CO
Chemi luminescence
Ultraviolet
photometry
Reagent
10(20)% KI
buffered at
pH = 6.8
2% KI
buffered at
pH = 6.8
Ethyl ene,
gas-phase
Rhodamine-B
None
Response
Total
oxidants
Total
oxidants
03-specific
03-specific
03-specific
Minimum
detection limit
0.010 ppm
0.010 ppm
0.005 ppm
0.001 ppm
0.005 ppm
Response
time, 90% FSa
3 to 5 minutes
1 minute
< 30 seconds
< 1 minute
30 seconds
Major
interferences
N02(+20%, 10%RI)
S02(-100%)
N02(+6%)
S02(-100%)
Noneb
None
Species that
absorb at 254 nm
References
Littman and Benoliel (1973)
Tokiwa et al. (1972)
Brewer and Mil ford (1960)
Mast and Saunders (1962)
Tokiwa et al. (1972)
Nederbragt (1965)
Stevens and Hodgeson (1973)
Regener (1960, 1964)
Hodgeson et al. (1970)
Bowman and Horak (1974)
FS = full response.
bA signal enhancement of 3 to 12% has been reported for measurement of 03 in humid versus dry air (California Air Resources Board, 1976).
GNo significant interferences have been reported in routine ambient air monitoring. If abnormally high concentrations of species that
absorb at 254 nm (e.g., aromatic hydrocarbons and mercury vapor) are present, some positive response may be expected. If high aerosol
concentrations are sampled, inlet filters must be used to avoid a positive response.
-------
5.6.4.3 Calibration Methods. All the analyzers discussed above must be periodi-
cally calibrated with ozonized air streams, in which the ozone concentration
has been determined by some absolute technique. This includes the UV absorption
analyzer, which, when used for continuous ambient monitoring, may experience
ozone losses in the inlet system because of contamination.
An ozone calibration system for a primary laboratory calibration system con-
sists of a clean air source, an ozone generator, and a sampling manifold. The
ozone generator most often used is a photolytic source employing a mercury
pen-ray lamp that irradiates a quartz tube through which clean air flows at a
controlled rate (Hodgeson et al., 1972b). The ozone concentration may be
varied by means of a mechanical sleeve over the lamp envelope or by changing
the lamp voltage or current. Once the output of the generator has been cali-
brated by a primary reference method, it may be used to calibrate 0^ transfer
standards, which are portable generators, instruments, or other devices used
to calibrate analyzers in the field. Reference calibration procedures that
have been used for total oxidants and ozone-specific analyzers in this country
are summarized in Table 5-15.
The original reference calibration procedure promulgated by EPA was the
1 percent neutral buffered potassium iodide (NBKI) method (U.S. Environmental
Protection Agency, 1971). This technique was employed in most of the United
States, with the exception of California, which routinely used a 2 percent
NBKI procedure that was quite similar to the EPA method except for the use of
humidified air through the ozone source (California Air Resources Board, 1976).
The Los Angeles Air Pollution Control District (LAAPCD) used a 1 percent un-
buffered KI procedure and measured the iodine produced by a titration technique
rather than the photometric technique used in the California and EPA methods.
A number of studies conducted between 1974 and 1978 revealed several deficien-
cies with KI methods, the most notable of which were poor precision or inter-
laboratory comparability and a positive bias of NBKI measurements relative to
simultaneous absolute UV absorption measurements. The positive bias was also
observed with respect to gas-phase titration (GPT) of ozone with standard
nitric oxide (NO) samples. The positive bias observed is peculiar to the use
of phosphate buffer in the NBKI techniques. The bias was not observed in the
unbuffered LAAPCD method (which nevertheless suffered from poor precision), nor
in the 1 percent EPA KI method without phosphate buffer (Hodgeson et al., 1977),
nor in a KI procedure that used boric acid as buffer (Flamm, 1977). A summary
019QQ/A 5-99 6/19/84
-------
TABLE 5-15. OZONE CALIBRATION TECHNIQUES
Method
1* NBKI
2* NBKI
1* Unbuffered
KI
UV photometry
i— >
o
i_j
Gas-phase
titration (GPT)
]* BAKI
Reagent
1% KI,
phosphate buffer
pH = 6.8
2% KI
phosphate buffer
pH = 6.8
1% KI
pH = 7
None
No standard
reference gas
1% KI,
boric acid buffer
pH = 5
Primary standard3
Reagent grade
arsenious oxide
03 absorptivity at
Hg 254 nm emission
line
No SRM (50 ppm in N2)
from NBS
Standard KI03f
solutions
Method used,
organization,
and dates
EPA
1971-1976
CARB
until 1975
LAAPCD
until 1975
All
1979-present
EPA, States
1973-present
EPA
1975-1979
Bias,
Purpose [Oa^./COg]^
Primary reference 1.12 ± 0.05
procedure
Primary reference 1.20 ± 0.05
procedure
Primary reference 0.96°
procedure
Primary reference
procedure
Alternative reference 1.030 ± 0.0156
procedure (1973-1979)
Transfer standard (1979-present)
Alternative reference 1.00't 0.05
procedure
In the case of the iodometric methods, the primary standard is the reagent used to prepare or standardize iodine solutions.
The uncertainty limits represent the range of values obtained in several independent studies.
C0nly one study available (Demore et al., 1976).
UV photometry used as reference method by CARB since 1975. This technique used as an interim, alternative reference procedure by
EPA from 1976 to 1979.
This is the value reported in the latest definitive study (Fried and Hodgeson, 1982). Previous studies reported biases ranging from
0 to 10 percent (Burton et al., 1976; Paur et al., 1979).
This procedure also recommended a standard I3~ solution absorptivity to be used instead of the preparation of standard iodine solutions.
-------
of results of these prior studies was presented in the previous criteria docu-
ment (U.S. Environmental Protection Agency, 1978) and in a review by Burton
et al. (1976). Correction factors for converting NBKI calibration data to a
UV photometry basis were given in Table 5-5 and discussed in section 5.5.5.2.1.
Subsequently, EPA evaluated four alternative reference calibration proce-
dures based on UV photometry, GPT with excess NO, GPT with excess ozone and the
boric acid buffered KI technique (BAKI). The results of these studies (Rehme
et al., 1981) showed that UV photometry was superior in accuracy, precision,
and simplicity of use; and in 1979 regulations were amended to specify UV
photometry as the reference calibration procedure (U.S. Environmental Protection
Agency, 1979). Laboratory photometers used as reference systems for absolute
0- measurements have been described by Demore and Patapoff (1976), Bass et al.
(1977), and Paur and Bass (1983).
These laboratory photometers contain long path cells (1 to 5 m) and employ
sophisticated digital techniques for making effective double beam measurements
of small absorbancies and low ozone concentrations. A primary standard UV
photometer, such as those above, is one that meets the requirements and specifi-
cations given in the revised ozone calibration procedures (U.S. Environmental
Protection Agency, 1979). Since these are currently available in only a few
laboratories, EPA has allowed the use of transfer standards, which are devices
or methods that can be calibrated against a primary standard and transferred to
another location for calibration of 0- analyzers. Examples of transfer stand-
ards that have been used are commercial 03 photometers, calibrated generators,
and GPT apparatus. Guidelines on transfer standards have been published by EPA
(McElroy, 1979).
5.6.4.4 Relationships of Total Oxidants and Ozone Measurements. The temporal
and quantitative relationship between simultaneous total oxidants and ozone
measurements has been examined in this chapter because of the existence of a
data base obtained by "total oxidants" measurements. Such a comparison is com-
plicated by the relative scarcity of data, the presence of both positive (NOp
and negative (S0_) interferences in total oxidants measurements, and the change
in the basis of calibration. In particular, the presence of N0_ and S02 inter-
ferences prevents the establishment of a definite quantitative relationship be-
tween ozone and oxidants data. Nevertheless, some interesting conclusions can
be drawn and are summarized below.
019QQ/A 5-101 6/19/84
-------
An expected relationship between total oxidants and specific 0_ measurements
O
can be predicted based upon the known response of oxidant instrumentation to
oxidizing and reducing species in the atmosphere. The predicted relation in
this document assumes that NO- is the only significant positive interference
and that S02 is the predominant negative interference. Because of the potential
presence of oxidizing (N02) and reducing (SOp) interferences, it is difficult
or impractical to intercompare measurements during evening and early morning
hours when ozone concentrations are at a minimum. The relationship is best
compared at the midday to early afternoon diurnal maximum of ozone when N0_
concentrations are approaching a minimum; the S02 concentration at this time
will depend on local emissions. A comparison of maximum hourly averages is
appropriate since the primary and secondary ambient air quality standards are
based on this value. If legitimate corrections or compensations have been made
for S02 and N02 interferences, the corrected total oxidants concentrations should
always be higher than simultaneous 0- concentrations by an amount dependent on
type and concentrations of other oxidants present. The major other oxidants
known to exist in the atmosphere are PAN and H-O^. Maximum concentrations of
these oxidants occur near the ozone diurnal maximum (chapter 6) with values
that are only a fraction of the 0» maximum (section 6.6 and 6.7). In addition,
both of these are classified as slow oxidants in that they release iodine at a
slow rate in aqueous solution. Therefore, if a contribution from these species
is discernible at all in the total corrected oxidants reading, it should be only
a small fraction of the ozone contribution. For most of the aerometric data
base, particularly outside the state of California, no attempts were made to
correct total oxidants concentrations for NOp and S0? because such corrections
were impractical or impossible. For uncorrected total oxidants data, the
counterbalancing effects of S02 and N0_ interferences make it even more diffi-
cult to discern contributions from oxidants other than ozone. The uncorrected
total oxidants data should then be either higher than or lower than correspond-
ing ozone data, depending on the relative concentrations of S02 and NO,,.
The simultaneous comparisons that have been made in large part confirm the
predictions above. Studies concluded in the early to mid-1970s were reviewed
in the previous criteria document (U.S. Environmental Protection Agency, 1978).
Averaged data showed fairly good qualitative and quantitative agreement between
diurnal variations of total oxidants and ozone. Monthly averaged data from
Los Angeles (Figure 5-4) and St. Louis (Figure 5-5) are illustrative. These
019QQ/A 5-102 6/19/84
-------
uncorrected data show distinct morning and evening peaks resulting from N0»
interference. Data taken from individual days of this study show considerably
more variation, with total oxidants measurements both higher and lower than
ozone measurements.
The most recent comparison in the literature involved simultaneous ozone and
total oxidant measurements in the Los Angeles basin by the California Air Resources
Board (1978) in 1974, 1976, and 1978. The maximum hourly data pairs were corre-
lated (Chock et al., 1982) and yielded the following regression equation for 1978,
in which a large number (927) of data pairs were available:
Oxidant (ppm) = 0.870 03 + 0.005
(Correlation coefficient - 0.92) (5-86)
The oxidant data were uncorrected for NO- and S02 interferences, and, again, on
individual days maximum oxidant averages were both higher than and lower than
ozone averages.
In summary, specific ozone measurements agree fairly well with total oxi-
dants corrected for NO- and SO- interferences, and in such corrected total oxi-
dants measurements ozone is the dominant contributor to total oxidants. Indeed,
it is difficult to discern the presence of other oxidants in most total oxidant
data. There can, however, be major temporal discrepancies between ozone and
oxidants data, which are primarily a result of oxidizing and reducing inter-
ferences with KI measurements. As a result of these interferences, on any
given day the total oxidant values may be higher than or lower than simultaneous
ozone data. The measurement of ozone is a more reliable indicator than total
oxidant measurements of oxidant air quality.
5-6.4.5 Methods for Sampling and Analysis of Peroxyacetyl Nitrate and Its
Homologues. Only two analytical techniques have been used to obtain
significant data on ambient peroxyacetyl nitrate (PAN) concentrations. These
are gas chroraatography with electron capture detection (GC-ECD) and long-path
Fourier transform infrared (FTIR) spectrometry. Atmospheric data on PAN con-
centrations have been obtained predominantly by GC-ECD because of its relative
simplicity and superior sensitivity. These techniques have been described in
this chapter along with attendant methods of sampling. Since PAN is thermo-
dynamically unstable, standards must be generated and analyzed by some absolute
technique for the purpose of calibrating the GC-ECD. Thus, PAN generation
019QQ/A 5-103 6/19/84
-------
techniques and absolute methods for analyzing these samples have also been
summarized.
By far the most widely used technique for the quantitative determination
of ppb concentrations of PAN and its homologues is GC-ECD (Stephens, 1969).
With carbowax or SE30 as a stationary phase, PAN, peroxypropionyl nitrate (PPN),
peroxybenzoyl nitrate (PBzN), and other homologues (e.g., peroxybutyrl nitrate)
s
are readily separated from components such as air, water, and other atmospheric
compounds, as well as ethyl nitrate, methyl nitrate, and other contaminants
that are present in synthetic mixtures. Electron-capture detection provides
sensitivities in the ppb and sub-ppb ranges. Table 5-16 shows parameters used
by several investigators to determine trace levels of PAN by GC-ECD. Sample
injection into the GC is accomplished by means of a gas-sampling valve with a
gas-sampling loop of a few milliliters' volume. Sample injection may be per-
formed manually or automatically. Typically, manual air samples are collected
in 50-200 ml ungreased glass syringes and purged through the gas-sampling valve.
Samples collected from the atmosphere should be analyzed as soon as possible
because PAN and its homologues undergo thermal decomposition in the gas phase
and at the surface of containers.
The recent work of Singh and Sal as (1983a, 1983b) on the measurement of PAN
in the free (unpolluted) troposphere (see chapter 6) is illustrative of current
capabilities for measuring low concentrations. A minimum detection limit of
0.010 ppb was obtained. The literature contains conflicting reports on the
effects of variable relative humidity on PAN measurements by GC-ECD. Some
investigators have reported a reduced response to PAN calibration samples when
dry diluent gas is used, whereas others have not observed this effect. The
reduced response has been attributed to losses of PAN to surfaces within the
inlet system and the GC. Presumably, water vapor may deactivate surfaces. For
the present, if a moisture effect is suspected in a PAN analysis, the bulk
of this evidence suggests that humidification of PAN calibration samples (to a
range approximating the humidity of the samples being analyzed) would be advisa-
ble.
Conventional long-path infrared spectroscopy and Fourier-transform infrared
spectroscopy (FTIR) have been used to detect and measure atmospheric PAN. Sen-
sitivity is enhanced by the use of FTIR. The most frequently used IR bands
have been assigned and the absorptivities shown in Table 5-17 permit the quan-
titative analysis of PAN without calibration standards. The absorptivity of the
019QQ/A 5-104 6/19/84
-------
TABLE 5-16. SUMMARY OF PARAMETERS USED IN DETERMINATION OF PAN BY GC-ECD
Reference
Heuss and Glasson,
1968
Grosjean, 1983
Darley et al . ,
1963
Stephens and Price,
1973
cn
CJ1
Lonneman et al . ,
1976
Holdren and Spicer,
1984
Peake and Sandhu,
1983
Singh and Sal as,
1983
Grosjean et al . ,
1984
Nielsen et al. ,
1982
Column
dimensions
and material
4 ft x 1/8 in
Glass
6 ft x 1/8 in
Teflon
3 ft x 1/9 in
Glass
1.5 ft. x 1/8 in
Teflon
3 to 4 ft x 1/8 in
Glass
5 ft x 1/8 in
Teflon
3.3 ft x 1/8 in
Glass
1.2 ft x 1/8 in
Teflon
1.7 ft x 1/8 in
Teflon
3.9 ft x 1/12 in
Glass
Stationary
phase
SE30
(3.8%)
10% Carbowax 400
5% Carbowax 400
5% Carbowax E 400
10% Carbowax 600
60/80 Mesh
Carbowax 600
5% Carbowax 600
10% Carbowax 600
10% Carbowax 400
5% Carbowax 400
Solid
support
80-100
Mesh
Diatoporte S
60/80 Mesh
Chromosorb P
100-200 Mesh
Chromosorb W
Chromosorb
G 80/100
mesh treated
with dimethyl
dichlorosilane
Gas Chrom Z
Gas Chrom Z
Chromosorb W
80/100 Mesh
Supelcoport
60/80 Mesh
Chromosorb G
Chromosorb
W - AW - DCMS
Column
temperature,
°C
25
30
35
25
25
35
33
33
30
25
Flow
rate, Carrier
ml/min gas
N.A.a N.A.a
40 NU
25 N2
60 Na
70 95% Ar
5% CH4
70 90% Ar
10% CH4
50 N2
30 95% AR
5% CH4
40 N2
40 N2
Elution
time,
min
N.A.a
N.A.a
2.17
1.75
2.7
3.00
N.A.3
N.A.a
5.0
6.0
Concentration
range
ppb
range
ppb
range
3 to 5
ppb
37
ppb
0.1 to 100 ppb
ppb
0.2 to 20
ppb
0.02-0.10
ppb
2-400
ppb
11 ppb
N.A. - not available in reference.
-------
TABLE 5-17. INFRARED ABSORPTIVITIES OF PEROXYACETYL NITRATE
o
CTl
Absorptivity, ppm-1
Frequency
cm-1
1842
1741
1302
1162.5
930
791.5
Mode
v(c=o)
vas(N02)
vs(N02)
v(c-o)
v(o-o)
6(N02)
Liquid PAN
(Bruckmann and
Willner, 1983)
12.4
32.6
13.6
15.8
N.A.a
13.4
PAN in air
(Stephens,
1969)
10.0
23.6
11.2
14.3
N.A.a
10.1
Frequency
cm-1
1830
1728
1294
1153
930
787
m-1 x 104
PAN in air
(Stephens and
Price, 1973)
10.0
23.6
11.4
13.9
1.8
10.3
PAN in octane
(Holdren and
Spicer, 1984)
9.44
26.5
9.44
9.66
N.A.a
10.1
Not available in reference.
-------
990 cm band of PBzN has been reported by Stephens (1969). Some recent simul-
taneous measurements of PAN and other atmospheric pollutants such as (L, HN(L,
HCOOH, and HCHO have been made by long-path FTIR spectrometry during smog epi-
sodes in the Los Angeles Basin. Tuazon et al. (1978) describes an FTIR system
operable at pathlengths up to 2 km for ambient measurements of PAN and other
trace constituents. This system employed an eight-mirror multiple reflection
cell with a 22.5-m base path. Detection of PAN was by bands at 793 and 1162 cm" .
—~i «.~i
The 793 cm band is characteristic of peroxynitrates, while the 1162 cm band
is reportedly attributable to PAN only (Hanst et al., 1982). Tuazon et al. (1981)
reported a detection limit for PAN of 3 ppb at a 900-meter pathlength.
Hanst et al. (1982) made measurements with a 1260-m folded optical path
system during a 2-day smog episode in Los Angeles in 1980. An upper limit of
1 ppb of peroxybenzoyl nitrate (PbzN) was placed, based on observations in the
vicinity of the PBzN band at 990 cm ; the maximum PAN concentration observed
was 15 ppb.
Sampling may constitute one of the major problems in the analysis of trace
reactive species, such as PAN, by long-path FTIR spectrometry. The large inter-
nal surface area of the White cells may serve to promote the decomposition or
irreversible adsorption of reactive trace species. High volume sampling rates
and inert internal surface materials are used to minimize these effects.
Because of the thermal instability of dilute PAN samples and the explosive
nature of purified PAN, calibration samples are not commercially available.
Each laboratory involved in making such measurements must prepare its own
standards. Calibration samples are usually prepared by various means at con-
centrations in the ppm range, and they must be analyzed by some absolute tech-
nique.
Earlier methods used to synthesize PAN have been summarized by Stephens
(1969). The photolysis of alkyl nitrites in oxygen was the most commonly used
procedure and may still be used by some investigators. As described by Stephens
et al. (1965), the liquefied crude mixture obtained at the outlet of the photol-
ysis chamber is purified by preparation-scale GC. [CAUTION: Both the liquid
crude mixture and the purified PAN samples are violently explosive and should
be handled behind explosion shields using plastic full-face protection, gloves,
and a leather coat at all times.] The pure PAN is usually diluted to about
1000 ppm in nitrogen cylinders at 100 psig and stored at reduced temperatures,
019QQ/A 5-107 6/19/84
-------
Gay et al. (1976) have used the photolysis of Cl2:aldehyde:NO„ mixtures
in air or oxygen for the preparation of PAN and a number of its homologues at
high yields. This procedure has been utilized in a portable PAN generator
that can be used for the calibration of GC-ECD instruments in the field (Grosjean,
1983; Grosjean et al., 1984).
The other technique for PAN preparation in current use involves the nitra-
tion of peracetic acid. Several investigators have recently reported on a con-
densed-phase synthesis of PAN with peracetic acid that produces high yields of
a pure product free of other alky! nitrates (Hendry and Kenley, 1977; Kravetz
et al., 1980; Nielsen et al., 1982; Holdren and Spicer, 1984). Most of these
procedures call for the addition of peracetic acid (40 percent in acetic acid)
to a hydrocarbon solvent (pentane, heptane, octane) maintained at 0°C in a dry-
ice acetone bath, followed by acidification with sulfuric acid and slow addition
of sodium nitrate. After the nitration is complete, the hydrocarbon fraction
containing PAN concentrations of 2 to 4 mg/ml can be stored at -20°C for
periods longer than a year (Holdren and Spicer, 1984).
The most direct method for absolute analysis of these PAN samples is by
infrared absorption, using absorptivities given in Table 5-17. Conventional IR
instruments and 10-cm gas cells can analyze gas standards of concentrations
>35 ppm. Liquid microcells can be used for the analysis of PAN in octane
solutions.
The alkaline hydrolysis of PAN to acetate ion and nitrite ion in quantita-
tive yield (Nicksic et al., 1967) provides a means independent of infrared for
the quantitative analysis of PAN. Following hydrolysis, nitrite ion may be
quantitatively analyzed by the Saltzman colorimetric procedure (Stephens, 1969).
The favored procedures now use ion chromatography to analyze for nitrite (Nielsen
et al., 1982) or acetate (Grosjean, 1983, 1984) ions. Another calibration pro-
cedure has been proposed that is based on the thermal decomposition of PAN in the
presence of excess NO (Lonneman et al., 1982; Lonneman and Bufalini, 1982). The
peroxyradical, CH3C(0)02> and its decomposition products rapidly oxidize NO to
N0? with a stoichiometry that has been experimentally measured. By the use of
NO standard mixtures and the measurement by chemiluminescence of the NO con-
sumed, the absolute PAN concentration can be determined.
5.6.4.6 Methods for Sampling and Analysis of Hydrogen Peroxide. Hydrogen perox-
ide (HpOp), like ozone and PAN, is formed as a product of the photooxidation of
hydrocarbons and reaches maximum concentrations during daylight hours. There
019QQ/A 5-108 6/19/84
-------
are some early reports of H?Q- concentrations as high as 180 ppb at an ozone
maximum of 650 ppb, but it now appears more likely that maximum HpO» concentra-
tions are in the range of 10 to 50 ppb and are only a small fraction (< 10%)
of the corresponding ozone maximum. Applied and potentially useful techniques
for the measurement of ambient H?0? are summarized in Table 5-18.
With the exception of Fourier-transform infrared (FTIR) studies, all of
the techniques that have been used for atmospheric H?0? measurements have em-
ployed aqueous traps for sampling. Recent studies have indicated that this
approach leads to interference from ozone, which will always be present at
higher concentrations. Absorbed 0- has been observed to promote both the for-
mation and destruction of FLO^ in aqueous media (Zika and Saltzman, 1982).
Therefore, an obvious research need in H_0? measurements is a clear delinea-
tion of the nature of any 0_ interferences and the development of means for
their prevention.
Of the procedures given in Table 5-18, only the titanium colorimetric,
enzyme-catalyzed, and FTIR methods have been used for atmospheric sampling.
The other procedures do not appear promising for ambient air analysis. The
titanium sulfate-8-quinolinol reagent has been used in several earlier studies
on atmospheric \\^^ (Bufalini et al., 1972; Gay et al. , 1972a; Gay et al. , 1972b;
Kok et al., 1978a). Hydrogen peroxide in air is scrubbed in a coarse-fritted
bubbler containing aqueous titanium sulfate-ammonium sulfate-sulfuric acid solu-
tion. After sampling, the solution is extracted with an aliquot of 8-quinolinol
in chloroform. The absorbance at 450 nm of the titanium (IV)-H_02-9-quinolinol
complex in chloroform is determined. A positive interference is expected from
any compound that can liberate H?0» via acid hydrolysis (Pobiner, 1961), e.g.,
t-butylhydroperoxide. Of the major atmospheric pollutants investigated (S0?,
0_, N0?, NO, and hydrocarbons), only S0? at high concentrations gave a small
(0.7 percent) negative interference (Gay et al., 1972b).
In the titanium tetrachloride method, samples are collected in a midget
impinger containing an aqueous TiCl.-HCl solution. A stable TiCl.-HpO- complex
is formed immediately, and after the solution is diluted to a known volume, the
absorbance of the complex at 410 nm is determined. The principal difficulty
with this method is the formation of fine particles, presumably TiO?, which scat-
ter visible radiation and create an apparent absorption. In an intercomparison
of H_0_ measurement methods, Kok et al. (1978a) reported rather poor agreement
between the two titanium reagents and between these and chemiluminescence.
019QQ/A 5-109 6/19/84
-------
TABLE 5-18. MEASUREMENT METHODS FOR HYDROGEN PEROXIDE
en
i
i— •
t-*
o
Method
Titanium
colorimetry
Chemi 1 umi nescence
Enzyme-catalyzed
Enzyme-catalyzed
Enzyme- catalyzed
Fourier-transform
infrared absorption
Electrochemical
H202-olefin
reactions
Mixed-ligand
complex reagents
Reagent(s)
(1) Titanium Sulfate
-8-Quinolinol
(2) Titanium Tetrachloride
Luminol, Cu(II)
basic solution
Scopoletin, horseradish-
peroxidase (HRP)
Leuco crystal violet,
HRP
3-(p-hydroxyphenyl)
propionic acid
None
Aqueous solutions
l,2-di-(4-pyridyl)
ethyl ene
Vanadium and
uranium hydroxamic
acid chelates
Limits of
detection3
(1) 1.6 x 10-6 M
(2) ca 10-6 M
0.001 to 1 ppm
1.5 x 10-11 M
10-8 M
10-6 to 10-4 Mf
0.005 ppm (est. )
5 x 10-6 to 1 M
10-6 to 5 x 10-4 M
10-6 M
Interferences"
Positive
Al kyl hydro-
peroxides
PANd
NA
NA
NA
NA9
NA
03
NA
Negative
S02C?
S02e
NA
NA
NA
None
NA
NA
NA
Applications Primary reference
Atmospheric (1) Gay et al. (1972a, 1972b)
(2) Pilz and Johann (1974);
Kok et al. (1978a)
Atmospheric, Armstrong and Humphreys (1965);
rainwater Kok et al . (1978a,1978b)
Atmospheric, Andreae (1966); Perschke
rainwater and Broda (1961); Zika and
Saltzman (1982)
Motto! a et al . (1970)
Zaitsu and Ohkura (1980)
Atmospherich Hanst et al . (1982)
— Pisarevskii and Polozova (1980)
Hauser and Kolar (1968)
Csanyi (1981);
Meloan (1961)
aExcept where noted, detection limits are in moles/1iter(M) in aqueous solution.
b03 may be both a positive and negative interference in all these procedures using aqueous sampling. See Text. NA = not available.
cThe S02 interferences is reported to be small at high S02 concentrations (Gay et al., 1972b). Studies of potential positive and
negative interferences are incomplete for these methods.
dThe reported PAN interference is small (Kok et al., 1978b).
eThe report of an S02 interference is undocumented.
fThe lower limit could presumably be reduced by the use of larger samples.
9With sufficient resolution, there should be no interferences. IR absorption by atmospheric water vapor is the major analytical
limitation.
hH202 bands have not been observed in any long-path FTIR studies. The estimated lower limit of detection in these studies is
approximately 0.005 ppm.
-------
A promising method for the measurement of hydrogen peroxide in the atmos-
phere at very low concentrations is based on the chemiluminescence obtained from
the Cu(II)-catalyzed oxidation of luminol (5-amino-2,3 dihydro-l,4-phthalazine-
dione) by H-O™ (Armstrong and Humphreys, 1965). The product of the reaction with
Hj,CL is 3-amino-phthalic acid, a nitrogen molecule, and a photon of light at
450 nm. A small positive interference was reported for PAN (Kok et a!., 1978b).
If 03 absorption leads to the formation of H20_ as reported (Zika, 1982; Heikes
et a'L, 1982), then 0- is a major interference. There have also been undocumented
reports of a negative interference from SO,,.
Perhaps the most promising chemical approach for the measurement of trace
concentrations of H?0,, employs the catalytic acitivity of the enzyme, horse-
radish peroxidase (HRP), on the oxidation of organic substrates by H»0_. This
general method involves three components: a substrate that is oxidizable, HRP,
and H202. Three substrates that have been used are scopoletin (6-methoxy-7-
hydroxy-l,2-benzopyrone), 3-(p-hydroxyphenyl)propionic acid (HPPA), and leuco
crystal violet (LCV). The scopoletin reagent has recently been used in atmos-
pheric analysis. The disappearance of scopoletin fluorescence, upon oxidation
of scopoletin by H?0_, is monitored and the fluorescence intensity is used to
obtain the concentration of H000 from a calibration curve. The most signifi-
-11
cant advantage of the scopoletin method is the sensitivity (ca. 16 M). The
chief disadvantage of the method is that the concentration of H^O,, must be
within a narrow concentration range in order to obtain an accurately measurable
decrease in fluorescence. This limits the usefulness of the technique in
determining unknown H?0_ concentrations over several orders of magnitude. With
the leuco crystal violet (LCV) substrate, intensely colored crystal violet is
formed from the reaction of H?0? with LCV, catalyzed by HRP. The absorbance is
measured at 596 nm, where the molar absorption coefficient of crystal violet is
10 M cm , a very high value and an inherent advantage of this method.
Finally, Zaitsu and Ohkura (1980) have tested a number of 4-hydroxy phenyl com-
pounds and found that 3-(p-hydroxyphenyl) propionic acid (HPPA) provided a sen-
sitive and rapid means for determining HJ)-, A product is formed that fluo-
resces at 404 nm, and the intensity of this fluorescence is monitored as a func-
tion of H~0? concentration. The detection limit was reported to be 10 mole
H-O- with a test solution of only 0.1 ml volume used. The molar sensitivity
could presumably be improved by the use of large sample volumes.
019QQ/A 5-111 6/19/84
-------
The enzymatic methods appear to be the most promising colorimetric methods
of H?0? and have considerably greater sensitivity than the methods employing
titanium reagents. However, studies of potential atmospheric interferences
have apparently not been conducted for any of these three substrates.
Hydrogen peroxide can be monitored directly in the gas phase by FTIR absorp-
tion at 1250 cm , where the absorption coefficient is 8.4 cm atm at 1 cm
resolution (Hanst et a!., 1981). One FTIR measurement of the possible presence
of 0.070 ppm H_0_ was reported during an intense smog episode in Pasadena,
California (Hanst et a!., 1975). Unfortunately, minimum detection limits at
1 km pathlength are degraded to 0.040 ppm because of neighboring absorption
bands of H20 and CH4 (Hanst et al., 1981).
As with 0~, H?02 calibration standards are not commercially available and
are usually prepared at the time of use. The most convenient method for pre-
paring aqueous samples containing micromolar concentrations of H_0p is simply
the serial dilution of commercial grade 30 percent HJ)2 (Fisher Analytical
Reagent). Techniques for the convenient generation of gas-phase standards are
not available. A technique often used for generating ppm concentrations of
H»0» in air involves the injection of microliter quantities of 30 percent H^O^
in air involves the injection of microliter quantities of 30 percent H-O^ solu-
tion into a metered stream of air that flows into a Teflon bag. Aqueous and
gas-phase samples are then standardized by conventional iodometric procedures
(Allen et al., 1952; Cohen et al., 1967).
Hydrogen peroxide liberates iodine from an iodide solution quite slowly,
but in the presence of a molybdate catalyst the reaction is rapid. The iodine
liberated can be determined by titration with standard thiosulfate at higher
concentrations or by photometric measurement of the tri-iodide ion at low con-
centrations. The molar absorption coefficient of the tri-iodide ion at 350 nm
has been determined to be 2.44 x 10 (Armstrong and Humphreys, 1965). The
stoichiometry is apparently 1 mole of iodine released per mole of H202- How-
ever, definitive studies of the stoichiometry have not been performed for H202
as they have for the iodometric determination of O^.
019QQ/A 5-112 6/19/84
-------
5.7 REFERENCES FOR CHAPTER 5
Adamson, P.; Guenthard, H. H. (1980) Reinvestigation of the matrix IR spectrum
of peroxyacetyl nitrate. Spectrochim. Acta Part A 36A: 473-475.
Aikin, A. C.; Herman, J. R.; Maier, E. 0. R.; McQuillan, C. J. (1983) Influence
of peroxyacetyl nitrate (PAN) on odd nitrogen in the troposphere and
lower stratosphere. Planet. Space Sci. 31: 1075-1082.
Akimoto, H.; Bandow, H.; Sakamaki, F.; Inoue, G.; Hoshino, M.; Okuda, M.
(1980) Photooxidation of propylene-NO -air system studies by long-path
Fourier transform infrared spectromet^y. Environ. Sci. Techno!. 14:
172-179.
Allen, A. 0.; Hochanadel, C. J.; Ghormley, J. A.; Davis, T. W. (1952) Decomposi-
tion of water and aqueous solutions under mixed fast-neutron and y-radia-
tion. J. Phys. Chem. 56: 575-586.
Anbar, M.; Taube, H. (1954) Interaction of nitrous acid with hydrogen peroxide
and with water. J. Am. Chem. Soc. 76: 6243-6247.
Andreae, W. A. (1955) A sensitive method for the estimation of hydrogen peroxide
in biological materials. Nature (London) 175: 859-860.
Andrews, A. W.; Thibault, L. H.; Lijnsky, W. (1978) The relationship between
mutagenicity and carcinogenicity of some nitrosamines. Mutat. Res. 51:
319-326.
Appel, B. R. (1973) A new and more sensitive procedure for analysis of peroxy-
benzoyl nitrate. J. Air Pollut. Control Assoc. 23: 1042-1044.
Appel, B. R.; Tokiwa, Y.; Haik, M. (1981) Sampling of nitrates in ambient air.
Atmos. Environ. 15: 283-289.
Armstrong, W. A.; Humphreys, W. G. (1965) A L.E.T. independent dosimeter based
on the chemiluminescent determination of H^Op- Can. J. Chem. 43: 2576.
Atkinson, R.; Darnall, K. R.; Pitts, J. N., Jr. (1978) Rate constants for the
reactions of OH radicals and ozone with cresols at 300 ± 1 K. J. Phys.
Chem. 82: 2759-2761.
Atkinson, R.; Darnell, K. R.; Lloyd, A. C.; Winer, A. M.; Pitts, J. N., Jr.
(1979) Kinetics and mechanisms of the reactions of the hydroxyl radical
with organic compounds in the gas phase. Adv. Photochem. 11: 375-488.
Atkinson, R.; Aschmann, S. M.; Winer, A. M.; Pitts, J. N., Jr. (1981) Rate
constants for the gas-phase reactions of 0~ with a series of carbonyls at
296 K. Int. J. Chem. Kinet. 13: 1133-11427
Atkinson, R.; Aschmann, S. M.; Fitz, D. R.; Winer, A. M.; Pitts, J. N., Jr.
(1982) Rate constants for the gas-phase reactions of 0- with selected
organics at 296 K. Int. J. Chem. Kinet. 14: 13-18.
OZONER/A 5-113 6/20/84
-------
Atkinson, R.; Aschmann, S. M. (1984) Rate constants for the reactions of CL
and OH radicals with a series of alkynes. Int. J. Chem. Kinet. 16:
259-268.
Atkinson, R.; Carter, W. P. L. (1984) Kinetics and mechanisms of the reactions
of ozone with organic compounds in the gas phase, submitted to Chem. Rev.
Atkinson, R.; Lloyd, A. C. (1984) Evaluation of kinetic and mechanistic data
for modeling of photochemical smog. J. Phys. Chem. Ref. Data (in press).
Atkinson, R.; Plum, C. N.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N. , Jr.
(1984a) Rate constants for the gas phase reactions of N0_ radicals with a
series of organics in air at 298 ± 1 K. J. Phys. Chem. (in press).
Atkinson, R.; Pitts, J. N., Jr.; Aschmann, S. M. (1984b) Tropospheric reactions
of dimethyl sulfide with NO- and OH radicals. J. Phys. Chem. (in press).
Atkinson, R.; Aschmann, S. M.; Winer, A. M.; Pitts, J. N., Jr. (1984c) Kinetics
of the gas phase reactions of NO., radicals with a series of dialkenes,
cycloalkenes and monoterpenes at 295 ± 1 K. Environ. Sci. Techno!, (in
press).
Atkinson, R.; Carter, W. P. L.; Plum, C. N.; Winer, A. M.; Pitts, J. N., Jr.
(1984d) Kinetics of the gas phase reactions of NO. radicals with a series
of aromatics at 296 ± 2 K. Int. J. Chem. Kinet. (in press).
Babko, A. K. ; Volkova, A. I. (1951) The colored complex of titanium with
hydrogen peroxide. Zh. Obshch. Khim. 21: 1949-1956.
Bailey, P. A. (1978) Ozonation in Organic Chemistry. Vol. I. Olefinic Com-
pounds. In: Blomquist, A. T.; Wasserman, H. H. , eds. Organic Chemistry:
Vol. 39-1. New York, NY: Academic Press.
Bailey, P. A. (1982) Ozonation in Organic Chemistry. Vol. II. Nonolefinic
Compounds. In: Blomquist, A. T.; Wasserman, H. H., eds. Organic Chem-
istry: Vol. 39-11. New York, NY: Academic Press.
Ballard, L. F.; Tommerdahl, J. B. ; Decker, C. E. ; Royal, T. M.; Nifong, D. R.
(1971a) Evaluation of new air pollution monitoring systems: The Los
Angeles Study. Interim Report, Contract No. CPA 70-101, U.S. Environmental
Protection Agency, Research Triangle Park, NC.
Ballard, L. F.; Tommerdahl, J. B. ; Decker, C. E. ; Royal, T. M.; Matus, L. K.
(1971b) Field evaluation of new air pollution monitoring systems: St.
Louis Study, Phase 1, Contract No. CPA 70-101, U.S. Environmental Protec-
tion Agency, Research Triangle Park, NC.
Bass, A. M.; Ledford, A. E., Jr.; Whittaker, J. K. (1977) Ultraviolet photometer
for ozone calibration. In: Dimitriades, B., ed. International Conference
on Photochemical Oxidant Pollution and Its Control; January. Volume I.
Research Triangle Park, NC: U.S. Environmental Protection Agency; pp. 13-17;
EPA report no. EPA-600/3-77-001a.
OZONER/A 5-114 6/20/84
-------
Beard, M. E.; Margeson, J. H.; Ellis, E. C. (1977) Evaluation of 1 Percent Neu-
tral Buffered Potassium Iodide Procedure for Calibration of Ozone Monitors.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-600/4-77-005.
Bergshoeff, G. (1970) Ozone Determination in Air. Pittsburgh Conf. Analytical
Chemistry and Applied Spectroscopy. Publication No. 342, IG-TNO Delft,
The Netherlands.
Bersis, D.; Vassiliou, E. (1966) A chemiluminescence method for determining
ozone. Analyst 91: 499-505.
Bowman, L. D.; Horak, R. F. (1972) A continuous ultraviolet absorption ozone
photometer. Anal. Instrum. 10: 103-108.
Bowman, R. L.; Alexander, N. (1966) Ozone-induced chemiluminescence of organic
compounds. Science (Washington, DC) 154: 1454-1456.
Brewer, A. W.; Milford, J. R. (1960) The Oxford Kew ozone sonde. Proc. R. Soc.
London Ser. A256: 470-495.
Brock, J. R.; Durham, J. L. (1984) In S02 NO and N02 Oxidation Mechanisms:
Atmospheric Considerations. Calvert, J. G. (ea)., Ann Arbor, MI; Ann
Arbor Science Publishers. Also in: Acid Precipitation Series, vol. 3.
Boston, Butterworth Publishers. Chapter 5.
Bruckmann, P. W.; Winner, H. (1983) Infrared spectroscopic study of peroxy-
acetyl nitrate (PAN) and its decomposition products. Environ. Sci.
Techno!. 17: 352-357.
Bryan, R. J.; Romanovsky, J. C. (1956) Instrumentation for air pollution.
Instruments and Automation 29.
Bufalini, J. J.; Brubaker, K. L. (1969) Symposium on Chemical Reactions in Urban
Atmospheres; October. Warren, MI: General Motors Research Laboratories.
Bufalini, J. J.; Gay, B. W., Jr; Brubaker, K. L. (1972) Hydrogen peroxide
formation from formaldehyde photooxidation and its presence in urban
atmospheres. Environ. Sci. Technol. 6: 816-821.
Burdo, T. G.; Seitz, W. R. (1975) Mechanism of cobalt catalysis of luminol
chemiluminescence. Anal. Chem. 47: 1639-1643.
Burton, C. S.; Bekowies, P. J.; Pollack, R. I.; Connell, P. (1976) Oxidant/
ozone ambient measurement methods: an assessment and evaluation.
EF76-111R, Systems Applications, Inc., San Rafael, CA. August.
California Air Resources Board. (1976) A study of the effect of atmospheric
humidity on analytical oxidant measurement methods. Presented at the 15th
Conference on Methods in Air Pollution Studies, J. J. Wesolowski, Chairman,
California Air Resources Board, Air and Industrial Hygiene Laboratory,
Berkeley, CA.
California Air Resources Board. (1978) California Air Quality Data: July-August-
September. 10(13): 2-3. Available from Technical Service Div., California
Air Resources Board, Sacramento, CA.
OZONER/A 5-115 6/20/84
-------
Calvert, J. G.; Su, F.; Bottenheim, J. W.; Strausz, 0. P. (1978) Mechanism of
the homogeneous oxidation of sulfur dioxide in the troposphere. Atmos.
Environ. 12: 197-226.
Calvert, J. G.; Stockwell, W. R. (1983) Acid generation in the troposphere by
gas-phase chemistry. Environ. Sci. Techno!. 17: 428A-443A.
Calvert, J. G., ed. (1984) SO-, NO and NO- Oxidation Mechamisms: Atmospheric Con-
siderations. Ann Arbor science Publishers, Acid Precipitation Series -
Vol. 3, Butterworth Publishers, Boston.
Carroll, H.; Dimitriades, B.; Ray, R. M. (1972) A chemiluminescence detector
for ozone measurement. Report of Investigation. RI 7650, U.S. Department
of the Interior, Bureau of Mines, Washington, DC.
Carter, W. P. L.; Winer, A. M,; Pitts, J. N., Jr. (1981a) Major atmospheric
sink for phenol and the cresols. Reaction with the nitrate radical.
Environ. Sci. Techno!. 15: 829-830.
Carter, W. P. L.; Tuazon, E. C.; Winer, A. M.; Pitts, J. N., Jr. (1981b) Gas
phase reactions of N,N~dimethylhydrazine with ozone and NO in simulated
atmospheres. In: Scanlan, R. A.; Tannenbaum, S. R., eds. N-Nitroso Com-
pounds. Washington, DC: American Chemical Society; pp. 118-131; American
Chemical Society Symposium Series no. 174.
Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr. (1981c) Effect of peroxyacetyl
nitrate on the initiation of photochemical smog. Environ. Sci. Techno!.
15: 831-834.
Chameides, W. L.; Davis, D. D. (1982) The free radical chemistry of cloud
droplets and its impact upon the composition of rain. JGR J. Geophys.
Res., 87: 4863-4877.
Cherniack, I.; Bryan, R. J. (1965) A comparison study of various types of
ozone and oxidant detectors which are used for atmospheric air sampling.
J. Air Pollut. Control Assoc. 15: 351-354.
Chock, D. P.; Kumar, S. ; Herrmann, R. W. (1982) An analysis of trends in
oxidant air quality in the South Coast Air Basin of California. Atmos.
Environ. 16: 2615-2624.
Cholak, J.; Schafer, L. J.; Yeager, P. W. (1956) The concentration of ozone in
certain American cities. J. Air Pollut. Control Assoc. 5: 227-232.
Clark, T. A.; Baumgardner, R. E.; Stevens, R. K.; Krost, K. J. (1974) Evaluation
of new ozone monitoring instruments by measuring in non-urban atmospheres.
In: Instrumentation for Monitoring Air Quality. Philadelphia, PA: American
Society for Testing and Materials; pp. 101-111.
Clements, J. B. (1975) Summary report: Workshop on ozone measurement by
potassium iodide method. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-650/4-75-007.
Clyne, M. A. A.; Coxon, J. A. (1968) Kinetic studies of oxyhalogen radical
systems. Proc. R. Soc. London Ser. A 303: 207-231.
OZONER/A 5-116 6/20/84
-------
Cohen, I. R.; Purcell, T. C. (1967) Spectrophotometric determination of hydrogen
peroxide with 8-quinolinol. Anal. Chem. 39: 131-132.
Cohen, I. R.; Purcell, T. C.; Altshuller, A. P. (1967) Analyses of the oxidant
in photooxidation reactions. Environ. Sci. Technol. 1: 247-252.
Cox, R. A.; Burrows, J. P. (1979) Kinetics and mechanism of the disproportiona-
tion of H02 in the gas phase. J. Phys. Chem. 83: 2560-2568.
Cox, R. A.; Roffey, M. J. (1977) Thermal decomposition of peroxyacetyl nitrate
in the presence of nitric oxide. Environ. Sci. Technol. 11: 900-906.
Csanyi, L. J. (1981) Analytical applications of mixed-ligand complexes. III.
Vanadium (V)-xylenol orange reagent mixture for the Spectrophotometric
determination of traces of hydrogen peroxide. Microchem. J. 26: 10-16.
Crutzen, P. J. (1982) The global distribution of hydroxyl. In: Goldberg, E. D.,
ed. Atmospheric Chemistry. New York, NY: Springer-Verlag; pp. 313-328.
Crutzen, P. J.; Fishman, J. (1977) Average concentrations of OH in the tropos-
phere, and the budgets of CHA, CO, H9, and CFLCC1-. Geophys. Res. Lett.
4: 321-324. * * * d
Darley, E. F.; Kettner, K. A.; Stephens, E. R. (1963) Analysis of peroxyacyl
nitrates by gas chromatography with electron capture detection. Anal.
Chem. 35: 589-591.
Dasgupta, P. D. (1980a) Discussion on "The importance of atmospheric ozone and
hydrogen peroxide in oxidizing sulphur dioxide in cloud and rainwater."
Atmos. Environ. 14: 272-275.
Dasgupta, P. K. (1980b) The importance of atmospheric ozone and hydrogen
peroxide in oxidizing sulfur dioxide in cloud and rainwater—further
discussion on Penkett et a!., 1979. Atmos. Environ. 14: 620-621.
Davis, D. R.; Jensen, R. E. (1976) Low level ozone and weather systems. In:
Specialty Conference on Ozone/Oxidants--Interactions with the Total
Environment. Philadelphia, PA: Air Pollution Control Association; pp.
242-251.
Delumyea, R. G. (1974) Analytical and physical investigations of the luminol
reaction. Detroit, MI: Wayne State University. Available from: Univer-
sity Microfilms, Ann Arbor, MI; Publication No. 75-13, 309. Dissertation.
Demerjian, K. L.; Kerr, J. A.; Calvert, J. G. (1974) The mechanism of photo-
chemical smog formation. In: Pitts, J. N., Jr.; Metcalf, R. L.; Lloyd,
A. C., eds. Advances in Environmental Science and Technology: vol. 4.
New York, NY: John Wiley and Sons, Inc.; pp. 1-262-
Demore, W. B.; Romanovsky, J. C.; Feldstein, M.; Hamming, W. J.; Mueller, P.
K. (1976) Interagency comparison of iodometric methods for ozone determi-
nation. In: Calibration in Air Monitoring. Philadelphia, PA: American
Soceity for Testing and Materials; pp. 131-143; ASTM Special Technical
Publication 598.
OZONER/A 5-117 6/20/84
-------
Demore, W. B.; Patapoff, M. (1976) Comparison of ozone determinations by
ultraviolet photometry and gas-phase titration. Environ. Sci. Techno!.
10: 897-899.
Deutsch, S. (1968) Acid potassium iodide method for determining atmospheric
oxidants. J. Air Pollut. Control Assoc. 18: 78-83.
Dickinson, J. E. (1961) Air quality of Los Angeles County. Technical Progress
Report Vol. II. Los Angeles Air Pollution control District, Los Angeles,
CA.
Erickson, R. E.; Yates, L. M.; Clark, R. L.; McEwen, D. (1977) The reaction of
sulfur dioxide with ozone in water and its possible atmospheric significance.
Atmos. Environ. 11: 813-817.
Finlayson-Pitts, B. J.; Pitts, J. N., Jr. (1982) An assessment of the atmospheric
chemistry of oxides of sulfur and nitrogen: acid deposition and ozone.
Final Report to Office of Technology Assessment, Congress of the United
States. Available from: NTIS, Springfield, VA.
Finlayson, B. J.; Pitts, J. N., Jr.; Atkinson, R. (1974) Low pressure gas phase
ozone-olefin reactions. Chemiluminescence, kinetics, and mechanisms. J.
Am. Chem. Soc. 96: 5356.
Flamm, D. L. (1977) Analysis of ozone at low concentrations with boric acid
buffered KI. Environ. Sci. Techno!. 11: 978-983.
Foulger, B. E.; Simmonds, P. G. (1979) Drier for field use in the determination
of trace atmospheric gases. Anal. Chem. 51: 1089-1090.
Fried, A.; Hodgeson, J. A. (1982) Laser photoacoustic detection of nitrogen dio-
xide in the gas-phase titration of nitric oxide and ozone. Anal. Chem.
54: 278-282.
Galloway, J. N.; Likens, G. E.; Keene, W. C.; Miller, J. M. (1982) The composi-
tion of precipitation in remote areas of the world. JGR J. Geophys. Res. 82:
8771-8786.
Gay, B. W., Jr.; Bufalini, J. J. (1972a) Photochemical smog and ozone reactions.
R. F. Gould, ed., Am. Chem. Soc., Adv. in Chem. Ser. 113: 225.
Gay, B. W. , Jr.; Bufalini, J. J. (1972b) Hydrogen peroxide in the urban atmos-
phere. Environ. Lett. 3: 21-24.
Gay, B. W.; Noonan, R. C.; Bufalini, J. J.; Hanst, P. L. (1976) Photochemical
synthesis of peroxyacyl nitrates in gas phase via chlorine - aldehyde
reaction. Environ. Sci. Techno!. 10: 82-85.
Gertler, A. W.; Miller, D. F.; Lamb, D.; Katz, U. (1984) Studies of sulfur dio-
xide and nitrogen dioxide reactions in haze and cloud. In: Durham, J.L., ed.
Chemistry of Particles, Fogs and Rain. Boston, MA: Butterworth Publishers;
Acid Precipitation Series, Vol. 2.
Graedel, T. E.; Farrow, L. A.; Weber, T. A. (1976) Kinetic studies of the photo-
chemistry of the urban troposphere. Atmos. Environ. 10: 1095-1116.
OZONER/A 5-118 6/20/84
-------
Graedel, T. E.; Weschler, C. J. (1981) Chemistry within aqueous atmospheric
aerosols and raindrops. Rev. Geophys. Space Phys. 19: 505-539.
Grosjean, D. (1981) Critical evaluation and comparison of measurement methods
for nitrogenous compounds in the atmosphere. Final Report by Environmental
Research and Technology, Inc. to Coordinating Research Council, Atlanta,
GA.
Grosjean, D. (1982) Formaldehyde and other carbonyls in Los Angeles air. Environ.
Sci. Techno!. 16: 254-262.
Grosjean, D. (1983) Distribution of atmospheric nitrogenous pollutants at a
Los Angeles area smog receptor site. Environ. Sci. Techno!. 17: 13-19.
Grosjean, D.; Fung, K.; Collins, J.; Harrison, J.; Breitung, E. (1984) Portable
generator for on-site calibration of peroxyacetyl nitrate analyzers. Anal.
Chem. 56: 569-573.
Halfpenny, E.; Robinson, P. L. (1952a) Pernitrous acid. The reaction between
hydrogen peroxide and nitrous acid, and the properties of an intermediate
product. J. Chem. Soc. 928-938.
Halfpenny, E. ; Robinson, P. L. (1952b) Nitration and hydroxylation of aromatic
compounds by pernitrous acid. J. Chem. Soc. 939-946.
Hampson, R. F.; Braun, W.; Brown, R. L.; Garvin, D.; Herron, J. T.; Huie, R. E. ;
Kurylo, M. J.; Laufer, A. H.; McKinley, J. D.; Okabe, H.; Scheer, M. D.;
Tsang, W.; Stedman, D. H. (1973) Survey of photochemical and rate data
for twenty-eight reactions of interest in atmospheric chemistry. J. Phys.
Chem. Ref. Data 2: 267-311.
Hanst, P. L. ; Gay, B. W.; Spence, J. W. (1981) Atmospheric measurement of trace
pollutants. Long path Fourier-transform infrared spectroscopy. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/3-81-026.
Hanst, P. L.; Wilson, W. E.; Patterson, R. K. ; Gay, B. W., Jr.; Chaney, L. W.
(1975) A spectroscopic study of California smog. Research Triangle Park,
NC. Publication No. EPA-650/4-75-006.
Hanst, P. L.; Wong, N. W.; Bragin, J. (1982) A long-path infrared study of
Los Angeles smog. Atmos. Environ. 16: 969-981.
Harding, L. B.; Goddard, W. A., III. (1978) Mechanisms of gas-phase and liquid-
phase ozonolysis. J. Am. Chem. Soc. 100: 7180-7188.
Harrison, R. M.; Laxen, D. P. H. (1978) Sink processes for tetralkyl1ead com-
pounds in the atmosphere. Environ. Sci. Technol. 12: 1384-1392.
Hauser, T. R.; Kolar, A. M. (1968) Anal. Chem. 40: 231.
Hegg, D. A.; Hobbs, P. V. (1981) Cloud water chemistry and the production of sul-
fates in clouds. Atmos. Environ. 15: 1597-1604.
OZONER/A 5-119 6/20/84
-------
Helkes, B. G. ; lazirus, A. L. ; Kok, G. L.; Kunen, S. M. ; Gandrud, B. W. ; Gitlin,
S. N.; Sperry, P. D. (1932) Evidence for aqueous phase hydrogen peroxide
synthesis in the troposphere. JGR J. Geophys. Res. 87: 3045-3051.
Hendry, D. G.; Kenley, R. A. (1977) Generation of peroxy radicals from peroxy
nitrates (RO?NO?). Decomposition of peroxyacyl nitrates. Atmos. Environ.
99: 3198-3193.
Herron, J. T.; Martinez, R. I.; Huie, R. E. (1982) Kinetics and energetics of
the Criegee intermediate in the gas phase. I. The Criegee intermediate
in ozone-alkene reactions. Int. J. Chem. Kinet. 14: 201-224.
Hersch, P.; Deuringer, R. (1963) Galvanic monitoring of ozone in air. Anal.
Chem. 35: 897:899.
Heuss, J. M.; Glasson, W. A. (1968) Hydrocarbon reactivity and eye irritation.
Environ. Sci. Techno!. 2: 1109-1116.
Hochanadel, C. J. (1952) Effects of cobalt T-radiation on water and aqueous
solutions. J. Phys. Chem. 56: 587-594.
Hodgeson, J. A. (1972) Review of analytical methods for atmospheric oxidants
measurements. Int. J. Environ. Anal. Chem. 2: 113-132.
Hodgeson, J. A. (1976) A survey of calibration techniques for atmospheric ozone
monitors. Washington, DC: National Bureau of Standards; publication no.
NBSIR 76-1191.
Hodgeson, J. A.; Krost, K. J.; O'Keeffe, A. E.; Stevens, R. K. (1970) Chemi-
luminescent measurement of atmospheric ozone. Anal. Chem. 42: 1795-1802.
Hodgeson, J. A.; Stevens, R. K.; Martin, 8. E, (1972) A stable ozone source
applicable as a secondary standard for the calibration of atmospheric
monitors. ISA Trans. 11: 161-167.
Hodgeson, J. A.; Hughes, E. E.; Schmidt, W. P.; Bass, A. M. (1977) Methodology
for standardization of atmospheric ozone measurements. In: Dimitriades,
B. , ed. International Conference on Photochemical Oxidant Pollution and
Its Control, Vol I. Research Triangle Park, NC: U.S. Environmental Pro-
tection Agency; EPA report no. EPA-600/3-77-001a; pp. 3-12.
Hoffman, M. R.; Edwards, J. 0. (1975) Kinetics of the oxidation of sulphite by
hydrogen peroxide in acidic solution. J. Phys. Chem. 79: 2096-2098.
Hoffman, M. R.; Jacob, D. J. (1984) In: Calvert, J. G. , ed. SO*, NO and NO^
Oxidation Mechanisms: Atmospheric Considerations. Boston, MA: Butterworth
Publishers; Acid Precipitation Series, Vol. 3.
Hoigne, J.; Bader, H. (1976) The role of hydroxyl radical reactions in ozonation
processes in aqueous solutions. Water Res. 10: 377-386.
Holdren, M. W. ; Rasmussen, R. A. (1976) Moisture anomaly of peroxyacetyl nitrate
(PAN). Environ. Sci. Techno!. 10: 185-187.
OZONER/A 5-120 6/20/84
-------
Holdren, M. W.; Spicer, C. W. (1984) Field compatible calibration procedure for
peroxyacetyl nitrate. Environ. Sci. Techno!. 18: 113-116.
Holdren, M. W. ; Spicer, C. W.; Hales, J. M. (1984) Peroxyacetyl nitrate solu-
bility and decomposition rate in acidic water. Atmos. Environ. (In press).
Intersociety Committee. (1970) Manual and Methods of Air Sampling and Analysis:
Tentative Method for Continuous Monitoring of Atmospheric Oxidants with
Amperometric Instruments. Health Lab. Sci. 7: 1.
*
Japar, S. M. ; Niki, H. (1975) Gas-phase reactions of the nitrate radical with
olefins. J. Phys. Chem. 79: 1629-1632.
Jeong, K.-M.; Kaufman, F. (1982) Kinetics of the reaction of hydroxyl radical
with methane and with nine Cl- and F-substituted methanes. 1. Experimen-
tal results, comparisons and applications. J. Phys. Chem. 86: 1808-1815.
Joseph, D. W.; Spicer, C. W. (1978) Chemiluminescence method for atmospheric
monitoring of nitric acid and nitrogen oxides. Anal. Chem. 50: 1400-1403.
Kacmarek, A. J.; Solomon, I. J.; Lustig, M. (1978) Preparation and properties
of peroxyacetyl nitrate. J. Inorg. Nucl. Chem. 40: 574-576.
Kamens, R. M. ; Gery, M. W.; Jeffries, H. E.; Jackson, M.; Cole, E. I. (1982)
Ozone-isoprene reactions. Product formation and aerosol potential. Int.
J. Chem. Kinet. 14: 955-975.
Katz, M. (1976) Nitrogen compounds and oxidants. In: Stern, A. C. , ed. Air
Pollution. Vol III: Measurement, Monitoring and Surveillance of Air Pollu-
tion. New York, NY: Academic Press; pp. 259-305.
Kelly, T. J.; Stedman, D. H.; Kok, G. L. (1979) Measurements of HJD- and HNO_
in rural air. Geophys. Res. Lett. 6: 375-378.
Kilpatrick, M. L.; Herrich, C. C.; Kilpatrick, M. (1955) The decomposition of
ozone in aqueous solution. J. Am. Chem. Soc. 78: 1784-1789.
Kirk-Othmer Encyclopedia of Chemical Technology, Vol. 16, 3rd Ed. (1981) John
Wiley & Sons, Inc.; p. 701.
Kok, G. L. ; Darnall, K. R.; Winer, A. M.; Pitts, J. N., Jr.; Gay, B. W. , Jr.
(1978a) Ambient air measurements of hydrogen peroxide in the California
South Coast Air Basin. Environ. Sci. Techno!. 12: 1077-1080.
Kok, G. L.; Holler, T. P.; Lopez, M. B.; Nachtrieb, H. A.; Yuan, M. (1978b)
Chemiluminescent method for determination of hydrogen peroxide in the
ambient atmosphere. Environ. Sci. Techno!. 12: 1072-1076.
Kravetz, T. M.; Martin, S. W.; Mendenhall, G. D. (1980) Synthesis of peroxyacetyl
and peroxyaroyl nitrates. Complexion of peroxyacetyl nitrate with benzene.
Environ. Sci. Technol. 14: 1262-1264.
Kunen, S. M.; Lazrus, A. L.; Kok, G. L.; Heikes, B. G. (1983) Aqueous oxidation
of S02 by hydrogen peroxides. J. Geophys. Res. 88: 3671-3674.
OZONER/A 5-121 6/20/84
-------
Larson, T. V.; Horike, N. R.; Harrison, H. (1978) Oxidation of S02 by CL and CL
in aqueous solution: a kinetic study with significance to atmospheric rate
processes. Atmos. Environ. 12: 1597-1611.
Lee, Y.-N.; Schwartz, S. E. (1981) Reaction kinetics of nitrogen dioxide with
liquid water at low partial pressure. J, Phys. Chem. 85: 840-848.
Leighton, P. A. (1961) The photochemistry of air pollution. New York, NY:
Academic Press. 300 p.
Littman, F. E.; Benoliel, R. W. (1953) Continuous oxidant recorder. Anal. Chem.
25: 1480-1483,
Lonneman, W. A.; Bufalini, J. J.; Namie, G. R. (1982) Calibration procedure for
PAN based on its thermal decomposition in the presence of nitric oxide.
Environ. Sci. Techno!. 16: 655-660.
Lonneman, W. A.; Bufalini, J. J. ; Seila, R. L. (1976) PAN and oxidant measure-
ment in ambient atmospheres. Environ. Sci. Techno!. 10: 274-280.
Lovelock, J. E.; Maggs, R. J. ; Adlar, L. R. (1971) Gas-phase coulometry by
thermal electron attachment. Anal. Chem. 43: 1962-1965.
Ludwig, F. L.; Shelar, E. (1978) Site selection for the monitoring of photo-
chemical air pollutants. Research Triangle Park, NC; U.S. Environmental
Protection Agency; EPA report no. EPA-450/3-78-013.
Martin, R. (1984) In: Calvert, J. G. , ed. S02, NO and N0« Oxidation Mechanisms:
Atmospheric Considerations. Boston, MAT Butterwortn Publishers; Acid Pre-
cipitation Series, Vol. 3.
Martin, R. L.; Damschen, D. E. (1981) Aqueous oxidation of sulfur dioxide by
hydrogen peroxide at low pH. Atmos. Environ. 15: 1615-1621.
Mast, G. M.; Saunders, H. E. (1962) Research and development of the instru-
mentation of ozone sensing. ISA Trans. 1: 325-328.
McElroy, F. F. (1979) Transfer standards for calibration of ambient air moni-
toring analyzers for ozone: technical assistance document. Research Tri-
angle Park, NC. U.S. Environmental Protection Agency; EPA report no.
EPA-600/4-79-056.
Meloan, C. E.; Mauck, M.; Huffman, C. (1961) Spectrophotometric determination
of traces of hydrogen peroxide. Anal. Chem. 33: 104-106.
Mori, M.; Shibata, M.; Kyuno, E.; Ito, S. (1956) Reaction of hydrogen peroxide
with titanium (IV) at different pH values. Kaguku to Kogyo (Tokyo)
29: 904-907.
Mottola, H. A.; Simpson, B. E.; Gorin, G. (1970) Absorptiometric determination
of hydrogen peroxide in submicrogram amounts with leuco crystal violet and
peroxidase as catalyst. Anal. Chem. 42: 410-411.
OZONER/A 5-122 6/20/84
-------
Mudd, J. 8. (1976) The role of free radicals in toxicity of air pollutants
(peroxyacyl nitrates). In: Pryor, W. A., ed. Free Radicals in Biology.
Vol. II. New York, NY: Academic Press, Inc.; pp. 203-211.
Mudd, J. B.; Leavitt, R.; Ongun, A.; McManus, T. T. (1969) Reaction of ozone with
ami no acids and proteins. Atmos. Environ. 3: 669-682.
Mueller, F. X.; Loeb, L.; Mapes, W. H. (1973) Decomposition rates of ozone in
living areas. Environ. Sci. Techno!. 7: 342-346.
Mueller, P. K.; Tokiwa, Y. ; deVera, E. R.; Wehrmeister, W. J.; Belsky, T.;
Twiss, S. ; Inada, M. (1973) A guide for the evaluation of atmospheric ana-
lyzers. Research Triangle Park, NC: U.S. Environmental Protection Agency,
Quality Assurance and Environmental Monitoring Laboratory; EPA report no.
EPA-650/4-74-014. Available from: NTIS, Springfield, VA; PB 233048.
National Academy of Sciences, National Research Council. (1977) Ozone and other
photochemical oxidants. Washington, DC: National Academy Press; pp. 45-
125 and chapters 4 and 6.
Neal, R.; Severs, R. K.; Wenzel, L.; McKenzie, K. (1976) Simultaneous chemi-
luminescent ozone and KI oxidant measurements in Houston, Texas, 1975.
In: Speciality conference on Qzone/Oxidants—Interaction with the Total
Environment. Philadelphia, PA: Air Pollution Control Association;
pp. 180-188.
Nederbragt, G. W.; Van der Horst, A.; Van Duijn, J. (1965) Rapid ozone deter-
mination near an accelerator. Nature (London) 206: 87.
Nicksic, S. W.; Harkins, J.; Mueller, P. K. (1967) Some analyses for PAN and
studies of its structure. Atmos. Environ. 1: 11-18.
Nieboer, H.; van Ham, J. (1976) Peroxyacetyl nitrate (PAN) in relation to ozone
and some meteorological parameters at Delft in The Netherlands. Atmos.
Environ. 10: 115-120.
Nielsen, T. ; Hansen, A. M.; Thomsen, E. L. (1982) A convenient method for pre-
paration of pure standards of peroxyacetyl nitrate for atmospheric analysis.
Atmos. Environ. 16: 2447-2450.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1976) Infrared Fourier-
transform spectroscopic studies of atmospheric reactions. Paper N2, 12th
Informal conference on Photochemistry, National Bureau of Standards,
Gaithersburg, Maryland, June 28-July 1.
Niki, H.; Maker, P. D. ; Savage, C. M.; Breitenbach, L. P. (1980) An FTIR study
of the mechanism for the gas-phase reaction between H0? radicals. Chem.
Phys. Lett. 73: 43-46.
Niki, H.; Maker, P. D.; Savage, C. M. ; Breitenbach, L. P. (1983) Atmospheric
ozone-olefin reactions. Environ. Sci. Technol. 17: 312A-322A.
Okita, T.; Inugami, M. (1971) Measurement of ozone by chemiluminescent methods.
Koshu Eiseiin Kenkyu Hokoku 20: 207-211.
OZONER/A 5-123 6/20/84
-------
Ott, W. (1977) Development of criteria for siting air monitoring stations.
J. Air Pollut. Control Assoc. 27: 543-547.
Overton, J. H.; Durham, J. (1982) Acidification of rain in the presence of
S02, HpOp, 0- and HNO^. Vol. II. Energy and Environmental Chemistry.
Acid Rain. Reith, L. K. , ed. Ann Arbor, MI: Ann Arbor Science. Part
3: 245-262.
Pate, C. T.; Atkinson, R.; Pitts, J. N., Jr. (1976) The gas phase reaction of
CL with a series of aromatic hydrocarbons. J. Environ. Sci. Health
All: 1-10.
Paur, R. J.; Bass, A. M. (1983) Standard reference photometer for the assay of
ozone. Presented at the National Meeting of the American Chemical Society;
October; Washington, DC.
Paur, R. J.; McElroy, F. F. (1979) Technical assistance document for calibration
of ambient ozone monitors. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/4-79-057.
Peake, E.; Sandhu, H. S. (1982) The formation of ozone and peroxyacetyl nitrate
(PAN) in the urban atmospheres of Alberta. Presented at 65th Annual Con-
ference Chem. Inst. Canada; May 30 - June 2,1982; Toronto, Canada.
Penkett, S. A. (1983) Atmospheric chemistry: PAN in the natural atmosphere.
Nature (London) 302: 293-294.
Penkett, S. A.; Jones, M. R.; Brice, K. A.; Eggleton, A. E. J. (1979) The
importance of atmospheric ozone and hydrogen peroxide in oxidizing sulphur
dioxide in cloud and rainwater. Atmos. Environ. 13: 123-127.
Perschke, H.; Broda, E. (1961) Determination of very small amounts of hydrogen
peroxide. Nature (London) 190: 257-258.
Pilz, W.; Johann, I. (1974) Measurement of very small quantities of hydrogen
peroxide in air. Int. J. Environ. Anal. Chem. 3: 257-270.
Pisarevskii, A. M.; Polozova, I. P. (1980) Electrode for potentiometric deter-
minations of hydrogen peroxide in alkaline media. Zh. Anal. Khim. 35:
2187-2194.
Pitts, J. N., Jr.; Darnall, K. R.; Carter, W. P. L. ; Winer, A. M.; Atkinson, R.
(1979) Mechanisms of photochemical reactions in urban air. Research Tri-
angle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/3-79-110.
Pobiner, H. (1961) Determination of hydroperoxides in hydrocarbon by conversion
to hydrogen peroxide and measurement by titanium complexing. Anal. Chem.
33: 1423-1426.
Previero, A.; Bordignon, E. (1964) Modifica controllata di triptofano, metionina,
cistina, e tirosina in peptidi naturali e proteine. Gazz. Chim. Ital.
9: 630-638.
OZONER/A 5-124 6/20/84
-------
Pryor, W. A.; Dooley, M. M.; Church, D. F. (1983) Mechanisms for the reaction of
ozone with biological molecules: the source of the toxic effects of ozone.
In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., (eds). Proceedings of an
International Symposium on the Biomedical Effects of Ozone and Related
Photochemical Oxidants; March 14-18, 1982; Pinehurst, NC. Princeton, NO:
Princeton Scientific Publishers, Inc. (1982). (Advances in Modern Envi-
ronmental Toxicology, vol. 5).
Pryor, W. A.; Dooley, M. M.; Church, D. F. (1982) Mechanisms for the reaction
of ozone with biological molecules: the source of the toxic effects of
ozone. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., (eds). Proceed-
ings of International Symposium on the Biomedical Effects of Ozone and
Related Photochemical Oxidants. Princeton, NJ: Princeton Scientific
Publishers (1982). (Advances in Modern Environmental Toxicology, vol. 5).
Pryor, W. A.; Prier, E. G.; Church, D. F. (1981) Radical production from the
interaction of ozone and PUFA as demonstrated by electron spin resonance
spin trapping techniques. Environ. Res. 24: 42-52.
Pryor, W. A.; Stanley, J. P.; Blair, E. ; Cullen, G. B. (1976) Autoxidation of
polyunsaturated fatty acids. Part I. Effect of ozone on the autoxidation
of neat methyl linoleate and methyl linoleate. Environ. Res. 31: 201-210.
Purcell, T. C.; Cohen, I. R. (1967) Microdetermination of peroxides by kinetic
colorimetry. Environ. Sci. Techno!. 1: 431-433.
Rao, T. K.; Young, J. A.; Luinsky, W.; Epler, J. L. (1979) Mutagenicity of ali-
phatic nitrosamines in Salmonella typhimurium. Mutat. Res. 16: 1-7.
Regener, V. H. (1960) On a sensitive method for the recording of atmospheric
ozone. JGR J. Geophys. Res. 65: 3975-3977.
Regener, V. H. (1964) Measurement of atmospheric ozone with the chemiluminescent
method. JGR J. Geophys. Res. 69: 3795-3800.
Rehme, K. A.; Puzak, J. C.; Beard, M. E.; Smith, C. F.; Paur, R. J. (1981)
Project summary: Evaluation of ozone calibration procedure. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/S4-80-050.
Renzetti, N. A.; Romanovsky, J. C. (1956) Comparative study of oxidants in the
Los Angeles atmosphere. J. Air Pollut. Control Assoc. 6: 154-159.
Richards, L. W.; Anderson, J. A.; Blumenthal, D. L.; McDonald, J. A.; Kok, G. L.;
Lazrus, A. L. (1983) Hydrogen peroxide and sulfur (IV) in Los Angeles
cloud water. Atmos. Environ. 17: 911-914.
Roth, J. A.; Sullivan, D. E. (1981) Solubility of ozone in water. Ind. Eng.
Chem. Fundam. 20: 137-140.
Saltzman, B. E.; Wartburg, A. F., Jr. (1965) Absorption tube for removal of
interfering sulfur dioxide in analysis of atmospheric oxidant. Anal. Chem.
37: 779-782.
020NER/A 5-125 6/20/84
-------
Scheutzle, D.; Rasmussen, R. A. (1978) The molecular composition of secondary
aerosol particles formed from terpenes. J. Air Pollut. Control Assoc.
28: 236-240.
Schwartz, S. E. (1984) In SO,,, NO and NOp Oxidation Mechanisms: Atmospheric
Considerations. In: CaTvert, 0. G., ed. Vol. 3. Boston, MA: Butterworth
Publishers.
Seiler, W.; Fishman, J. (1981) The distribution of carbon monoxide and ozone
in the free troposphere. JGR J. Geophys. Res. 86: 7255-7265.
Severs, R. K. (1975) Simultaneous total oxidant and ozone measurements in
ambient air. J. Air Pollut. Control Assoc. 25: 394-396.
Shanley, E. S. (1951) Hydrogen peroxide. J. Chem. Educ. 28: 260-266.
Singh, H. B. ; Hanst, P. L. (1981) Peroxyacetyl nitrate (PAN) in the unpolluted
atmosphere: An important reservoir for nitrogen oxides. Geophys. Res.
Lett. 8: 941-944.
Singh, H. B.; Ludwig, F. L.; Johnson, W. B. (1978) Tropospheric ozone: Concen-
trations and variabilities in clean remote atmospheres. Atmos. Environ.
12: 2185-2196.
Singh, H. B.; Sal as, L. J. (1983a) Peroxyacetyl nitrate in the free troposphere.
Nature (London) 302: 326-328.
Singh, H. B.; Salass L. J. (1983b) Methodology for the analysis of peroxyacetyl
nitrate (PAN) in the unpolluted atmosphere. Atmos. Environ. 17: 1507-
1516.
Spicer, C. W.; Holdren, M. W.; Keigley, G. W. (1983) The ubiquity of peroxy-
acetyl nitrate in the continental boundary layer. Atmos. Environ. 17:
1055-1058.
Standing Air Monitoring Work Group. Air Monitoring Work Group. Air monitoring
strategy for implementation plans. U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards. Research Triangle Park, NC.
June 1977.
Steinberg, M. ; Dietz, R. N. (1969) Sr-90 ozone generator for sub-ppm concentra-
tion range—instruction manual. Prepared for the U.S. Environmental Protec-
tion Agency, Research Triangle Park, NC by Brookhaven National Laboratory,
Department of Applied Science, Upton, Long Island, NY, 26 pp.
Stephens, E. R. (1964) Absorptivities for infrared determination of peroxyacyl
nitrates. Anal. Chem. 36: 928-929.
Stephens, E. R. (1967) The formation of molecular oxygen by alkaline hydrolysis
of peroxyacetyl nitrate. Atmos. Environ. 1: 19-20.
Stephens, E. R. (1969) The formation, reactions, and properties of peroxyacyl
nitrates (PANs) in photochemical air pollution. In: Pitts, J. N.; Metcalf,
R. L., eds. Advances in environmental sciences and technology: vol. 1.
New York, NY: Wiley-Interscience; pp. 119-146.
OZONER/A 5-126 6/20/84
-------
Stephens, E. R.; Burleson, F. R.; Cardiff, E. A. (1965) The production of pure
peroxyacyl nitrates. J. Air Pollut. Control Assoc. 15: 87-89.
Stevens, R. K.; Clark, T. A.; Decker, C. E.; Ballard, L. F. (1972a) Field per-
formance of advanced monitors for oxides of nitrogen, ozone, sulfur
dioxide, carbon monoxide, methane and nonmethane hydrocarbons. Presented
at the 1972 Annual Meeting of the Air Pollution Control Association, June;
Miami, FL.
Stephens, E. R.; Darley, E. F.; Taylor, 0. C.; Scott, W. E. (1961) Photochemical
reaction products in air pollution. Int. J. Air Water Pollut. 4: 79-100.
Stephens, E. R.; Hanst, P. L.; Doerr, R. C.; Scott, W. E. (1956) Reactions of
nitrogen dioxide and organic compounds in air. Ind. Eng. Chem. 48: 1498-
1504.
Stephens, E, R.; Price, M. A. (1973) Analysis of an important air pollutant --
peroxyacetyl nitrate. J. Chem. Educ. 50: 351-354.
Stephens, E. R.; Scott, W. E.; Hanst, P. L.; Doerr, R. C. (1956) Recent develop-
ments in the organic chemistry of the atmosphere. Proc. Am. Pet. Inst.
36(111): 288-297.
Stevens, R. K. ; Hodgeson, J. A. (1973) Application of chemiluminescent reactions
to the measurement of air pollutants. Anal. Chem. 45: 443A-446A, 449A.
Stevens, R. K.; Hodgeson, J. A.; Ballard, L. F.; Decker, C. E. (1972b) Ratio
of sulfur dioxide to total gaseous sulfur compounds and ozone to total
oxidants in the Los Angeles atmosphere - An instrument evaluation study.
In: Shults, W. D., ed. Determination of Air Quality. New York, NY:
Plenum Publishing Co.; pp. 109-130.
Su, F.; Calvert, J. G.; Lindley, C. R.; Uselman, W. M.; Shaw, J. H. (1979)
A Fourier-transform infrared kinetic study and HOC1 and its absolute
integrated infrared band intensities. J. Phys. Chem. 83: 912-920.
Taube, H.; Bray, W. C. (1940) Chain reactions in aqueous solutions containing
ozone, hydrogen peroxide and acid. J. Am. Chem. Soc. 62: 3357-3373.
Tokiwa, Y. ; Twiss, S.; deVera, E. A.; Mueller, P. K. (1972) Atmospheric ozone
determination by amperometry and colorimetry. In: Mamantov, G. ; Shults,
W. D., eds. Determination of air quality. New York, NY: Plenum Pub-
lishing Co.; pp. 109-130.
Toyama, Y.; Kobayashi, J. (1966) Notes on an ozone generator and its calibra-
tor. Pap. Meteorol. Geophys. 17: 65-75.
Tuazon, E. C.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr. (1981a) Reac-
tions of hydrazines with ozone under simulated atmospheric conditions.
Environ. Sci. Techno!. 15: 823-828.
Tuazon, E. C.; Carter, W. P. L.; Brown, R. V.; Atkinson, R.; Winer, A. M.;
Pitts, J. N., Jr. (1982) Atmospheric reaction mechanisms of amine fuels.
U.S. Air Force Final Report, ESL-TR-82-17.
OZONER/A 5-127 6/20/84
-------
Tuazon, E. C. ; Graham, R. A.; Winer, A. M.; Easton, R. R. ; Pitts, J. N., Jr.
Hanst, P. L. (1978) A kilometer pathlength Fourier-transform infrared sys-
tem for the study of trace pollutants in ambient and synthetic atmospheres.
Atmos. Environ. 12: 865-875.
Tuazon, E. C.; Winer, A. M.; Graham, R. A.; Pitts, J. N., Jr. (1980) Atmos-
pheric measurement of trace pollutants by kilometer-pathlength FT-IR
spectroscopy. Adv. Environ. Sci. Techno!. 10: 259-360.
Tuazon, E. C.; Winer, A. M.; Pitts, J. N., Jr. (1981b) Trace pollutant concen-
trations in a multi-day smog episode in the California South Coast
Air Basin by long pathlength FT-IR spectroscopy. Environ. Sci.
Techncl. 15: 1232-1237.
Tuesday, C. S. (1961) The atmospheric photooxidation of trans-butene-2 and nitric
oxide. In: Chemical reactions in the lower and upper atmosphere: pro-
ceedings of an international symposium; April; San Francisco, CA. Inter-
science Publishers; pp. 15-49.
U.S. Department of Health, Education, and Welfare, National Air Pollution Control
Administration. (1970) Air Quality Criteria for Photochemical Oxidants.
Publication No. AP-63, Washington, DC; U.S. Government Printing Office.
182 pp.
U.S. Environmental Protection Agency. (1971a) National primary and secondary
air quality standards. Fed. Regist. 36: 8186-8201, April 30. (See also:
Title 40, CFR, Part 50.)
U.S. Environmental Protection Agency. (1971b) Reference Methods. Code of Federal
Regulations, Title 40, Part 50, Appendix D. Washington, DC: U.S. Government
Printing Office.
U.S. Environmental Protection Agency. (1975) Ambient air monitoring reference
and equivalent methods. Fed. Regist. 40: 7042-7070, February 18.
U.S. Environmental Protection Agency. (1976) Measurement of photochemical oxi-
dants in the atmosphere. Calibration of reference methods. Fed. Regist.
41: 44049-44057, October 6.
U.S. Environmental Protection Agency. (1977) Quality assurance handbook for
air pollution measurement systems. Research Triangle Park, NC: U.S.
Environmental Protection Agency; EPA repot no. EPA-600/4-77-027.
U.S. Environmental Protection Agency. (1974) Comparability of nine methods for
monitoring N07 in the ambient air. Washington, DC: U.S. Environmental Pro-
tection Agency^ EPA report no. EPA-650/4-74-012.
U.S. Environmental Protection Agency. (1978a) Air quality criteria for ozone
and other photochemical oxidants. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; pp. 116-135; EPA report no. EPA-600/8-78-004.
U.S. Environmental Protection Agency. (1978b) Air quality data - 1976 annual
statistics. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no. EPA-450/2-78-009.
OZONER/A 5-128 6/20/84
-------
U.S. Environmental Protection Agency. (1978c) Screening procedures for ambient
air quality data. Unpublished report. Research Triangle Park, NC.
U.S. Environmental Protection Agency. (1979a) Fed. Regist. 44(92): 27558, Thurs-
day, May 10.
U.S. Environmental Protection Agency. (1979b) Fed. Regist. 44(219): 65069, Fri-
day, November 9.
U.S. Environmental Protection Agency. (1979c) Fed. Regist. 44(242): 72589,
Friday, December 14.
U.S. Environmental Protection Agency. (1979d) Calibration of ozone reference
methods. Fed. Regist. 44: 8221-8233, February 8.
U.S. Environmental Protection Agency (1979e) Air quality data - 1978 annual
statistics. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no. EPA-450/4-79-037.
U.S. Environmental Protection Agency. (1982) Air quality criteria for oxides
of nitrogen. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no. EPA-600/8-82-026.
Van de Wiel, H. J.; Reijnders, H. F. R.; Seifert, R.; Gies, H.; Lanting, R. W.;
Rudolf, W. (1979) International comparative programme on ozone reference
methods. Atmos. Environ. 13: 555-557.
o
Vigroux, E. (1952) Spectrophotometry absorption of ozone at 18°C below 3130A.
Compt. Rend. pp. 2529-2530.
Warren, G. J.; Babcock, G. (1970) Portable ethylene chemiluminescence ozone
monitor. Rev. Sci. Instrum. 41: 280-282.
Wartburg, A. F.; Brewer, A. W.; Lodge, J. P., Jr. (1964) Evaluation of a cou-
lometric oxidant sensor. Int. J. Air Water Pollut. 8: 21-28.
Watanabe, I.; Stephens, E. R. (1978) Reexamination of moisture anomaly in analysis
of peroxyacetyl nitrate. Environ. Sci. Techno!. 12: 222-223.
Weast, R. C., ed. (1977) Handbook of Chemistry and Physics, 57th Ed. Boca Raton,
FL: CRC Press; p. D-194.
Wendschuh, P. H.; Fuhr, H.; Gaffney, J. S.; Pitts, J. N., Jr. (1973) Reaction of
peroxyacetyl nitrate with amines. Chem. Commun. pp. 74-75.
Wendschuh, P. H.; Pate, C. T.; Pitts, J. N., Jr. (1973a) The reaction of pero-
xyacetyl nitrate with aldehydes. Tetrahedron Lett. 31: 2931-2934.
West, E. S.; Todd, W. R.; Mason, H. S.; Van Bruggen, J. T. (1966) Textbook of
Biochemistry, 4th Ed. New York, NY; The Macmillan Co.
Winer, A. M.; Atkinson, R.; Pitts, J. N., Jr. (1984) Gaseous nitrate radical:
a possible nighttime atmospheric sink for biogenic organic compounds.
Science (Washington, DC) 224: 156-159.
OZONER/A 5-129 6/20/84
-------
Yamamoto, Y.; Niki, E.-, Shiokawa, H.; Kamiya, Y. (1979) Ozonation of organic
compounds. 2. Ozonation of phenol in water. J. Org. Chera. 44: 2137-2142.
Zaitsu, K.; Okhura, Y. (1980) New fluorogenic substrates for horseradish pero-
xidase: rapid and sensitive assays for hydrogen peroxide and peroxidase.
Anal. Biochem. 109: 109-113.
Zhang, J.; Hatakeyama, S.; Akimoto, H. (1983) Rate constants of the reaction of
ozone with trans-1, 2-dichloroethene and vinyl chloride in air. Int. J.
Chem. Kinet. 15: 655-668.
Zika, R. G.; Saltzman, E. S. (1982) Interaction of ozone and hydrogen peroxide
in water: implications for analysis of H?0? in air. Geophys. Res. Lett.
9: 231-234.
Zika, R.; Saltzman, E.; Chameides, W. L.; Davis, 0. D. (1982) H_02 levels in
rainwater collected in South Florida and the Bahama Islands. JGR J.
Geophys. Res. 87: 5015-5017.
Zocher, H.; Kautsky, H. (1923) Luminescence during chemical reactions. Natur-
wissenschaften 11: 194-199.
OZONER/A 5-130 6/20/84
-------
6. CONCENTRATIONS OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS IN AMBIENT AIR
6.1 INTRODUCTION
The data presented in this chapter on the concentrations of ozone and
other photochemical oxidants in ambient air are intended to support and com-
plement information presented in subsequent chapters on the effects of these
compounds. Thus, this chapter describes potential exposures of human popula-
tions, crops, ecosystems, and nonbiological materials in general terms for the
entire nation and in specific terms for selected areas of the country. Since
the health and welfare effects of ozone have been much more thoroughly document-
ed than those of other related oxidants, primary emphasis in this section has
been placed on the concentrations of ozone found in ambient air. Potential
exposures are described by presenting data on peak (or second-highest) and
average concentrations nationwide and on seasonal and diurnal patterns in
selected urban and nonurban areas. The patterns of sustained or recurring
concentrations of respective incremental levels of ozone have been examined
for selected sites in order to aid in understanding the significance of health
and welfare effects documented in subsequent chapters. Likewise, data have
been included that portray representative urban and rural concentrations by
season and by time of day. In addition, data are presented on ozone or total
oxidant concentrations that are relevant to epidemiological studies (chapter
12). Spatial variations in ozone concentrations are briefly addressed since
latitude, altitude, and indoor-outdoor gradients are pertinent to the assess-
ment of potential exposures of human populations, and except for the indoor-
outdoor gradients, of crops and ecosystems.
Ozone is the only photochemical oxidant, other than nitrogen dioxide,
that is routinely monitored and for which a comprehensive aerometric data base
exists. Data for peroxyacetyl nitrate (PAN) and its homologues and for hydro-
gen peroxide (H?0_) and formic acid (HCOOH) have all been obtained as part of
special research investigations. Consequently, no nationwide patterns are
available for these oxidants, nor are data available that would permit the
extensive examination of the occurrence of these oxidants or of the corre-
lations of levels and patterns of these oxidants with those of ozone. Sections
6.6 and 6.7 present, however, a reasonably extensive review of concentration
data for these other oxidants.
0190DL/A 6-1 6/15//84
-------
As documented In the discussions that follow, the occurrence in ambient
air of ozone at concentrations above 0.12 ppm is both widespread and persistent
in many areas of the country. High concentrations (i.e., above 0.12 ppm) may
occur not only as single 1-hour exposures but also as recurring or sustained
high levels that are present for several hours on any one day, and which,
because of the persistence of meteorological patterns, are often repeated on
consecutive days.
The concentrations of ozone and related photochemical oxidants observed
in ambient air are the net result, as shown in the preceding chapter, of a
combination of any or all of a variety of atmospheric processes, including:
1. Local photochemical production from oxides of nitrogen and
reactive volatile organic compounds.
2. Transport of ozone produced photochemically but not locally.
3. Intrusion into the troposphere, even to ground level, of ozone-
rich air from a ubiquitous stratospheric reservoir.
4. Formation of ozone photochemically in the mid-troposphere, with
subsequent intrusion into the boundary layer.
5. Chemical scavenging in the atmosphere of ozone and other oxidants;
e.g., the reaction of ozone with nitric oxide (NO) or the
reaction of H202 with sulfur dioxide (S02).
6. Physical scavenging in the atmosphere of ozone and other oxidants;
e.g., the temperature-dependent decomposition of PAN, the
precipitation scavenging of HJ)-, and the photolytic dissociation
of ozone.
7. Combined physical and chemical scavenging processes at the
earth's surface; e.g., the deposition of ozone on reactive
biological or nonbiological surfaces, such as vegetation,
soils, or certain polymers.
These processes include, obviously, both manmade and natural processes
and driving mechanisms. Although the occurrence of high ozone concentrations
is most commonly associated with recognized meteorological conditions that
involve intense sunlight and elevated temperatures, the variety of processes
that may be involved contribute to strong diurnal cycles, but peak concentra-
tions have been observed to occur at almost any time of day. Ozone may be
transported after Us formation for distances up to 1000 km or more. Likewise,
PAN and other oxidants can be transported long distances. As a result, high
0190DL/A 6-2 6/15//84
-------
concentrations of ozone and related oxidants occur not only near large sources
of precursors but also in downwind nonurban areas. Ozone, and apparently PAN,
as well, can be transported at night above the surface pollutant and nocturnal
inversion layer (chapter 4). Thus, low morning concentrations of ozone are no
measure of the potential for high concentrations later in the day because
downward mixing from the transported ozone reservoir aloft will occur. During
daylight hours, ozone can be transported considerable distances at or near
ground level.
The analysis of quantitative apportionment of observed concentrations between
manmade and natural sources is germane primarily to understanding how stringent
the control of controllable (manmade) sources must be and to formulating
control strategies for attaining promulgated standards. Consequently, the
emphasis of this chapter is on documentation of concentrations rather than on
explanations of possible causes of observed concentrations. Probable causes
and explanations are mentioned where they are pertinent to the discussion.
Most of the data presented in this chapter to characterize both nationwide
and site-specific ozone concentrations in ambient air were obtained after
1978, although some older data are cited for purposes of historical and general
comparisons. Two factors influenced the use of post-1978 data. First, the
current Federal Reference Method for ozone, chemiluminescence, and the equi-
valent UV method were almost universally employed by 1979. Second, EPA pro-
mulgated a UV calibration method for ozone in 1979 Thus, these data form a
relatively homogeneous set for purposes of intercomparison. Because of the
well-recognized difficulties in converting from older data sets to the current
reference method, the chief pre-1979 aerometric data for ozone presented in
this section are those for specific sites and specific years that provide some
background information for epidemiologic studies (chapter 12). In addition,
some historical data on trends in ozone concentrations are given to help put
more recent data in perspective.
6.2 HISTORICAL DATA ON OZONE/OXIDANT CONCENTRATIONS AND TRENDS IN AMBIENT AIR
6.2.1 Summary of Urban Oxidant Data, 1964 through 1975
Aerometric data published in the 1970 and 1978 criteria documents for
ozone and other photochemical oxidants help provide perspective on the extent
of the photochemical oxidant problem in the United States over a decade or
0190DL/A 6-3 6/15//84
-------
deleted
average oxidant concentration, the peak concentration (less than 1-hour averag-
ing time), and the number and percentage of days on which the maximum 1-hour-
average oxidant concentration exceeded the specified values. Data in this
table were obtained from one site per city (U.S. Department of Health, Educa-
tion, and Welfare, 1970). The measurements were made using potassium iodide
total oxidants methods (chapter 5). As this table indicates, in the mid-1960s
the percentage of days on which oxidant concentrations exceeded 0.15 ppm in
Los Angeles and Pasadena, California, was an order of magnitude greater than
in other cities both within California and in other locations where monitoring
was carried out. In addition, during this 1964-1967 monitoring period the
maximum hourly average observed in Pasadena, Los Angeles, and San Diego was
typically twice that recorded elsewhere.
Table 6-2 shows the range of second-highest 1-hour-average concentrations
(potassium iodide methods) in selected major cities in 1974 and 1975 (U.S.
Environmental Protection Agency, 1978). This table, although tabulated in a
different manner than Table 6-1, shows that for comparable cities there was no
apparent increase in maximum concentrations between the 1964-1967 and 1974-1975
periods. The data are a gross indication of trends, but should not be over-
interpreted, since the data may not have been obtained from the same sites or
from the same number of observations.
6.2.2 Summary of Rural and Remote Ozone Data, 1957 through 1975
The 1978 criteria document for ozone and other photochemical oxidants
(U.S. Environmental Protection Agency, 1978) pointed out clearly that the
ozone concentrations observed at nonurban sites could be the result of either
transported manmade pollutants or naturally generated ozone, or combinations
of both. Thus, it is not possible to categorize nonurban sites arbitrarily as
being indicative of the natural atmospheric background; in fact, many rural
areas can be shown to be strongly affected by upwind urban pollutant sources.
Historical data on "remote" or "rural" or "nonurban" ozone/oxidant concen-
trations have been taken from the 1970 and 1978 criteria documents and are
shown in Tables 6-3, 6-4, and 6-5. The locations in Tables 6-3 and 6-4 are a
mixture of obviously remote locations, such as Antarctica and Mauna Loa; and
rural central continental locations that may or may not be immune from urban
transport problems, such as Pocahontas County, West Virginia, and Arosa,
0190DL/A 6-4 6/15//84
-------
TABLE 6-1. SUMMARY OF MAXIMUM OXIDANT CONCENTRATIONS RECORDED
IN SELECTED CITIES, 1964-1967
CT>
Total days with maximum hourly average equal
to or greater than concentration specified
Station
Pasadena
Los Angeles
San Diego
Denver3
St. Louis
Philadelphia
Sacramento
Cincinnati
Santa Barbara
Washington, D.C.
San Francisco
Chicago
11 months of data
Source: U.S. Depa»
Total days
of available
data
728
730
623
285
582
556
711
613
723
577
647
530
beginning February
0.15 ppm
No.
days
299
220
35
14
14
13
16
10
11
7
6
0
1965.
"tment of Health, Education,
Percent
of days
41.1
30.1
5.6
4.9
2.4
2.3
2.3
1.6
1.5
1.2
0.9
0
and Wei fare
0.10 ppm
No.
days
401
354
130
51
59
60
104
55
76
65
29
24
, 1970.
Percent
of days
55.1
48.5
20.9
17.9
10.1
10.9
14.6
9.0
10.5
11.3
4.5
4.5
0.05 ppm
No.
days
546
540
540
226
362
233
443
319
510
313
185
269
Percent
of days
75.0
74.0
74.0
79.3
62.2
41.9
62.3
52.0
70.5
54.2
28.6
50.8
Maximum
hourly
average, ppm
0.46
0.58
0.58
0.25
0.35
0.21
0.26
0.26
0.25
0.21
0.18
0.13
-------
TABLE 6-2. OXIDANT CONCENTRATIONS OBSERVED IN SELECTED
URBAN AREAS OF THE UNITED STATES, 1974-1975
Urban areas
New York, NY - Northeastern NJ
Los Angeles - Long Beach, CA
Chicago, IL - Northwestern IN
Philadelphia, PA
Detroit, MI
Boston, MA
Washington, DC
Cleveland, OH
Minneapolis - St. Paul, MN
Houston - Galveston, TX
Baltimore, MB
Dallas - Fort Worth, TX
Milwaukee - Racine, WI
Seattle - Tacoma, WA
Cincinnati, OH - Northern KY
Denver, CO
Total
no. of
valid
sites
8
3
6
10
2
7
8
5
2
4
2
2
7
4
6
6
Range of second-highest
1-hr values
ug/m3
259-510
255-784
163-427
216-625
455-514
186-376
363-451
245-411
141-206
304-588
314-372
274-323
332-425
118-235
284-412
212-349
ppm
0.13-0.26
0.13-0.40
0.08-0.22
0.11-0.32
0.23-0.26
0.09-0.19
0.18-0.23
0.12-0.21
0.07-0.10
0.16-0.30
0.16-0.19
0.14-0.16
0.17-0.22
0.06-0.12
0.14-0.21
0.11-0.18
aOnly sites having a minimum of 4000 observations were included in this
summary.
Source: U.S. Environmental Protection Agency, 1978.
0190DL/A
6-6
6/15//84
-------
TABLE 6-3. SUMMARY OF OXIDANT CONCENTRATIONS IN AMBIENT AIR
AT RURAL AND REMOTE SITES, 1957 THROUGH 1967
Location
Period
Concentrations
and averaging times
Geographic
South Pole
through May 1958
1961 through
1964
1963 and 1964
Petawawa Forest,
Chalk River,
Ontario
1967
Greenknob,
North Carolina
June 15, 1962-
July 11, 1962
(0.01 to 0.034 ppm);
mean monthly avg.
at surface
40 to 80 ug/m3
(0.02 to 0.04 ppm);
monthly mean, Regener
method
20 to 60 |jg/m3
(0.01 to 0.03 ppm);
monthly mean, Mast
meter
•v-20 to -vSO ug/m3
(M).01 to -v.0.04 ppm),
15-min measurements
(n = 6865), 24 hr/day
mean of 22 ug/m3
(0.011 ppm) and
maximum 15-minute
of 120 ug/m3
(0.06 ppm)
33 ug/m3 (0.017
ppm), mean cone;
140 ug/m3 (0.07
ppm), maximum
instantaneous
measurement
(n = 2394)
Reference
Greenland
Antarctica
July 1960 25 ug/m3 (0.013 ppm); McKee
instantaneous maximum (1961)
April 1957 20 to 67 ug/m3 Odishaw
(1959)
Aldaz
(1967)
Canada,
Dept. of
Forestry
and Rural
Develop-
ment (1%7)
U.S. Dept.
of Interior,
Southeastern
Forest Exp.
Station
(1967)
0190DL/A
6-7
6/15//84
-------
TABLE 6-3. SUMMARY OF OXIDANT CONCENTRATIONS IN AMBIENT AIR
AT RURAL AND REMOTE SITES, 1957 THROUGH 1967 (continued)
Location
Period
Concentrations
and averaging times
Reference
Pocahontas County,
West Virginia
June 6, 1961-
July 6, 1961
Remote ground-level
sites
49 ug/m3
(0.025 ppm),
mean cone;
125 ug/m3
(0.064 ppm),
maximum instan-
taneous measurement
(n = 2880)
20 to 60 ug/m3
(0.01 to 0.03 ppm),
instantaneous
measurements
U.S Dept.
of Interior,
Southeastern
Forest Exp.
Station
(1967)
Junge
(1963)
Source: Tabulated from data in U.S. Department of Health, Education, and
Welfare (1970).
0190DL/A
6-8
6/15//84
-------
TABLE 6-4. CONCENTRATIONS OF TROPOSPHERIC OZONE BEFORE 1962
References
Cbtz and Volz (1951)
Regener (1957)
Regener (1957)
Ehmert (1952)
Teichert (1955)
Kay (1953)
Brewer (1955)
Rice and Pales (1959)
Wexler et al. (1960)
Location, time, and remarks
Arosa, Switzerland 1950-1957, high
valley; daily maximum values.
Mt. Capilio and Albuquerque,
New Mexico, 1951-1952.
O'Neil, Nebraska, 1953.
Weissenau, Bodensee, Germany, 1952.
Lindenberg, Obs. , Germany, 1953-1954.
Farnborough, England, 1952-1953.
Tromso, Norway 1954.
Mauna Loa Observatory, Hawaii.
Little American Station, Antarctica.
Altitude3
1860 m
3100 m
1600 m
12.5 m above ground
20 m
above ground
80 m
above ground
0-12,000 m
0-10,000 m
3000 m
100 m
Os,
Range
19-90
18-85
3-120
30-100
0-90
0-70
0-50
0-50
26-50
60-70
30-62
20-60
ug/m3
Average
50
45
36
60
36
30
30
27
36
65
45
45
oa
Range
9.7-45.9
9.2-43.4
1.5-61.2
15.3-51.0
0-45.9
0-35.7
0-25.5
0-25.5
13.3-25.5
30.6-35.7
15.3-31.6
10.2-30.6
. ppb
Average
25.5
23.0
18.4
30.6
18.4
15.3
15.3
13.8
18.4
33.2
23.0
23.0
Above mean sea level, except as noted.
As interpreted from the published data. The values sometimes represent absolute maxima, sometimes mean maxima.
Source: U.S. Environmental Protection Agency, 1978.
-------
TABLE 6-5. SUMMARY OF OZONE DATA FROM RESEARCH TRIANGLE INSTITUTE STUDIES, 1973 THROUGH 1975
01
i
Average
Station
McHenry, MD
Kane, PA
Coshocton, OH
Lewisburg, W VA
Wilmington, OH
McConnelsville, OH
Wooster, OH
McHenry, MD
DuBois, PA
Bradford, PA
Lewisburg, W VA
Creston, IA
Wolf Point, MT
De Ridder, LA
Poynette, WI
Port 0' Conner, TX
Year
1973
1973
1973
1973
1974
1974
1974
1974
1974
1975
1975
1975
1975
1975
1975
1975
Station
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
ppm
0.074
0.065
0.056
0.052
0.052
0.057
0.047
0.057
0.056
0.040
0.038
0.035
0.028
0.030
0.038
0.027
ug/m3
145
127
110
187
102
112
92
112
110
78
74
69
55
59
74
53
No. of hours
>0.08 ppm
600
639
357
249
259
262
262
262
341
100
59
17
0
38
126
99
Total
hours
1662
2131
1785
1663
1751
2011
1878
2011
1667
2332
2386
2117
2160
2994
2663
2912
% Hours
ppm > 0.08
37.0
30.0
20.0
15.0
14.9
13.0
14.0
13.0
20.5
4.3
2.5
0.8
0.0
1.3
4.7
3.4
Source: U.S. Environmental Protection Agency, 1978, modified.
-------
Switzerland. On the basis, however, of the rather extensive studies that have
been made of both nonurban and background ozone concentrations since the 1978
document, the patterns of maximum observed concentrations at these various
sites are not outside the range of those that could occur as a result of
natural ozone sources.
In contrast to Tables 6-3 and 6-4, Table 6-5 shows a number of examples
of sites in rural areas where ozone concentrations clearly seem to be influenced
by manmade pollutants. Examples of probable manmade influences are seen in
the 1973 data from McHenry, Maryland, and Kane, Pennsylvania, where 30 percent
or more of the samples over a 2- to 3-month period equalled or exceeded 0.08 ppm,
a value that had not been equalled in the older and more remote sampling data.
Wolf Point, Montana, data from 1975, as listed in Table 6-5, show results that
are more characteristic of true natural background concentrations, based on
the previous tabulations. All, however, of the data shown in Table 6-5 for
1975 from this Research Triangle Institute research program are much lower than
the data from earlier years. Whether this is because of the sampling period
selected, relevant weather factors, measurement calibration biases, or a
combination of these, is unknown.
Thus, as stated above, ozone/oxidant data from a variety of rural and
more remote nonurban sites show that these locations may experience a wide
variety of ozone concentration patterns. The assumption that a nonurban site
will be exposed only to natural or background atmospheric ozone concentrations
cannot be made. This point will be discussed in more detail in a later section
when newer research is introduced.
6.2.3 Seasonal and Diurnal Variations in Ozone or Oxidants Prior to 1970
Seasonal variations in tropospheric ozone concentrations have been observed
for many years. In the absence of anthropogenic influences peak concentrations
most frequently occurred in the spring months and were attributed to the
influence of seasonal changes in both the stratospheric source and vertical
transport mechanisms. Figures 6-1 and 6-2 show the average monthly ozone
concentrations at Quillayute, Washington, a sea-level coastal station; and
Mauna Loa, Hawaii, a mountain observatory 11,300 ft above MSL. Although data
for only 2 years or less are shown in these figures, a springtime relative maxi-
mum is discernable, especially in the data from the higher-altitude station at
Mauna Loa.
0190DL/A 6-11 6/15//84
-------
OZONE CONCENTRATION, ppm
OZONE CONCENTRATION, ppm
S
p
2
p p
CO O
ro
8
m 2
^m Jp
I?
3 -i
3 <'
CD *
3 °-
(D —•
2
O N)
0) '
WO
II
O
o
N
O
3
CD
0)
-«
a
5
3
CO
>i2
Z
O
ill
10 C
So
i T i i i i i i i rn^
1 1 1 1
o
z
?S
m m
55
o o
m m
O O
^ J
? O
> C
XK
I I I I H
cw
cog
51?
5'°
H
li
tt 3/
il
i?
l-f (Q
6' 3"
>*
CQ Q)
CD 7~
3
O I*
< to
^** ^J
^A ^J
«o?r
00 o.
CD
Q.
D3!
c_co
f 3
5?>
CD -*
wO
2oo
p
2
p
§>
S
O
O
p
*_k
S3
i- 5SE1
«,-o
Q) 3
CO
(0
3 CD
11
o
3
O
N
O
3
CD
Q)
"i
S"
5'
3
CO
SC
£>z
53g
~*
5 -'O
ll§
o »2
5a
I=s
SOD
wO
||
I X -
5 s
S 3
£ >
o o
m m
D >
> £
? P
m
CO
1
-------
For areas influenced by manmade sources, ozone concentrations tend to
follow a long-recognized pattern, with maximum average concentrations occurring
mid- to late summer. This pattern is attributable to seasonal meteorological
conditions that tend to favor the accumulation of higher concentrations of
pollutants and the occurrence of favorable photochemical reaction conditions
during the late summer months. Figure 6-3 shows monthly average hourly ozone
concentrations for three urban areas, Los Angeles, Denver, and Phoenix, averaged
over 1- to 2-year periods prior to 1970. Los Angeles, with a peak concentra-
tion in August and a secondary peak in October, and Denver, with a July maximum
concentration, both show the typical photochemically related summertime ozone
cycle.
An illustration of a nonurban area where the average monthly ozone concen-
trations seems to follow a typical urban seasonal cycle is shown in Figure 6-4,
which depicts 1973-1974 data for Whiteface Mountain in upstate New York. The
fact that this region is affected frequently by air masses traveling from the
highly urbanized regions located south and southwest of Whiteface Mountain would
tend to support strong manmade influences at this site. This cannot be supported
specifically for this 1973-1974 time period, however, and some investigators
(e.g., Coffey and Stasiuk, 1975a,b) have argued otherwise.
The diurnal concentrations of ozone follow a characteristic pattern in
urban source areas. This pattern is one in which the maximum concentrations
occur near midday, from the late morning to the early afternoon. Figure 6-5
shows two urban area examples of this diurnal pattern, one from Denver and one
from Philadelphia. This sort of pattern is usually explained by the fact that
ozone-forming photochemical reactions require several hours to produce a
maximum concentration and also that in mid-afternoon atmospheric dilution
processes and wind transport usually reach a maximum, thus producing the
afternoon decline in concentrations.
At a background site, a midday maximum can also be seen frequently. At
such a location the diurnal cycle may result from the photochemical cycle of
natural HC and NO precursor reactions and/or from the diurnal mixing cycle in
which the deeper midday mixing replenishes the ozone concentrations at the
surface from the mid-tropospheric layers. This mixing compensates for the
ozone scavenged at the earth's surface during the nighttime period of ground-
layer stability.
0190DL/A 6-13 6/15//84
-------
a
a
0.06
2 0.05
cc
z
u
o
o
H
<
a
x
o
<
vu
5
0.04
0.03
0.02
0.01
I \ I I I I I I \ T
LOS ANGE
19641965,
I I I I I I I I I I
JAN. FEB. MAR. APR. MAY JUN. JUL AUG. SEP. OCT.lNOV. DEC.
MONTH
Figure 6-3. Monthly variation of mean hourly oxidant
concentrations for Los Angeles and Denver.
Source: U.S. Department of Health, Education, and
Welfare (1970), modified
a
a
o
s
ui
O
U
O
N
O
0.07
0.06
0.05
0.04
0.03
0.02
0.01
I I I I I M Mill TITT
iiiiilllllilil.il
J FMAMJJA SONDJ FMAM
1973
1974
YEAR AND MONTH
Figure 6-4. Average monthly ozone
concentrations recorded at Whiteface
Mountain in New York.
Source: Singh et al. (1977); cited in U.S.
Environmental Protection Agency
119781 6-14
-------
0.20
o.
^ 0.18 —
O 0.16
<( 0.14
0.12
0.10
0.08
O
z
o
Jf 0.06
< 0.04 —
g 0.02-
i i i i I i m i i i i
PHILADELPHIA 6/15/68
I I I
12-1 2-3 4-5 6-7 8-9 10-11 12-1 2-3 4-5 6-7 8-9 10-11 12-1
a.m.t »|« p.m-1
HOUR OF DAY
Figure 6-5. Diurnal variation of hourly oxidant concentrations in
Philadelphia and Denver.
Source: U.S. Department of Health, Education, and Welfare (1970)
6-15
-------
Transport has been mentioned as a factor that influences the urban diurnal
cycle. This transport, by horizontal wind movement, serves to reduce the
concentration in the source area and to increase subsequently the concentrations
in a non-source area at a later time and in a downwind direction. Thus, when
the diurnal pattern for a given site shows a peak concentration rather late in
the day or at night, it is logical, though not necessarily correct, to attribute
this to transport influences.
It also is obvious that the maximum ozone concentrations that are observed
at a given site are a complex function of precursor sources and meteorological
conditions. Thus, the extrapolation of observational data from one region to
another or even within a given region is not a simple task and should be
approached with caution, especially when assessing potential ozone exposures
to which human populations, crops, ecosystems, and other receptors are subjec-
ted. Diurnal, day-to-day, and longer-term seasonal cycles, as well as regional
influences, may all be important in exposure assessment.
6.2.4 Trends in Nationwide Ozone and Oxidant Concentrations
Data presented in the preceding section, while covering a number of
years, do not permit the evaluation of actual trends in ozone or oxidant
concentrations. Determination of whether ozone concentrations in ambient air
are static, rising, or declining trends can only be determined from statistical
tests using comparable aerometric data for a number of years (chapter 5). The
trend in nationwide concentrations of ozone over the period 1975 through 1981
is shown in Figure 6-6 (Hunt and Curran, 1982).
Evaluation of national trends, as well as local or regional trends, in
concentrations of ozone in ambient air over the past 5 to 10 years is compli-
cated by several factors: (1) a change in calibration procedure recommended
by EPA in 1978 and promulgated in 1979 (see chapter 5); (2) the possible
effects on aerometric data of quality assurance procedures instituted by EPA
in 1979; (3) the influence of diverse regional meteorological conditions; and
(4) changes in precursor emissions.
Figure 6-6 indicates a small decline in the composite average level of
the second-highest 1-hour ozone concentration even when the possible effects
of the above factors are considered. The 209 sites included in this analysis
(Hunt and Curran, 1982) reported at least 50 percent of the possible hourly
values in at least 5 of the 7 years from 1975 to 1981.
0190DL/A 6-16 June 1984
-------
0.18
0.16
GL
a
p 0.14
-------
Assessment of how much of the observed decline in ozone concentrations
from 1975 through 1981 should be attributed to the 1979 promulgation of the
ultraviolet (UV) calibration method as the Federal Reference Method is not a
simple matter of applying a correction factor to existing aggregated aerometric
data. The monitoring practices at each of the 209 sites would have to be
examined in detail. Not all monitoring sites switched to the use of the UV
method simultaneously. The state of California, for example (in EPA Region IX),
had already been using the UV method before it was promulgated in February
1979. In addition, other states, in other regions, may have used the boric
acid-potassium iodide (BAKI) method before, after, or both before and after
promulgation of the UV method, since the BAKI procedure was allowed by EPA as
an interim method for 18 months following the 1979 UV promulgation (see
chapter 5). Likewise, other states used gas-phase titration prior to 1979 but
either BAKI or UV procedures following the UV promulgation. The relationship
among these three methods, even if monitoring practices at individual sites
were known, is complex and would preclude the simple application of a single
correction factor (see chapter 5). Hunt and Curran (1982) have noted that
Region IX is the only region that showed improvement in ozone air quality
between 1980 and 1981 but not long-term improvement. California, whicn domi-
nates Region IX, changed calibration in 1975.
The majority of ambient air monitoring stations in the nation are operated
by state and local agencies, but there is a small group of National Air Moni-
toring Stations (NAMS) (chapter 5) that is responsible directly to EPA. The
trend line for the subset of 49 NAMS ozone stations is also shown in Figure
6-6 and tracks fairly closely the line for all 209 stations.
6.3 OVERVIEW OF OZONE CONCENTRATIONS IN URBAN AREAS
An overview of nationwide urban ozone concentrations for 1981 is provided
in Figures 6-7 and 6-8, which depict graphically average daylight concentrations.
Figure 6-7 shows data for spring and summer months, the months which comprise
the smog season in most if not all areas of the nation, and Figure 6-8 shows
daylight concentrations during the fall and winter months. The daylight period
of 6:00 a.m. to 8:00 p.m. includes the hours of greatest human activity out-
doors; the hours when exposure of vegetation and ecosystems would be expected
0190DL/A 6-18 June 1984
-------
cr>
i
14 - .16 PPM
.16 - 18 PPM
Figure 6-7. Average daylight (6:00 a.m. to 8:00 p.m.) concentrations of ozone in the second and 01 Apr I983
third quarters (April through September), 1981.
Source: G. Ouggan, OAQPS, U.S. Environmental Protection Agency
-------
ro
o
00 - .02 PPM
02 - .04 PPM
.04 - .06 PPM
.06 - 08 PPM
.08 - 10 PPM
.10 - 12 PPM
.12 - .14 PPM
14 - 16 PPM
.16 - .18 PPM
Figure 6-8. Average daylight (6:00 a.m. to 8:00 p.m.) concentrations of ozone in the first and fourth Q1 Apr 1983
quarters (January through March and October through December), 1981.
Source: G. Duggan, OAQPS, U.S. Environmental Protection Agency
-------
to have the greatest consequences (stomata are open in daylight and photosyn-
thesis is taking place; see chapter 7); and the hours of greatest local forma-
tion of ozone and other oxidants via photochemistry in the atmosphere (see
chapter 4). The average concentrations during the spring and summer months
(second and third quarters of the year) are clustered mainly in the 0.04 to
0.06 ppm range. Averages for the winter and fall months (first and fourth
quarters) are clustered mainly in the 0.02 to 0.04 ppm range, which is within
the range of natural background concentrations.
The stations used in Figures 6-7 and 6-8 reported at least 75 percent of
the possible 1-hour values per quarter. Some stations, however, monitor ozone
only during the months when the potential for photochemical ozone formation is
significant in those localities. Also, certain areas of the United States are
not monitored routinely for ozone because of the lack of emission sources or
transport events and thus the low potential for significant ozone or oxidant
concentrations. The Great Basin and the Great Plains, for example, are such
areas.
Figure 6-9 shows the collective nationwide frequency distribution of the
second highest 1-hour 0~ concentration for 1979, 1980, and 1981. Only data
collected by the Federal Reference Method (chemiluminescence) or the equiva-
lent UV method (see chapter 5) have been used in this analysis. A "valid
site" is one reporting at least 75 percent of the 8760 possible 1-hour values
in a year. There were 282 such sites in 1979, 266 in 1980, and 358 in 1981
(U.S. Environmental Protection Agency, 1980, 1981, 1982). As shown by
Figure 6-9, 50 percent of the second-highest 1-hour values in this 3-year
period were 0.12 ppm or less and 10 percent were equal to or greater than
0.20 ppm. While the third-highest values are also of interest, the highest
and second-highest 1-hour values are of greater consequence in relation to
the existing ozone standard and, thus, in relation to their health and wel-
fare implications. The second-highest value determines whether an area is
in compliance with the present ozone standard.
Table 6-6 lists the second-highest 1-hour 0- values reported for 1979
through 1982 for the 80 most populous Standard Metropolitan Statistical Areas
(SMSAs), grouped by population. Collectively these SMSAs account for 54
percent of the 1980 United States population of 226.5 million. The significant
observation to be drawn from this table of second-highest values is that the
lowest median concentration in 1981, 0.12 ppm for SMSAs having populations of
0190DL/A 6-21 6/18//84
-------
cr>
ro
rv>
a
Z
O
99.99
0.45
0.40
0.35
0.30
< 0.25
LU
O
Z 0.20
O
O
111
O 0.15
O
0.10
0.05
99.9 99.8
~TT
99 98 95 90 80706050403020 10
2 1 0.5 0.2 0.1 0.05 0.01
M I I I I I I II I—I I MI/HI
HIGHEST
2nd-HIGHEST
3rd-HIGHEST
I I I i 1 I I I
I I I I I III II
II
0.01 0.05 0.1 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.8 99.9 99.99
STATIONS WITH PEAK 1-hour CONCENTRATIONS < SELECTED VALUE, percent
Figure 6-9. Collective distributions of the three highest 1-hour ozone concentrations for
3 years (1979, 1980, and 1981) at valid sites (906 station-years).
Source: U.S. Environmental Protection Agency, SAROAD data files for 1979,1980,1981
-------
TABLE 6-6. SECOND-HIGHEST 1-hour OZONE CONCENTRATIONS REPORTED FOR 80
STANDARD METROPOLITAN STATISTICAL AREAS HAVING POPULATIONS >0.5 MILLION
Ozone concentration, ppm
Standard Metropolitan Statistical Area
Population >2 million
New York, NY - NJ
Los Angeles - Long Beach, CA
Chicago, IL
Philadelphia, PA - NJ
Detroit, MI
San Francisco - Oakland, CA
Washington, DC - MD - VA
Dallas - Fort Worth, TX
Houston, TX
Boston, MA
Nassau - Suffolk, NY
St. Louis, MO - IL
Pittsburgh, PA
Baltimore, MD
Minneapolis - St. Paul, MN - WI
Atlanta, GA
Summary statistics:
Minimum 1-hour value
Median 1-hour value
Maximum 1-hour value
Population 1 to < 2 million
Newark, NJ
Anaheim - Santa Ana - Garden Grove, CA
Cleveland, OH
San Diego, CA
Miami, FL
Denver - Boulder, CO
Seattle - Everett, WA
Tampa - St. Petersburg, FL
Riverside - San Bernardino - Ontario, CA
Phoenix, AZ
Cincinnati, OH - KY - IN
Milwaukee, WI
Kansas City, MO - KS
San Jose, CA
Buffalo, NY
Portland, OR - WA
New Orleans, LA
Indianapolis, IN
Columbus, OH
1979
0.19
0.44
0.22*
0.18a
0.12a
0.14a
0.18a
0.17
0.24
0.22a
0.18a
0.16a
0.17a
0.14a
0.10a
0.16
0.10
0.175
0.44
0.15
0.35
0.14a
0.36
0.05a
0.16
0.13
0.11
0.42a
0.12a
0.13
0.17.
0.12a
0.17a
O.lla
0.11
0.12
0.12
0.10
1980
0.18
0.44
0.34
0.24a
0.15
0.18
0.19
0.18
0.30
0.15
0.17
0.18
0.17a
0.18a
0.13
0.15
0.13
0.18
0.44
0.15
0.29
0.12
0.22
0.15
0.13
0.09
0.13
0.38
0.15
0.16
0.14
0.16
0.19
0.14
0.10
0.12
0.14
0.12
1981
0.18
0.35
0.14
0.17
0.15
0.14
0.15
0.15
0.23
0.13
0.14
0.15
0.16
0.17
0.10
0.14
0.10
0.15
0.35
0.14
0.31
0.12
0.24
0.14
0.13
0.12
0.11
0.34
0.16
0.13
0.17
0.12
0.14
0.12
0.15
0.11
0.13
0.11
1982
0.17
0.32
0.12
0.18
0.16
0.14
0.15
0.17
0.21
0.16a
0.13
0.15
0.14
0.14
0.10
0.14
0.10
0.15
0.32
0.17
0.18
0.12
0.21
0.14
0.14
0.09
0.11
0.32
0.12
0.13
0.13
0.10
0.14
0.11
0.12
0.17
0.12
0.13
0190DL/A
6-23
6/15//84
-------
TABLE 6-6. SECOND-HIGHEST 1-hour OZONE CONCENTRATIONS REPORTED FOR 80
STANDARD METROPOLITAN STATISTICAL AREAS HAVING POPULATIONS >0.5 MILLION
(continued)
Ozone concentration, ppm
Standard Metropolitan Statistical Area
San Juan, PR
San Antonio, TX
Fort Lauderdale - Hollywood, FL
Sacramento, CA
Summary statistics:
Minimum 1-hour value
Median 1-hour value
Max i mum 1-hour value
Population 0.5 to < 1 million
Rochester, NY
Salt Lake City - Ogden, UT
Providence - Harwick - Pawtucket, RI - MA
Memohis, TN - AR - MS
Louisville, KY - IN
Nashville - Davidson, TN
Birmingham, AL
Oklahoma City, OK
Dayton , OH
Greensboro - Winston-Sal em - High Point, NC
Norfolk - Virginia Beach - Portsmouth, VA - NC
Albany - Schenectady - Troy, NY
Toledo, OH - MI
Honolulu, HI
Jacksonvi lie, FL
Hartford, CT
Orlando, FL
Tuisa, OK
AKron, OH
Gary - Hammond - East Chicago, IN
Syracuse, NY
Northeast Pennsylvania
Charlotte - Gastonia, NC
Allentown - Bethlehem - Easton, PA - NJ
Richmond, VA
Grand Rapids, MI
New Brunswick - Perth Amboy - Sayreville, NJ
West Palm Beach - Boca Raton, FL
Omaha, NE - IA
Greenville - Spartanburg, SC
Jersey City, NJ
Austin, TX
1979
NDb
0.11
0.10a
0.16a
0.05
0.125
0.42
0.12
0.15
0.17,
O.lla
0.16a
0.09a
NDD
O.ll3
0.14a
o.ioa
0.10
0.13
0.15
0.04a
0.13
0.20
0.10a
0.13
0.15
0.133
0.13a
0.11
0.123
0.17a
0.13a
0.11
0.103
0.083
o.ioa
O.ll3
0.15a
0.12a
1980
NDb
0.12
0.12
0.17
0.09
0.14
0.38
0.12
0.17
0.21
0.13
0.19
0.13
0.16
0.12
0.13
0.12
0.12
0.13
0.14
0.04
0.12
0.24
0.09
0.15
0.11
0.15
0.11
0.15
0.14
0.15
0.13
0.11
0.19
0.09
0.14
0.11
0.16
0.13
1981
0.07
0.12
0.11
0.17
0.07
0.13
0.34
0.12
0.15
0.15
0.12
0.14
0.13
0.16
0.11
0.12
0.11
0.11
0.13
0.13
0.04
0.10
0.15
0.10
0.15
0.24
0.14
0.11
0.10
0.12
0.12
0.12
0.11
0.13
0.09
0.08
0.11
0.14
0.12
1982
0.02
0.14
0.09
0.16
0.02
0.13
0.32
0.11
0.14
0.15
0.12
0.17
0.11
0.15
0.11
0.16
0.11
0.10
0.12
0.12
0.04
0.11
0.16
0.09
0.13
0.14
0.13
0.12
0.16
0.12
0.14
0.12
0.11
0.16
0.09
0.09
0.11
0.14
0.11
0190DL/A
6-24
6/15//84
-------
TABLE 6-6. SECOND-HIGHEST 1-hour OZONE CONCENTRATIONS REPORTED FOR 80
STANDARD METROPOLITAN STATISTICAL AREAS HAVING POPULATIONS >0.5 MILLION
(continued)
Ozone concentration, ppm
Standard Metropolitan Statistical Area
Youngstown - Warren, OH
Tucson, AZ
Raleigh - Durham, NC
Springfield - Chicopee - Holyoke, MA - CT
Oxnard - Simi Valley - Ventura, CA
Wilmington, DE - NJ - MD
Flint, MI
Fresno, CA
Long Branch - Asbury Park, NJ
Summary statistics:
Minimum 1-hour value
Median 1-hour value
Maximum 1-hour value
1979
0.13
0.10
0.10a
0.16a
0.19
0.16a
0.11
0.18a
0.143
0.04
0.13
0.20
1980
0.12
0.10
0.13
0.15
0.18
0.17a
0.11
0.19a
0.169
0.04
0.13
0.24
1981
0.13
0.12
0.12
0.16
0.20
0.12a
0.11
0.17
ND6
0.04
0.12
0.27
1982
0.11
0.12
0.09
0.15
0.22
0.16
0.11
0.16
NDB
0.04
0.12
0.22
Fewer than 90 days of data.
bND = no data.
Source: U.S. Environmental Protection Agency, SAROAD data files for 1979-1982
0190DL/A
6-25
6/15//84
-------
0.5 to 1 million, equals the current national ambient air quality standard for
ozone. The suggestion of an increase with increasing population is largely
the disproportionate influence of southern California SMSAs having populations
greater than 1 million.
6.4 OVERVIEW OF OZONE CONCENTRATIONS IN NONURBAN AREAS
As mentioned in the preceding section, very few ozone monitoring stations
are located in nonurban areas. Consequently, the aerometric data base for
nonurban areas is not comparable to that for urban areas. The nonurban data
presented in this section were obtained from two special-purpose monitoring
networks that were designed to provide ozone concentrations at sites specifi-
cally selected to represent a variety of pristine or rural nonurban environ-
ments. These sites do not all represent areas totally unaffected by manmade
ozone or its precursors, as shown by the fact that some data records contain a
significant number of high values that are best explained as resulting from
the transport of ozone or its precursors from upwind urban areas. The data
given here are intended to show an overview of nonurban concentrations in
areas with relatively infrequent urban influences. Additional data on speci-
fic rural areas are presented in sections 6.5.1 and 6.5.2.
6.4.1 National Air Pollution Background Network (NAPBN)
The NAPBN consists of eight stations located in eight National Forests
(NF) across the country (Figure 6-10). The first three stations began opera-
tion in 1976 (Green Mountain NF, Vermont; Kisatchie NF, Louisiana; and Custer
NF, Montana); the second three in 1978 (Chequamegon NF, Wisconsin; Mark Twain
NFS Missouri; and Croatan NF, North Carolina); and the last two in 1979 (Apache
NF, Arizona; and Ochaco NF, Oregon). Yearly summaries of ozone concentrations
through 1980 are shown in Table 6-7 for the three sites established first
(Evans et al. , 1983). The principal points of interest in these summary sta-
tistics are the range of ozone concentrations and the arithmetic mean of the
values measured at these National Forest sites. The arithmetic mean concen-
trations for the three sites ranged from 0.027 ± 0.015 ppm at the Kisatchie NF
site in 1979 to 0.040 ± 0.011 ppm at the Custer NF site in 1979. The arithmetic
mean across all years and all three sites is 0.033 ppm. Fluctuations in the
observed concentrations from year-to-year and site-to-site are demonstrated by
0190DL/A 6-26 6/15//84
-------
CHEQUAMEGON NF
CROATAN NF
Figure 6-10. Locations of the eight national forest (NF) stations com-
prising the National Air Pollution Background Network (NAPBN).
Source: Evans et al. (1983)
6-27
-------
TABLE 6-7. ANNUAL OZONE SUMMARY STATISTICS FOR THREE NAPBN SITES
Site Year
Kisatchie NF, LA 1976
1977
1978
1979
1980
en Custer NF, MT 1976
^ 1977
00 i 1978
1979
1980
Green Mt. NF, VT 1976
1977
1978
1979
1980
No. 1-hour
measurements
3448
6793
5636
6993
4438
275
7603
7674
8488
7754
1058
6483
3671
6423
8574
% of
possible
l~hr meas.
39.4
77.5
64.3
79.8
50.7
3.1
86.8
87.6
96.9
88.5
12.1
74.0
41.9
73.3
97.9
Cone.
Min.
LDa
LD,
LD
LD
LD
0.020
LD
LD
LD
LD
LD
LD
LD
LD
LD
> ppm
Max.
0.125
0.135
0.125
0.100
0.105
0.060
0.080
0.075
0.070
0.070
0.060
0.145
0.105
0.105
0.115
Cone,
Arith.
Mean
0.032
0.033
0.034
0.027
0.028
0.039
0.040
0.030
0.032
0.037
0.029
0.038
0.029
0.032
0.032
, ppm
Arith.
std. dev.
0.021
0.023
0.021
0.015
0.016
0.008
0.011
0.017
0.012
0.012
0.011
0.021
0.018
0.017
0.017
Cone
Geom.
mean
0.024
0.025
0.027
0.023
0.023
0.038
0.039
0.023
0.029
0.035
0.026
0.031
0.024
0.027
0.027
. , ppm
Geom.
std. dev.
2.19
2.25
2.14
1.92
1.94
1.22
1.37
2.14
1.59
1.41
1.76
2.00
2.01
1.86
1.90
LD = less than detectable.
Source: Evans et al. (1983)
-------
the range of concentrations measured and by the size of the standard deviations
as well. The lowest concentrations seen were below the limits of detection of
the chemiluminescence monitor employed, but the highest concentrations observed
at the Kisatchie NF and Green Mt. NF sites were both above the present ozone
standard of 0.12 ppm.
These summary statistics show somewhat higher mean concentrations, lower
maximum concentrations, and lower standard deviations in data obtained at the
Custer NF site than at the other two, which may indicate that meteorological
conditions are less variable at that site or that the site is much less affec-
ted, if not altogether unaffected, by manmade ozone or its precursors. Previous
data shown in Table 6-5 for Wolf Point, Montana, are generally consistent with
the Custer NF data.
During a 6-day period in 1979, the NAPBN site in the Mark Twain NF,
Missouri, showed ozone concentrations well in excess of typical values. A
1-hour value of 0.125 ppm, the maximum observed at any NAPBN site in 1979, was
measured at that site on July 21, 1979. Evans et al. (1983) have calculated
the trajectories of air masses reaching the site during the 6-day period of
July 18 through July 23, 1979. They ascribed the unusually high values,
including the peak value on the 21st, to pollutants picked up as the trajectory
passed over urban areas in the Ohio River Valley and the Great Lakes region.
Table 6-8 shows the peak 1-hour value for each of the 6 days. Figure 6-11
shows the trajectories for the air parcels reaching the Mark Twain NF site at
midnight (0000), 8 a.m. (0800), noon (1200), and 6 p.m. (1800) on July 21,
1979. On July 23, clouds and rain spread over the region and the air-flow
trajectories shifted to the east and south, reducing both the quantities of
transported precursors and the potential for photochemical ozone generation.
6.4.2 Sulfate Regional Experiment Sites (SURE)
As part of a comprehensive air monitoring project sponsored by the Electric
Power Research Institute (Martinez and Singh, 1979) ozone data were collected
by the chemi luminescence method in the last 6 months of 1977 at the nine
"nonurban" SURE sites in the eastern United States shown in Figure 6-12. On
the basis of diurnal NO patterns that indicated the influence of traffic
/\
emissions, five of the sites were classed as "suburban"; the other four were
classed as "rural." The ozone data from these nine stations are summarized in
Table 6-9. Martinez and Singh (1979) noted that the four rural stations
0190DL/A 6-29 6/15//84
-------
TABLE 6-8. CONCENTRATIONS OF OZONE DURING 6-day PERIOD OF HIGH
VALUES AT NAPBN SITE IN HARK TWAIN NATIONAL FOREST, MISSOURI, 1979
1-hr maximum
Date 03 concentration, ppm
July 18 0.080
July 19 0.100
July 20 0.115
July 21 0.120
July 22 0.125
July 23 0.050
Source: Evans et al. (1983)
0190DL/A 6-30 6/15//84
-------
MINNEAPOLIS
INCINNATI
^v
LOUISVILLE
KANSAS CITY
ST. LOUIS
Figure 6-11. Trajectory analysis plots at the
NAPNB site at Mark Twain National Forest, MO,
July 21, 1979 (distance between bars represents
12 hr).
6-31
-------
0 50 150 250
I ' ' ' ' »
km
Figure 6-12. Location of SURE Monitoring Stations.
Source: Martinez and Singh (1979)
6-32
-------
TABLE 6-9. SUMMARY OF OZONE CONCENTRATIONS MEASURED AT SULFATE REGIONAL EXPERIMENT
(SURE) NONURBAN STATIONS, AUGUST THROUGH DECEMBER 1977
CO
CO
Number of measurements
Rural sites
#1 Montague, MA
#4 Duncan Falls, OH
#6 Giles Co. , TN
#9 Lewisburg, WV
Suburban sites
f2~Scranton, PA
#3 Indian River, DE
#5 Rockport, IN
#7 Ft. Wayne, IN
#8 Research Triangle
Park, NC
Total no. of
measurements
3419
3441
3632
3459
3410
3017
3462
3438
3495
with concentrations:
>0.08 ppm
60
52
63
23
0
29
29
0
80
>0.10 ppm
33
2
5
3
0
0
0
0
10
>0.12 ppm
21
0
0
0
0
0
0
0
0
Mean
concn,. ppm
0.021
0.029
0.026
0.035
0.023
0.030
0.025
0.020
0.025
Mean of
daily
1-hour
maxima,
ppm
0.044
0.049
0.052
0.054
0.035
0.049
0.046
0.039
0.050
1-hour
maximum,
ppm
0.153
0.107
0.117
0.106
0.077
0.099
0.099
0.080
0.118
Source: Martinez and Singh (1979)
-------
occasionally recorded high values comparable with urban areas, but that the
incidence was low. They concluded that infrequent transport of ozone or its
precursors, or both, rather than local ozone generation, was the most probable
cause of these high values.
6.5 VARIATIONS IN OZONE CONCENTRATIONS: DATA FROM SELECTED URBAN AND NONURBAN
SITES
Variations of ozone concentrations by season and by time of day, as
discussed qualitatively in section 6.2, have been long known and are well
documented. First studied in smog chambers, diurnal patterns have since been
corroborated by field investigations, and exceptions to such general patterns
have been examined and documented. Likewise, field investigations have substan-
tiated general seasonal patterns and exceptions to them, and have also estab-
lished a number of spatial variations in concentration, such as those that
occur with latitude or with altitude. While it is difficult to discuss temporal
and spatial variations separately, this section is subdivided along those
lines for convenience.
6.5.1 Temporal Variations in Ozone Concentrations
In section 6.2, diurnal and seasonal data for ozone concentrations were
presented as reported in the 1978 criteria document for ozone and other photo-
chemical oxidants. More recent data showing such temporal variations in ozone
concentrations will be reported here to provide a more detailed discussion.
6.5.1.1 Diurnal Variations in Ozone Concentrations. By definition, diurnal
variations are those that occur during a 24-hour period. Diurnal patterns of
ozone may be expected to vary with location, depending on the balance among
the factors affecting ozone formation, transport, and destruction. Figure
6-13 shows the diurnal pattern of ozone concentrations on July 13, 1979, in
Philadelphia, Pennsylvania. On this day a peak 1-hour average concentration
of 0.20 ppm, the highest for the month, was reached at 2:00 p.m., presumably
as the result of local photochemical processes. The severe depression of
concentrations to below detection limits (less than 10 ppb ) between 3:00 and
6:00 a.m. is usually explained as resulting from the scavenging of ozone by
local nitric oxide emissions. In this regard, this station is typical of most
urban locations.
0190DL/A 6-34 6/15//84
-------
I I I I I I I I I I I I
12 1 234 567 89 10 11 t 1 234 56789 10 11
NOON
a.m. HOUR OF DAY p.m.-
Figure 6-13. Diurnal pattern of 1-hour ozone concen-
trations on July 13, 1979, Philadelphia, PA.
Source: U.S. Environmental Protection Agency,
SAROAD data file for 1979
6-35
-------
Diurnal profiles of ozone concentrations can vary from day to day at a
specific site because of changes in the various factors that influence con-
centrations. Such day-to-day variations are clearly demonstrated in Figure
6-14, which shows diurnal variations in ozone concentrations on 2 consecutive
days at the same monitoring site in Detroit, Michigan. Differences in timing
and magnitude occur that are especially noticeable between midnight and about
7:00 a.m. Transport is probably involved in these nighttime variations. The
afternoon peaks, the actual maxima for the 2 days, differ in magnitude but not
in timing.
Composite diurnal data, that is, concentrations for each hour of the day
averaged over multiple days or months, often differ markedly from the diurnal
cycle shown by concentrations for a specific day. In Figures 6-15 through
6-18, diurnal data for 2 consecutive days are compared with composite diurnal
data (1-month averages of hour-by-hour measurements) at each of two urban
(Washington, D.C., and St. Louis County, Missouri) and two agriculture-oriented
sites (Alton, Illinois, and N. Little Rock, Arkansas). Several obvious points
of interest present themselves in these graphs: (1) at some sites at least,
peaks can occur at virtually any hour of the day or night but these may not
show up strongly in the longer-term average data; (2) some sites may experience
multiple peaks during a 24-hour period; and (3) disparities, some of them
large, can exist between peaks (the diurnal data) and the 1-month mean (the
composite diurnal data) of hourly ozone concentrations. These are only exam-
pies of the differences that can occur between daily and monthly mean concen-
tracion patterns. Since these patterns differ from site to site, no conclusions
about comparative levels at a given site or between urban and rural sites can
be drawn from these figures, especially since the rural and suburban data in
these examples come from months generally considered to be outside the photo-
chemical smog season.
The effects of averaging are readily apparent when diurnal ozone con-
centrations are compared with "composite diurnal" ozone concentrations.
Figures 6-19 and 6-20, based on 3-month averages, demonstrate rather graphical-
ly, when compared with Figures 6-15 through 6-18 (daily values and 1-month
averages) the effects of lengthening the period of time over which values are
averaged. The figures show composite diurnal patterns calculated on the basis
of 3 months. While seasonal differences are seen, and will be discussed
later, the comparison of 3-month and 1-month composite diurnal concentrations
0190DL/A 6-36 6/15//84
-------
240
I I I I I I I I I I I
• SEPTEMBER 20
O SEPTEMBER 21
I I I I I I I I i I I I I I"
12 12 3456789 10 11 I 1234 56789 10 11
, NOON
H a.m. HOUR OF DAY P'm'
Figure 6-14. Diurnal patterns of ozone concentra-
tions, September 20 and 21, 1980, Detroit, Ml. (1960
3 = 1 ppm)
Source: U.S. Environmental Protection Agency,
SAROAD data file for 1980
6-37
-------
>o
= 5
»
I
i3
S?
(D
n
"
0
3
I
»
3
0
1
f
*
T3
CCQ
2 C
*2
T
o
0
-
•
T3 O
•*
C
3J
(D _l
(O
00
3 O
O 3
CD r+
33-
^
a o
^.0
11
0
-,
z
O
0
2
w i
2?£ 1
r* CD _1_
OZONE CONCENTRATION, ppm
p
S
OZONE CONCENTRATION, ppm
o
S
p
2
o
CO
p
1.*
ro
—i coco ^9:31 T
- 11 111 I
3? 3
<
'O
CD'<
CD
3 —
Oi3-2
CO CD
003
< -• —
co 9J
— -»c* !3S Q.
Protection Agency
1 -month composite
ne concentrations,
I
2
O
O
z
§!
•p
3
..
00
p
°
p
S
p
S
o
S
p
0
ro
00
1
S
:Ccl^
CENTER
^TTV
-------
W. If-
i o.io
a
Z
p 0.08
ec
K
2 0.06
O
Z
0 0.04
Ul
Z
N 0.02
O
0
III I M
1 1
1
III 1 1 1 1
SUBURBAN-RESIDENTIAL
OCT. 11-12
— — — 1 -month
—
—
;
/
— /
t
*
T T'X
— I if
^^k.j
I fl •*• I
r \
\
1
^
I "
f
vr
I •
^
-
—
-f __
^
• /
»^-*
• 1 •
_
\
^ I/*^
~~^k~~J '" - - ^^ **""'"
• 1 1 1 •*• 1 1 1 1 1 •*• 1 S
24 4 8 12 16 20 24 4 8 12 16 20 2'
Ud~-.a-m._NOON_n.m
^ a.m.— NOOM p.m. .^ 1
HOUR OF DAY
Figure 6-17. Diurnal and 1-month composite
diurnal variations in ozone concentrations, Alton,
IL, October 1981 (fourth quarter).
Source: U.S. Environmental Protection Agency,
SAROAD data file for 1981
a
a
O
Ul
O
Z
o
o
Ul
I
O
U. If.
0.10
0.08
0.06
0.04
0.02
0
I
i i i i i i i
1 1 1 1 1
1
i i i
1
RURAL-NEAR URBAN
_ NOV. 22-23
—— — "• 1 -month
— —
/\l
"~ 7 \
i
t
— -
to— •
%
i
^^
V^r'-1
•
i >
r
[ •
,r— \
>
i i
• •
-p
**
/
/
\ /*
\/
^>
i
*'
•*
\ .
~\
i
•
. i •
"\ T
\
'xx
\
| -
/
/
h J
f3i
A
L 1 '
—
'
S^
1
—
^~'
^>
1
24 4 8 12 16 20 24
NOON*— p.m.
8 12 16 20 24
-a.m.—NOON—p.m.—>»|
HOUR OF DAY
Figure 6-18. Diurnal and 1 -month composite
diurnal variations in ozone concentrations, N.
Little Rock, AR, November 1981 (fourth quarter).
Source: U.S. Environmental Protection Agency,
SAROAD data file for 1981
6-39
-------
\l.l\l
P 0.09
1 0.08
2
0 0.07
P
< Oft
Z 0.05,
UI
Z 0.04
O
0 0.03
UI
•y
0 0.02
N
0.01
0
2
I I I I I I I I
- ALTON, IL
_
—
—
_
—
— — -. *
— >n — •
I I I I I I I I
42468
I I I I I I I II
.-*-• .3rd
/'x'*2nd^<>
6'
...-••"4th Q "'•-.
I I I I I I I 1 I
10 12 14 16
-NOON,
I \
Q.
V
\
*
I \
18
n.m.i
I I I 1
—
„_
_
—
MB
—
—
^»SB.T-^!
—
1 1 1 1
20 22 2
HOUR OF DAY
Figure 6-19. Composite diurnal patterns by quarter
of ozone concentrations at a rural agricultural
site, Alton, IL, 1981.
Source: U.S. Environmental Protection Agency,
SAROAD data file for 1981
E
a
o.
0
\
^^^-^^^^^
'" •••'4th Q
—
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
42 4 6 8 10 12 14 16 18 20 22
I
—
—
—
_
•~™
—
"^
^
—
|
2-
HOUR OF DAY
Figure 6-20. Composite diurnal patterns by quarter
of ozone concentrations at a rural agricultural
site, N. Little Rock, AR, 1981.
Source: U.S. Environmental Protection Agency,
SAROAD data file for 1981
6-40
-------
at the same sites readily demonstrates the smoothing out of peak concentrations
as the averaging period is lengthened. Thus, a fourth pertinent point of
interest, related to the third point, emerges from the data presented above:
that is, increasing the averaging time obscures the magnitude and time of
occurrence of peak ozone concentrations. This is an obvious and familiar
result in the statistical treatment of sampling data, but one that is highly
pertinent to the protection of human health and welfare from the effects of
ozone.
Quantitative analyses of the relationships among 1-hour peak concen-
trations or second-highest 1-hour peak concentrations and daylight, diurnal,
monthly, seasonal, and yearly average ozone concentrations lie outside the
scope of this document. The relationships of peak and mean concentrations
assume more or less significance, depending upon whether the health and wel-
fare effects of exposure to ozone are solely concentration-dependent, heavily
concentration-dependent, or both concentration- and time-dependent. Neverthe-
less, if acute exposures are the chief cause of adverse health and welfare
effects, careful attention will have to be paid to the relationship of 1-hour
versus other averaging times.
Regarding other points of interest seen in these several figures, the
significance of the occurrence of peak concentrations outside of daylight
hours is lessened by the fact that most people spend the nighttime hours
indoors and the fact that the stomata of green plants are apparently closed at
night. (Gradients between indoor and outdoor concentrations of ozone, relative
to human exposures, are briefly discussed in section 6.5.2. The relationship
of stomatal function to the effects observed with ozone exposure of green
plants is discussed in chapter 7.)
No attempt is made in this section to document the respective contribu-
tions of local formation of ozone versus transport of ozone; however, the
occurrence of multiple peak ozone concentrations within a 24-hour period is
usually construed as indicating the presence of ozone transported to the site
from elsewhere (as discussed in chapter 4). An illustration of the diurnal
variations that are seen when transport occurs is shown by Figure 6-21, where
dual peaks occur on each of three successive days at a site of the Sulfate
Regional Experiment (SURE) network.
An example is shown in Figure 6-21 of the occurrence of dual peaks of
high ozone concentrations on each of 3 consecutive days of high concentrations.
0190DL/A 6-41 6/15//84
-------
a.m. NOON p.m.
SATURDAY, 27 AUGUST
24
a.m. NOON p.m.
SUNDAY, 28 AUGUST
24
a.m. NOON p.m.
MONDAY, 29 AUGUST 1977
24
Figure 6-21. Three-day sequence of hourly ozone concentrations at Montague, MA, SURE station
showing locally generated midday peaks and transported late peaks.
Source: Singh and Martinez (1979)
-------
One of the more important questions regarding the effects of ozone on both
people and plants is the possible significance of high concentrations lasting
1 hour or longer on each of 2 or more consecutive days.
In human controlled exposures, attenuation of response to ozone has been
observed at about 0.20 to 0.50 ppm in exercising subjects upon repeated,
consecutive-day exposures (see chapter 11). That attenuation is lost after
exposures to those levels cease (see chapter 11 for the time course of loss of
attenuation). This finding raises the important question of what patterns of
repeated ambient air exposures might be experienced by communities in high-ozone
areas, as well as in other parts of the country.
Data records for 6:00 a.m.-to-8:00 p.m. ozone concentrations in the
second and third quarters of the year, 1979 through 1981, have been examined
for Pasadena and Pomona, California; Dallas, Texas; and Washington, D.C.
These cities were chosen because adequate aerometric data records were avail-
able, because they represent areas known to experience high ozone concentrations
(California), and because they represent different geographic regions of the
country (west, southwest, east). Similar data could be compiled for any city
for which sufficient aerometric data exist. The choice of the daylight period
and of the second and third quarters is consistent with known diurnal (see
Figure 6-22) and seasonal patterns (Figures 6-19 and 6-20) of ozone concentra-
tions and with patterns of typical human and crop or ecosystems exposures.
One can count from the data records the number of exposures to ozone that
occur for respective durations (2, 3, 4, etc., consecutive days) at concen-
trations equal to or greater than specified concentration ranges. A cumula-
tive tally of such consecutive-day exposures results in the kind of data shown
in Table 6-10. Plotting of the cumulative number of exposures of respective
durations (in days) for respective concentration cutoffs produces the log-
probability graphs shown in Figures 6-23 through 6-26.
For this discussion, an n-day period at or above the concentration cutoffs
is simply called an "exposure" and an n-day period below the respective cutoffs
is called a "respite."
The intuitive expectation is that for successively higher daily concentra-
tion limits, the number of "exposures" in a given time at a given place will
become smaller and the length of the intervening "respites" will increase.
Table 6-10 shows that this expectation is supported by data from a number of
locations across the country. This table shows that the total number of days
0190DL/A 6-43 6/15//84
-------
0.08
o.
a
d
O
LU
O
z
o
u
ui
Z
o
N
o
0.06
0.04
0.02
24
a.m.
NOON
p.m.
24
Figure 6-22. Composite diurnal ozone pattern at an
Argonne, IL, agricultural site, August 6 through September
30, 1980.
Source: Kress and Miller (1983)
6-44
-------
TABLE 6-10. TOTAL DAYS WHEN MAXIMUM DAILY OZONE CONCENTRATION EXCEEDED
OR WAS LESS THAN SPECIFIED CONCENTRATIONS
APRIL THROUGH SEPTEMBER, 1979 THROUGH 1981, AT PASADENA
AND POMONA, CALIFORNIA, AND AT WASHINGTON, D.C., AND DALLAS, TEXAS
Location Concentration (ppm)
<0.06 > 0.06 <0.12 > 0.12 <0.18 > 0.18
Pasadena, CA 44 488 160 372 303 229
Pomona, CA 70 472 207 335 373 169
Washington, DC 296 146 437 5 442 0
Dallas, TX 124 327 412 39 449 2
0190DL/A 6-45 6/15//84
-------
W
I
«,
O
ui
cc
CO
o
0.
X
uu
7
6
5
1
7'
3
Q.
(0
HI
CC
i i i i i i
(A)
"EXPOSURES"
0.18 ppm —
D o.12 ppm
O 0.06 ppm
(B)
"RESPITES"
_LJ I
-'l I I I I I II I LJ
99.99 989590 80706050403020 10 5 21 0.01
Figure 6-23. Probability that "exposures" and "respites" for
specified concentration cutoffs will persist for indicated or
longer period at Pasadena, CA, based on aerometric data for
April through September, 1979 through 1981.
6-46
-------
M
n
•o
UJ
tc.
U>
O
Q.
X
UJ
71
6
5
4
3
<0
•o
O
z
111
LLJ
E 2
UJ
cc
i i i i i i i
(A)
"EXPOSURES"
.'O
i rTT_
A 0.18 ppm
O- 0.12 ppm
O 0.06 ppm
(B)
"RESPITES"
.O
L.I I Lfxf I I I I I I I I I U
' ' ff • •
99.99 98 95 90 80 70 60 50 40 30 20 10 5 21 0.01
Figure 6-24. Probability that "exposures" and "respites" for
specified concentration cutoffs will persist for indicated or
longer period at Pomona, CA, based on aerometric data for April
through September, 1979 through 1981.
6-47
-------
7
in 6
(0 R
•O °
i4
LU 3
_rn l l i i i i i i i ".f r
UJ
oc
3
0)
O
a.
x
2
*
1
LU
2
(A)
"EXPOSURES"
/O
.•' 0.06 ppm
••o
/o
o-
7-P-
_ C
a-'
(B)
__ "RESPITES'
.-' 0.06 ppm
/o
I I I 4 I I I I I I I I I ..I
99.99 98 95 90 80 70 60 50 40 30 20 10 5 21 0.01
Figure 6-25. Probability that "exposures" and "respites" for
specified concentration cutoffs will persist for indicated or
longer period at Washington, DC, based on aerometric data for
April through September, 1979 through 1981.
6-48
-------
•a
eUrn 'i~i FITI rT°iT i I i'T
5
4
O
LU 3
O
a.
x
LJJ
1
71
Z 6
| 5
E 2
to
LU
QC
(A) or* 9
"EXPOSURES"
Q-' Px
0'
/
s
s
a --- 0.12 ppm
O ....... 0.06 ppm
O''
P
d
(B) °/
"RESPITES" o/
i i i i i i i i u
99.99 98 95 90 80 70 60 50,40 30 20 10 5 21 0.01
Figure 6-26. Probability that "exposures" and "respites" for
specified concentration cutoffs will persist for indicated or
longer period at Dallas, TX, based on aerometric data for April
through September, 1979 through 1981.
6-49
-------
associated with "exposures" decreases as the limiting concentration increases,
and vice versa for "respite" days. Figure 6-23 shows that for Pasadena the
lengths in days of "exposures" and "respites" have log-probability distribu-
tions. For example, this distribution of Pasadena "exposures" and "respites"
shows that there is a 70 percent probability that a 0.06 ppm "exposure" will
be 3 days or longer; and the probability of an "exposure" of 3 days or longer
is 55 percent for 0.12 ppm and 42 percent for 0.18 ppm. As would be expected,
the probability of a given length of "respite" at Pasadena increases with
increasing concentration; and while the probability of a "respite" at a concen-
tration lower than 0.06 and 3 days or longer is only 18 percent, a 3-day or
longer "respite" from 0.12 ppm or 0.18 ppm concentrations has probabilities of
34 percent and 52 percent, respectively.
Table 6-10 also shows total days of occurrence ("exposures") and relief
("respites") from concentrations of 0.06, 0.12, and 0.18 ppm for Pomona,
California; Washington, D.C.; and Dallas, Texas, for the period April through
September, 1979 through 1981. Figures 6-24, 6-25, and 6-26 show the probabili-
ties that "exposures" and "respites" of indicated lengths or numbers of days
will occur for Pomona, Washington, D.C., and Dallas, respectively. The results
for Pomona are not especially different from those shown for Pasadena, as might
be expected for sites close together in the Los Angeles basin. Both Washington,
D.C. and Dallas show significantly fewer "exposure" days than the California
stations at each of the three limiting concentrations; and Washington, D.C.
actually experienced no "exposures" equal to the 0.18 ppm limit. The probability
plots for Washington, D.C., and Dallas (Figures 6-25 and 6-26) do not show results
for the higher concentration limits because of the few occurrences in these cate-
gories.
These tabulations and probability plots have been presented to show the
probable distribution of concentrations by number of consecutive days for
these specific sites. These are descriptive, based on 3 years of data at
these specific sites. These tabulations and figures cannot be used to predict
the probable concentration that might accompany a given "exposure" or "respite"
duration.
6.5.1.2 Seasonal Variations in Ozone Concentrations. In addition to the
diurnal cycles and between-day variations discussed in the preceding section,
seasonal variations in ozone concentrations occur (for the reasons discussed
in chapter 4) and usually assume characteristic patterns.
0190DL/A 6-50 6/15//84
-------
In order to compile an assessment of potential ozone damage to the six
leading commercial crops in the United States (corn, soybeans, hay, wheat,
cotton, and tobacco), Lefohn (1982) surveyed 304 ozone monitoring stations and
identified 24 that (1) were located in counties producing significant quantities
of one or more of these six crops in 1978; (2) reported at least 50 percent of
possible hourly data in 1978; (3) reported an hourly maximum of at least 0.1
ppm 03; and (4) ranked high in cumulative ozone exposure for the period April
to October, 1978. Six of these sites represented counties high in soybean,
wheat, or hay production. Quarterly composite diurnal patterns for 6 of these
sites with reasonably complete (>75 percent) 1981 data are shown in Figure
6-27 (U.S. Environmental Protection Agency, SAROAD file). The average levels
are apparently comparable with the long-term averages at the NAPBN sites pre-
viously discussed (section 6.4.1). In addition, the diurnal patterns for
these sites clearly show the clustering of the afternoon levels into two
seasons, the low "winter" levels in the first and fourth quarters and the
higher "summer" levels in the second and third quarters of the year.
Although averaging causes details to be obscured, the average diurnal
patterns in Figure 6-27 show that the time of occurrence of peaks differs
among sites. Among the sites shown in Figure 6-27, ozone concentrations
appear to peak at 2:00 to 2:30 p.m. in Little Rock in the higher-concentration
second and third quarters. At Bakersfield in the second and third quarters,
there is evidence, even in these smoothed-out curves, of two peaks, the first
at about 1:00 p.m. and the second at 5:00 to 6:00 p.m. At the Clark County,
Ohio, site, the peak concentrations in the second and third quarters center
around about 5:00 p.m., but they do not return to "baseline" until after
midnight. These patterns appear to indicate transport into the areas. It is
also possible that single peaks that are shifted to mid- to late afternoon are
the product of transport. Depending upon proximity to urban centers and wind
speed and direction, rural areas usually experience their peak concentrations
later than those of urban areas, often but not always, within daylight hours.
Composite diurnal variations in ozone concentrations at a rural site in
Argonne, Illinois, over a 7-week period of the third quarter of 1980 were
shown in Figure 6-22. The actual day-to-day variations in ozone concentration
over the entire third quarter of 1980 at that site are shown for comparison in
Figure 6-28. As part of the National Crop Loss Assessment Network, the site
0190DL/A 6-51 6/15//84
-------
a
a
z
o
H
UJ
I
LLI
O
N
O
a
a
O
P
O
O
UJ
O
N
O
I
a
g
O
a
o
N
O
0.10
0.09
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
I I I I M M I M I
N. LITTLE ROCK, AR
(A)
1st a
2nd Q
3rd Q
4thQ „,
I I
24
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
I I I I I I I I I I I I I I I I I I I I I I
2 4 6 8 10 12 14 16 18 20 22 24
a.m. *^* p.m.
I I I I TITT IT I M I I I I I I I I I
SACRAMENTO, CA
(C)
24
0.10
0.09
k
2 4 6 8 10 12 14 16 18 20 22 24
a.m »-|*4 p.m., »-|
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
I I I I I I I ITT 1 I I I I I I I I I I I
CLARK CO., OH
(E)
1 1 1 1 1
I I I I I I I I I I I I I I I
24 2 4 6 8 10 12 14 16 18 20 22 24
|^ a.m »-}•« p.m. ^-|
HOUR OF DAY, LST
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
I I I I I I I I M I I I I I I I I I I I I I
- BAKERSFIELD, CA
24 2 4 6 8 10 12 14 16 18 20 22 24
L« a.m *~|^ p.m ^j
U.1U
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
I
I I
I I I M M I I I I I I I I I
Mil
ALTON, IL
—
—
—
—
—
—
5i
|
(D)
•— -:
| |
''~"I>
/ .•"""" """"'^
f • V.
t^ ^
^•-f ..-••"" ''--,
I I I I I I I I I I I I I I I 1
—
—
—
—
—
—
^*S"— "S3
MM
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
24 2 4 6 8 10 12 14 16 18 20 22 24
[^ a.m *.[*« p.m »~|
I I I I I I M I I I II I I II TT I
EDMOND, OK
(F)
I I
1 1 M T I I I I I I I I I I I I I I I I I I
24 2 4
u
6 8 10 12 14 16 18 20 22 24
—a.m. »>|«« p.m.
HOUR OF DAY, LST
Figure 6-27 (A-F). Quarterly composite diurnal patterns of ozone concentrations at
selected sites representing potential for exposure of major crops, 1981.
Source: U.S. Environmental Protection Agency, SAROAD data file for 1981
6-52
-------
0.15
7-hour
24-hour
JUL
AUG SEP
MONTH OF YEAR
OCT
Figure 6-28. Daily 7-hour and 24-hour average
ozone concentrations at a rural (NCLAN) site in
Argonne, IL, 1980.
Source: Kress and Miller (1983)
6-53
-------
at Argonne monitors ozone concentrations over a 7-hour daytime period approxi-
mating the period of peak photosynthesis in crops. The data in Figure 6-22
yield a 24-hour average ozone concentration of 0.026 ppm (Kress and Miller,
1983). The 7-hour average for the same 7-week period at the Argonne site is
0.042 ppm (Kress and Miller, 1983). The day-to-day variations in both the
7-hour and the 24-hour averages generally appear to be greater than the average
difference within a day for either the 7-hour or 24-hour periods (Figure 6-22).
The fluctuations in 1-hour values within a day or from day to day would be
larger than within-day or between-day variations in either the 7-hour or the
24-hour average. The 7-hour average will be higher than the 24-hour average
because the former excludes the low nighttime concentrations.
In Figure 6-29 (A-H), seasonal variations in ozone concentrations in 1981
are depicted using 1-month averages and the single 1-hour maximum concen-
tration within the month for eight sites across the nation (U.S. Environmental
Protection Agency, SAROAD data file). The data from most of these sites
exhibit the expected pattern of high ozone levels in the summer and low levels
in the winter. The seasonal rise and fall is not always a simple smooth
curve, however. Tampa, for example, shows a late spring maximum. Dallas data
also tend to be skewed toward higher spring concentrations. Averaging together
data for several years would give a smoother "characteristic" pattern but also
v/ould obscure the fact that local, and even national, weather in a particular
year plays at least as big a role in the formation of ozone as the regular
seasonal changes in the elevation of the sun and the resulting variations in
insolation. Because of seasonal changes in storm tracks from year to year,
the general weather conditions in a given year may be more favorable for Oo/O
formation than during the prior or following year. Thus, short-term concentra-
tion trends may not be indicative of real changes in air quality.
6.5.1.3 Weekday-Weekend Variations in Ozone Concentrations. Atmospheric
ozone concentrations represent the combined effects of emission sources and
meteorological conditions. The various sections of this document have been
based on the assumption that the ozone precursor sources were operating in a
generally steady state or at least on an average, repeatable diurnal cycle.
For the most part, urban source patterns of oxidant precursors appear reasonably
constant; however, in most urban areas there are decided changes that occur in
traffic and commercial patterns that are keyed to a weekday-weekend activity
cycle. The impacts of these changes have been observed in corresponding
changes in ozone concentration patterns.
0190DL/A 6-54 6/15//84
-------
0.3D
0.30
i
a
2 0.25
O
E 0.20
UJ
Z 0.15
O
U
UJ
§ 0.10
N
O
0.05
0
0.35
0.30
a
^ 0.25
O
HMB
£ 0.20
UJ
Z* 0.15
O
0
UJ
2 0.10
1
0.05
0
_
™*
-
—
—
—
—
^
_
—
TUCSON, AZ
(A)
D MAXIMUM
AVERAGE
m
*™^
i
ii
—
!*?
^1^^
!*!*?
yv!
—
—
;S5
w
^•^•*
W!
-
^•"^
1
—
~
-
r-
J FMAMJ JASOND
POMONA, CA
(C)
—
—
m
w
m
^
^
__
f¥¥
—
™
—
—
—
1
^_
_
—
—
—
U.JO
0.30
0.25
0.20
0.15
0.10
0.05
0
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0
LENNOX, CA
_ (B)
^
—
—
~n
N
j
_
*M
1
—
!£!
1
|
_^
I
-
1
—
^^
w
i
1
FMAMJ JAS
r~ir— i
—
O N D
DENVER, CO
_ (D)
—
^^
_
—
"D
_
^^
M
D
—
S
1
—
1
1
1
™
—
X;!
-
Pfltr
J FMAMJ JASOND
MONTH OF YEAR
JFMAMJJASOND
MONTH OF YEAR
Figure 6-29 (A-H). Seasonal variations in ozone concentrations as indicated by monthly
averages and the 1-hour maximum in each month at selected sites, 1981.
6-55
-------
U.O9
0.30
E
a
a
z- 0.25
g
< 0.20
h-
LU
1 0.15
0
0
Ul
Z 0.10
0
N
o
0.05
0
0.35
0.30
e
a
^ 0.25
Z
O
F
$ 0.20
WASHINGTON, DC
- (E) _
H MAXIMUM ~~
AVERAGE
__
J
F
M
MORRIS
—
m
mmm
1
$
—
1
—
T?
—
la"
AMJJASOND
CO., NJ
„ (G)
jit '
Z
2 °-'35f™
9 §
1 0.10 {-
0
N
o !
0.05 j—
o!
PPP
y
~
m
—
_
-*"
_
|fWI
w
—
!v"
!* *X
—
11
—
W! P?i|
U.JO
0.30
0.25
0.20
0.15
0.10
0.05
0
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0
—
—
-
-
—
TAMPA, FL
(F)
—
—
—
FW
I"W
1
—
i
m
•^
w»
—
-
PS
—
7^
—
!3
—
QD"
JFMAMJJASOND
—
—
_
•—
DALLAS, TX
(H)
—
I—I r— i —
—
7T!
—
m
m
,
i!
il
H?
-
"
^
1
—i —
iCT
JFMAMJJASOND
MONTH OF YEAR
JFMAMJJASOND
MONTH OF YEAR
Figure 6-29 (A-H) (continued). Seasonal variations in ozone concentrations as indicated by
monthly averages and the 1-hour maximum in each month at selected sites, 1981.
6-56
-------
In an analysis of data from the 1960s in the Los Angeles basin, Schuck
et al. (1966) noted a summertime shift in oxidant maximum values on the weekend
from the central basin commercial and urban areas to suburban and coastal
areas. These authors attributed this change to shifts in traffic pattern
to recreational travel on the weekends. Other early analyses revealed no
weekday/weekend difference in oxidant concentrations (as opposed to spatial
shifts) and apparently none in the Los Angeles Basin, based on the observation
that most alerts occurred on Friday and no alerts had ever occurred on Sunday
(Altshuller, 1975).
Figures 6-30 through 6-32 show average hourly ozone concentrations for
the summer months of July, August, and September, 1981, for Pomona and Lennox,
California, and Little Rock, Arkansas, in which the data have been separated into
Sundays and the other 6 weekdays. Data used in these figures are for 3 months
only, but the Sunday patterns differ somewhat from weekday patterns in each of
the months considered. For the most part, the Sunday daytime ozone concentra-
tions are higher than the corresponding weekday concentrations, although these
data are not definitive. In areas such as Little Rock, where even summer-
time ozone is close to typical background concentrations, as indicated by
Figure 6-32, day-of-the-week shifts are less pronounced than in California.
The concentration cycles for ozone that appear to be related to Sunday/
weekday changes are generally subtle, but they may have an influence on inter-
pretations of urban and suburban exposures and other effects data.
6.5.2 Spatial Variations in Ozone Concentrations
Ozone is commonly thought of as a regional pollutant. Data abound to
confirm that some urban regional airsheds have higher average ozone concentra-
tions than others. In addition, an examination of specific sites within an
airshed will also show that spatial variations in ozone concentrations occur
to cause differing microscale exposures both of human populations within the
same urban airshed and of crops and other vegetation in nonurban areas.
6.5.2.1 Urban versus Nonurban Variations. Data were presented in the 1978
document demonstrating that peak concentrations of ozone in rural areas are
generally lower than those in urban areas, but that "dosages or average concen-
trations in rural areas are comparable to or even higher than those in urban
areas" (U.S. Environmental Protection Agency, 1978). The diurnal concentration
0190DL/A 6-57 6/15//84
-------
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0
I I I I I I I I I I I I I
_ JUL
SUNDAY
I I I I U
•OTHER DAYSf
24
8 10 12 14 16 18 20 22
NOON
a.m. - »»<« - P-m.
I I I I I I I
I I I S I I I I
10 12 14 16 18 20 22
NOON
p.m.
I I I I I I I I I I I I I I I I I I I I I
I i I I I I
24 2 4 6 8 10 12 14 16 18 20 22
TIME OF DAY, LST
Figure 6-30. Composite diurnal data for Sunday versus
other 6 days for July through September 1981, Pomona,
CA.
Source: U.S. Environmental Protection Agency,
SAROAD data file for 1981
6-58
-------
u. 10
0.16
0.14
0.12
0.10
0.08
0.06
0.04
002
UiVA.
n
I I I I I I I I I I I
_ JUL.
—
—
—
— f
L-. -^i^
i i T-rTTi i i i i
I i I I I I I I I I
— •
_
OTHER DAYS
SUNDAY ~
—
_ —
N _
^^^^- ^ ***'*'* *^_^ ^
I I I I I I I I I I
24 2 4 6 8 10 12 14 16 18 20 22
E
CL
a
Z
O
H
CC
H
LLJ
O
O
O
LU
z
o
N
0
NOON
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0
I I I I I I I I I I I
I AUG
_
—
—
— — — —
~~ -^ -^"^
^ ^^^
^*>^r
W I I |LT I I I I I I I
I I I I I I I I I I
_
—
—
_ ^-— ""
^^ >^_ ^*^ «, .
k^
-s^-
i i i i i i i i i T^
24 2 4 6 8 10 12 14 16 18 20 22
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
n
NO
I I I I I I I I I I I
SEPT.
—
—
—
—
'-^I.
_ ^,
--t-k.J-'apS't— &-* — r\ i i
ON
I I I I I I I I I I
—
—
—
—
—
/•x"~"""~'»^ ""
^ ^"N^
^*^-=^-
l I I l l l I i FT —
24 2 4 6 8 10 12 14 16 18 20 22
NOON
TIME OF DAY, LST
Figure 6-31. Composite diurnal data for Sunday versus
other 6 days for July through September 1981, Lennox,
CA.
Source: U.S. Environmental Protection Agency, SAROAD
data file for 1981
6-59
-------
a.
a.
2
{£
Z
LU
O
z
O
O
iu
2
O
N
O
0.08
0.06
0.04
0.02
0
0.08
0.06
0.04
0.02
0
0.08
0.06
0.04
0.02
I I I I I I I I I 1 I I I I I I I I I I I I
. JUL.
OTHER DAYS /'
SUNDAYS '
I i I I I I i 1 I I I I 1 I I I I I I I I I
24 2 4 6 8 10 12 14 16 18 20 22 24
NOON
*— •— a.m.i »»-L4 p.m.i —
I I I I I I I ! I I I I I ill
AUG.
OTHER DAYS
. ---SUNDAYS
I I I i i I I i I I I I I I I I i I I I I
24 2 4 6 8 10 12 14 16 18 20 22 24
NOON
• a.m., »L« p.m «.
I I I I I I I I I I I I I I I I I I I I
OTHER DAYS
SUNDAYS ,'
s
i i i 5 ! I I I I I I I I i I I I i I I
I
24
6 8 10 12 14 16 18 20 22 24
NOON
. a.m.t »-!« p.m. • —
TIME OF DAY, LST
Figure 6-32. Composite diurnal data for Sunday versus
other 6 days for July through September 1981, Little
Rock, AR.
Source: U.S. Environmental Protection Agency, SAROAD
data file for 1981
6-60
-------
data presented in the preceding section indicate that peak ozone concentra-
tions can occur later in the day in rural areas than in urban, with the distance
downwind from urban centers generally determining in large measure how much
later the peaks occur. The data presented in the preceding section for Montague,
Massachusetts, in Figure 6-21 (Singh and Martinez, 1979) exemplify high late-
afternoon secondary peak concentrations resulting from transport.
The NAPBN and other nonurban data in section 6.4 illustrate the typical
urban-nonurban gradient that exists between peak ozone concentrations. While
corroboration of the statement in the 1978 document that dosages and average
concentrations in rural areas are higher than in urban areas would require
calculations of means or ppm-hours, Figure 6-27 supports that conclusion.
Ozone concentrations during nighttime and early-morning hours are lower in
Bakersfield and Sacramento, California, than at the other four sites. Both
areas represent crop-growing areas but both are essentially urban areas and
are affected by other urban areas upwind. The other four sites are either
suburban or rural. The lower nighttime and early-morning ozone concentrations
found in urban areas are typically explained as resulting from ozone scavenging
by reaction with nitric oxide.
Another consideration pertinent to potential exposures of crops in rural
areas is the well-documented fact that ozone persists longer in nonurban than
in urban areas (Coffey et al., 1977; Cleveland et a!., 1976; Wolff et al.,
1977; Isaksen et al., 1978). Again, the absence of chemical scavengers appears
to be the chief reason.
6.5.2.2 Intracity Variations. Despite relative intraregional homogeneity,
evidence exists for intracity variations in concentrations that are pertinent
to potential exposures of human populations and to assessing actual exposures
sustained in epidemiologic studies. Two illustrative pieces of data are
presented in this section, one a case of relative homogeneity in a city with a
population under 500,000 (New Haven, Connecticut) and one a case of relative
inhomogeneity of concentrations in a city of greater than 9 million population
(New York City) (U.S. Department of Commerce, 1982).
New Haven, Connecticut was the site of an epidemiological study in 1976
by Zagraniski et al. (1978). Symptoms recorded in subjects' diaries were
correlated with ozone concentrations measured by the chemiluminescence method
at a downtown New Haven site characterized as Center City-Residential.
Table 6-11 shows several percentiles in the distribution of hourly values for
0190DL/A 6-61 6/15//84
-------
TABLE 6-11. OZONE CONCENTRATIONS AT SITES IN AND AROUND
NEW HAVEN, CONNECTICUT, 1976
(CHEMILUMINESCENCE METHOD, HOURLY VALUES IN ppm)
Site (SAROAD No.)
New Haven, CT:
(070700123F01)
Derby, CT:
(070190123F01)
Harden, CT:
(070400001F01)
No.
Measurements
4119
5698
3853
% of values < stated concentration
50%
0.021
0.023
0.030
90%
0.035
0.038
0.045
95%
0.091
0.071
0.075
99%
0.162
0.095
0.098
Max Concn.
0.274
0.290
0.240
Source: U.S. Environmental Protection Agency, SAROAD data file for 1976.
0190DL/A
6-62
6/18//84
-------
that site plus two other sites in the county that were operating at the time,
one in Derby, Connecticut, 9 miles west of New Haven, and one in Hamden,
Connecticut, 6 miles north. The Derby site also is characterized as Center
City-Commercial, the Hamden site as Rural-Agricultural. The general similarity
of values among the three sites appears to substantiate the New Haven data
used in the epidemiological study since there was probably a reasonable temporal
correlation between these close sites. This shows that wherever study subjects
might have traveled about the county, they probably incurred similar exposures
to ambient ozone. This conclusion is reinforced by the data in Table 6-12,
showing the date and time of the maximum hourly concentrations by quarter at
these three sites. The significant data are those for the second and third
quarters when the potential for 03 formation and for exposure is the greatest.
Differences in peak concentrations varied from 0.006 ppm in the fourth quarter
to 0.055 ppm in the third quarter among sites.
The source of much of the ozone found in the New Haven, Connecticut, area
is the greater New York City area (e.g., Cleveland et al., 1976) and an urban
plume transported over considerable distance would tend to be relatively uniform
TABLE 6-12. QUARTERLY MAXIMUM 1-HOUR OZONE VALUES AT SITES
IN AND AROUND NEW HAVEN, CONNECTICUT, 1976
(CHEMILUMINESCENCE METHOD, HOURLY VALUES IN ppm)
New Haven^ CT
(No. measurements
Max 1-hr, ppm
(Hour of day)
Date
Derby, CT
No. measurements
Max 1-hr, ppm
(Hour of day)
Date
Hamden, CT
No. measurements
Max 1-hr, ppm
(Hour of day)
Date
1
10
0.045
11:00 a.m.
3/29
11
0.015
11:00 p.m.
3/31
56
0.050
Noon
3/29
Quarter of
2
1964
0.274
2:00 p.m.
6/24
2140
0.280
2:00 p.m.
6/24
2065
0.240
3:00 p.m.
6/24
Year
3
2079
0.235
2:00 p.m.
8/12
2187
0.290
2:00 p.m.
8/12
1446
0.240
1:00 p.m.
7/20
4
66
0.066
10:00 p.m.
10/3
1360
0.060
7:00 p.m.
12/20
286
0.065
3:00 p.m.
10/7
Source: U.S. Environmental Protection Agency, SAROAD data file for 1976.
0190DL/A 6-63 6/15//84
-------
The highest or second-highest 1-hour maximum ozone concentration reported
from a given station during a given year frequently gives an indication of the
potential for repeated human exposure to high ozone levels. Nevertheless, a
one-to-one correspondence between peak levels and either the number of days or
the number of hours that a given level may be exceeded does not necessarily
exist. Data obtained in the metropolitan New York area illustrate this latter
fact (Smith, 1981). Data for 1980 are given in Table 6-13. These data were
obtained at the monitoring sites shown in Figure 6-33. The second highest
1-hour ozone readings at the Eisenhower Park and Queens College stations have
values only a few percentage points apart, yet there were 51 hours of ozone
concentrations exceeding 0.12 ppm and 15 days when ozone levels exceeded 0.12
for at least 1 hour at the Queens College Station; whereas corresponding values
were recorded at Eisenhower Park for only 7 hours during 2 days. At both sta-
tions, data for about 94 percent of possible hours were recorded as valid. Thus,
the pattern of repeated peak exposures is different between these two stations,
a fact having likely significance for health and welfare effects.
The range of first-, second-, third-, and fourth-highest values, along
with frequencies of values >0.12 ppm, establishes an apparent concentration
gradient in the area from sites 6 and 5 to site 4. Exposure of human popula-
tions living and working in metropolitan New York City could differ appreciably
if the residences were located and all activities were centered in lower
Brooklyn as opposed to the upper Bronx. Differences in peak concentrations at
the respective sites varied by date (6/14 to 8/28); and by level on the same
day (8/28), when ozone was 0.080 ppm at site 4 and 0.174 at site 7, a differ-
ence of 0.094 ppm.
6.5.2.3 Indoor-Outdoor Ozone Concentration Ratios. It has long been realized
that most people in the United States spend a large proportion of their time
indoors. Essentially all air pollution monitoring, however, is done on outdoor
air. A knowledge of actual exposures of populations to ozone is essential for
Optimal interpretation and use of the results of epidemiological studies. The
modeling of actual exposures, as opposed to potential exposures, necessitates
knowing general activity patterns and at least approximate indoor/outdoor
ratios (I/O) of ozone.
For slowly reacting compounds such as carbon monoxide, and in the absence
of indoor sources, the long-term average ratios of the indoor to the outdoor
concentration tends to be close to unity, although over short time periods the
0190DL/A 6-64 6/15//84
-------
I/O may be significantly different because of non-equilibrium factors (Yocom,
1982). In contrast, the situation for reactive pollutants such as ozone is
much more complex, and reported I/O values for ozone are highly variable.
Unfortunately, the number of experiments and kinds of structures examined
TABLE 6-13. PEAK OZONE CONCENTRATIONS AT EIGHT SITES IN NEW YORK CITY
AND ADJACENT NASSAU COUNTY, 1980a
Site
Site no.
Susan Wagner H.S. 1
Mabel Dean H.S. 2
Woolsey Post Office 3
Mamaroneck 4
P.S. 321 5
Sheepshead Bay H.S. 6
Queens College 7
Eisenhower Park 8
No. 1-hr
averages
>0.12 ppm
20
19
37
0
24
44
51
7
Days
with 1-hr
averages
>0.12 ppm
8
10
6
0
9
12
15
2
Four highest daily
values, ppm, and date
1st
0.174
(8/28)
0.155
(7/21)
0.188
(7/20)
0.092
(6/14)
0.148
(7/26)
0.184
(7/31)
0.174
(8/28)
0.175
(8/28)
2nd
0.152
(7/18)
0.154
(7/26)
0.163
(7/21)
0.080
(8/28)
0.146
(8/28)
0.173
(7/18)
0.164
(7/21)
0.158
(7/21)
3rd
0.140
(7/26)
0.144
(7/18)
0.151
(7/22)
0.076
(7/2)
0.165
(7/18)
0.165
(8/7)
0.163
(6/14)
0.119
(7/20)
4th
0.131
(9/1)
0.139
(8/28)
0.148
(8/28)
0.075
(7/26)
0.145
(7/9)
0.164
(7/14)
0.159
(8/24)
0.118
(8/24)
aSites monitored during the Northeast Corridor Monitoring Program (NECRMP); site
numbers assigned here are keyed to Figure 6-33. For NECRMP site numbers, see
Smith (1980).
Source: Smith (1981)
0190DL/A
6-65
6/15//84
-------
SITES
1. Susan Wagner High School
2. Mabel Dean High School
3. Woolsey Post Office (Astoria)
4. Mamaroneck
5. Public School 321
6. Sheepshead Bay High School
7. Queens College
8. Eisenhower Park (Nassau Co.)
NEW JERSEY
Figure 6-33. New York State air monitoring sites for Northeast
Corridor Monitoring Program (NECRMP).
Source: Smith (1981)
6-66
-------
provide only limited data for use in modeling indoor exposures. Yocom (1982)
has presented a chronological summary of studies in which either ozone or
photochemical oxidant indoor-outdoor gradients were measured. Studies have
been conducted over the period 1971 through the present (one ongoing study) by
five research organizations: University of California, the California Insti-
tute of Technology, GEOMET, Inc., Lawrence Berkeley Laboratory, and TRC Environ-
mental Consultants. Structures examined have included hospitals, schools,
office buildings, single-dwelling homes, "experimental" dwellings, apartments,
and mobile homes. Private homes included those with and without gas stoves
and fireplaces, and those inhabited by smokers versus nonsmokers. Areas of
the country in which the buildings were located ranged from Southern California
to Boston, including as well, Denver, Chicago, Washington, D.C., Baltimore,
Pittsburgh, and other unspecified locations (Yocom, 1982).
The results of a number of the studies conducted to determine I/O for
ozone in a variety of building types are shown in Table 6-14. These results
are highly variable, to say the least. The variability is not surprising,
considering the diversity of structures and locations included in the studies.
In this tabulation the highest I/O value of 0.80 was reported by Sabersky
et al. (1973) on the basis of smog-season measurements in a multistory, air-
conditioned building on the Pasadena campus of the California Institute of
Technology. Air exchange in this building was at a rate of 10 changes per
hour with 100 percent outside air (i.e., no recirculation of inside air). For
another Cal Tech building, in which there was a mix of 70 percent outside air
and 30 percent recirculated inside air, Sabersky et al. (1973) found an indoor-
outdoor ozone ratio of 0.65. The lowest indoor-outdoor ozone concentration
ratios shown in Table 6-14 are those reported by Berk et al. (1981), which
were in the range of 0.10 to 0.25. These data were the result of studies in
energy-efficient housing, where ventilation was restricted in various ways for
energy conservation. These experiments were carried out in Medford, Oregon.
The research of Moschandreas et al. (1978, 1981) was carried out on east-coast
residences and those results were also highly variable.
A relatively large number of factors can affect the difference in ozone
concentrations between the inside of a structure and the outside air. In
general, outside air infiltration or exchange rates, interior air circulation
rates, and interior surface composition (e.g., rugs, draperies, furniture)
affect the balance between replenishment and decomposition of ozone within
buildings (Thompson et al., 1973; Sabersky et al., 1973; Berk et al., 1980;
0190DL/A 6-67 6/15//84
-------
TABLE 6-14. SUMMARY OF REPORTED INDOOR-OUTDOOR OZONE RATIOS
Structure
Indoor-outdoor
ratio (I/O)
Reference
Residence
(with evaporative cooler)
Office
(air-conditioned; 100% outside
air intake)
(air-conditioned; 70% outside
air intake)
0.60*
0.80 + 0.10
0.65 + 0.10
Thompson et al. (1973)
Sabersky et al. (1973)
Residence
Residence
Residence
(gas stoves)
(all electric)
Office
School room
Residence
0.70
0.50 to 0.70
0.19
0.20
0.29
0.19 (max)
0.10 to 0.25
Moschandreas et al.
Moschandreas et al.
Berk et al. (1980)
Berk et al. (1981)
(1978)
(1981)
aMeasured as total oxidants.
0190DL/A
6-68
6/15//84
-------
Moschandreas et al., 1978). The rate at which exterior air enters a building
depends on local wind speed and direction, on how well-sealed the building is,
on how frequently doors and windows are opened, and on the operating character-
istics and cycles of heating/air conditioning/ventilating systems. A signifi-
cant factor that increases infiltration is an increasing temperature differen-
tial between warm interior air and cold outside air (Moschandreas et al.,
1978), although such a differential would be unusual for a photochemical smog
period. Moschandreas et al. (1978) reported exterior-interior exchange rates
ranging from ten changes per hour in an office building to one change every
5 hours in a residence. At the higher exchange rates, inducted ozone remains
at a level indoors that is closer to the outdoor level. As the exchange rate
decreases, surface decomposition processes can result in progressively lower
equilibrium ozone concentrations. Other factors, such as relative humidity,
also affect decomposition. The half-life of ozone inside residences has been
estimated at 2 to 6 minutes (Moschandreas et al., 1978; Mueller et al., 1973;
Sabersky et al., 1973), while its half-life in an office environment has been
estimated at 11 minutes (Mueller et al., 1973). These results are indicative
of the relatively rapid reaction rate that can be expected for ozone in a
building or room environment. The problem of I/O values in buildings was the
subject of a model development program by Shair and Heitner (1974) in which
they tried to account for ventilation and for losses by reactions and surface
scavenging. Considering the research results shown in Table 6-13 and summarized
by Yocom (1982), any estimates of indoor ozone exposures to occupants must be
considered as having a large degree of probable variability.
At present there are no long-term monitoring data on indoor air pollutant
concentrations that are comparable to the concentration or exposure patterns
that are available for outdoor locations. Thus, for estimates of the exposure
of building occupants to ozone and other photochemical oxidants, it is necessary
to rely on extrapolations of very limited I/O data such as those shown in
Table 6-13.
6.5.2.4 Macroscale Variations in Ozone Concentrations: Effects of Altitude
and Latitude. The 1978 criteria document presented discussions on the effects
of tropopause-folding events (TF), and of the seasonal tropopause adjustment
(STA) and small-scale eddy transport (SSET) mechanisms on stratospheric-
tropospheric exchange. As described previously in chapter 4, TF events would
be expected to produce rare increases in ground-level ozone concentrations,
0190DL/A 6-69 6/15//84
-------
resulting from strong incursions of stratospheric ozone, in the southern and
eastern United States (latitudes of about 37°N and less and longitudes of
about 90° and less) (U.S. Environmental Protection Agency, 1978). The STA and
TF mechanisms result in stratospheric-tropospheric interchange in the same
season (winter or winter-to-spring) and in the same latitudes.
Concentrations of ozone can be expected to vary with altitude and with
latitude. These variations occur because of the interchange mechanisms involved
in stratospheric-tropospheric exchange, the decay of stratospheric ozone as it
traverses the troposphere, and the known production in the apparently unpolluted
troposphere of ozone at certain altitude ranges above mean sea level (MSL).
Seiler and Fishman (1981) reported on ozone measurements taken on flights in
remote tropospheric air during July and August 1974. Averages of their data
show that ozone concentrations increase with increasing altitude and in general
substantiate the accepted belief that the earth's surface and the lower atmos-
phere act as ozone sinks. The value for the average ozone concentration in
the troposphere is about 0.035 ppm + 30 percent. Tropospheric ozone shows a
marked hemispheric asymmetry; the higher concentrations occur in the northern
hemisphere.
Logan et al. (1981) summarized earlier measurements of background tropos-
pheric ozone. Hemispheric asymmetry is readily apparent in their results, as
is the seasonal increase in lower tropospheric ozone in the summer at mid-
latitudes in the northern hemisphere.
6.5.2.5 Microscale Variations in Ozone Concentrations: Effects of Monitor
Placement. Just as macroscale variations in ozone concentrations have been
observed by measuring vertical and horizontal profiles at various altitudes,
so too have microscale variations been observed as a function of placement of
sampling probes.
Data drawn from a recent study on rural ozone concentrations illustrate
the possible effects of sampling probe location on the resulting concentration
data. The data given here illustrate lesser-known effects of placement of
monitoring probes that are pertinent for vegetation studies, in particular.
Pratt et al. (1983) studied concentrations of ozone and oxides of nitrogen
in the upper-midwestern part of the United States. Concentration data were
obtained over 4 years by means of monitors at two sampling heights (ca. 3 and
9 meters) at three air quality monitoring sites: LaMoure County, North Dakota;
Traverse County, Minnesota; and Wright County, Minnesota. All stations were
0190DL/A 6-70 6/15//84
-------
rural sites. The mean 0- concentrations did not differ greatly among the sites,
but in at least some instances the mean differences between sampling heights were
as large or larger than the differences among the scattered sites. Table 6-15
presents the mean ozone concentrations measured at two separate sampling heights
(Pratt et al., 1983). Annual average concentrations were 1 to 3 ppb lower at the
3.05-meter height than at the 6.10 and 9.14-meter heights, reflecting the
depletion of ozone near the surface. As might be expected, the gradient was
especially conspicuous at night because of the continued surface scavenging
and a decrease in the rate of transfer from layers aloft. The concentrations
of ozone occuring at these sites were near background in all years measured.
In areas with higher ozone concentrations, one would expect to see larger
absolute gradients between monitors at different heights. In fact, the careful
measurement of concentration gradients over distances of 1 to 10 meters is a
recognized method for estimating the scavenging potential of a surface.
6.6 CONCENTRATIONS OF PEROXYACETYL NITRATE (PAN) AND PEROXYPROPIONYL NITRATE
(PPN) IN AMBIENT AIR
6.6.1 Introduction
As noted in the introduction to this chapter (section 6.1), published data
on the concentrations in ambient air of photochemical oxidants other than ozone
are not comprehensive or abundant. Much more is known now, however, about their
atmospheric concentrations than was known when the 1978 criteria document for
ozone and other photochemical oxidants was published. Review of the data that
follow will show that peroxyacetyl nitrate (PAN), peroxypropionyl nitrate (PPN),
and hydrogen peroxide (H?0?) are the most abundant of the non-ozone oxidants in
ambient air in the United States other than the inorganic nitrogenous oxidants
such as nitrogen dioxide (N0?) and possibly nitric acid (HNO»), in some areas.
The inorganic nitrogenous oxidants found in ambient air are reviewed in air
quality criteria documents on the nitrogen oxides and are not treated in this
document except for the review of the role of nitric oxide (NO) and N02 in
atmospheric photochemistry.
In this section, the concentrations of PAN and PPN in ambient air will be
presented in order to examine potential exposures to these oxidants of human
populations, vegetation, and terrestrial ecosystems.
0190DL/A 6-71 6/15//84
-------
cr>
i
TABLE 6-15. MEANS AND STANDARD ERRORS OF OZONE CONCENTRATIONS MEASURED OVER 4 YEARS
AT TWO SAMPLING HEIGHTS AT THREE STATIONS IN THE RURAL, UPPER-MIDWESTERN UNITED STATES
(ppb/v/v)
Site
LaMoure
County, ND
Traverse
County, MN
Wright
County, MN
Sampling
height, m
3.05
9.14
3.05
9.14
3.05
6.10
1977
22.97 + 0.20
(6.413)3
24.14 + 0.19
(6,410)
24.44 + 0.19
(9,672)
26.29 + 0.19
(9,810)
-
1978
35.30 +• 0.20
(15,218)
35.67 + 0.19
(15,220)
35.2 + 0.19
(22,675)
36.77 + 0.19
(22,624)
34.64 + 0.23
(17,437)
35.61 + 0.23
(17,440)
1979
30.80 + 0.16
(21,064)
32.25 + 0.15
(20,470)
30.64 + 0.17
(19,900)
32.39 + 0.16
(19,289)
28.71 + 0.18
(17,771)
29.27 + 0.18
(17,775)
Years
1980
34.98 +0.25
(17,157)
37.53 + 0.25
(17,157)
. 34.66 +0.23
(22,629)
37.60 + 0.23
(22,625)
34.27 + 0.26
(18,222)
35.16 + 0.27
(18,280)
1981
31.92 + 0.26
(6,699)
34.23 + 0.26
(6,364)
28.78 + 0.23
(7,141)
31.08 + 0.22
(7,142)
31.60 + 0.24
(7,764)
32.28 + 0.24
(7,766)
1977-1981
' 32.26
(66,551)
33.82
65,597
32.09
(82,017)
34.21
(81,550)
32.42
(61,254)
33.21
(61,261)
aThe numbers in parentheses refer to the number of hours of monitoring included in the reported values. Values are based on
all valid data per site. For each sampling height at each site, values for three monitors separated by 76 m are included
in the calculations. Monitoring was conducted only during the second half of 1977 and only until 30 June in 1981.
Source: Pratt et al. (1983)
-------
No data indicating potentially adverse effects of PPN are available. It
exists in trace quantities and apparently no research has ever been undertaken
to determine potential toxicity. At least one study (Heuss and Glasson, 1968)
has reported that peroxybenzoyl nitrate (PBzN), like PAN, is a lachrymator.
No unambiguous identification of PBzN in the ambient air of the United States
has been made, however.
A number of facts about the toxicity and behavior of PAN are available,
however, and are pertinent to the examination of its concentrations in ambient
air:
On a concentration basis, PAN appears to be more phytotoxic
than ozone (chapter 7). Although data on the health effects of
PAN are sparse, PAN appears to be less toxic than ozone in
animals and man (chapters 10 and 11).
As with ozone, PAN apparently has to enter the leaf of the
plant to exert its toxic effects (chapter 7). Thus, to manifest
its toxicity in plants, PAN has to be present when the stomata
of the leaves are open, which is thought to be limited to the
daylight photoperiod. Important in this context is the fact
that no visible injury to plants from PAN will occur unless
light is present before, during, and after exposure to PAN.
This not true for ozone injury. Like ozone, however, PAN is
more toxic to herbaceous (e.g., crops) than to woody vegetation
(e.g., shrubs) (chapter 7).
Although PAN and ozone are individually toxic to man, animals,
and vegetation, data on possible interactive effects are quite
limited. In vegetation studies, higher concentrations of the
combined pollutants produce effects that are less than additive;
that is, the pollutants appear to be antagonistic (chapter 7).
Adsorption of PAN on reactive surfaces differs from that of
ozone. Deposition velocities for PAN on vegetation may be as
much as three times lower than for ozone (Hill and Chamberlain,
1976; in McMahon and Denison, 1979). Deposition velocities for
soils may be greater for PAN, but velocities depend on type of
soil and amount of soil water. More PAN is taken up by moist
soil than by dry (Garland and Penkett, 1976), but more ozone is
taken up by dry soil than moist (Garland, 1977).
No indoor PAN concentrations appear in the recent literature,
but given the known deposition velocities for the respective
pollutants, indoor/outdoor ratios for PAN may be higher than
those of ozone. The persistence of PAN is thermally dependent,
which also would influence indoor concentrations.
019SP/A
6-73
6/18/84
-------
6. Nitric oxide (NO) does not scavenge PAN as effectively as it
does ozone (U.S. Environmental Protection Agency, 1978); but NO
may serve as a PAN scavenger when NO is at higher concentrations
in the evening.
7. No damage to nonbiological materials has ever been attributed
to PAN, singly or in combination with other pollutants. Apparent-
ly no research has been done in this area.
Given the above information about PAN, it is apparent that the concentra-
tions of PAN that are of most concern are those in nonurban areas during day-
light hours, relative to vegetation, and those both indoors and outdoors at
all times in both urban and nonurban areas, relative to human populations.
Since concentrations of precursors to PAN are lower in nonurban areas, whether
PAN can, like ozone, be transported from urban areas is a major determinant of
the levels likely to be found in nonurban (agricultural) areas. Most of the
available data on concentrations of PAN in ambient air are from urban areas.
This section presents historical and recent data on the concentrations of PAN
and, where available, on PPN and on the patterns those concentrations assume.
6.6.2 Historical Data
In the 1970 criteria document for photochemical oxidants (U.S. Department
of Health, Education, and Welfare, 1970), concentrations for total oxidants
and PAN were reported for Los Angeles and Riverside, California. In Los
Angeles, composite diurnal data for PAN (obtained by GC-ECD) showed average
peak 1-hr concentrations of about 40 ppb in September 1965 and about 60 ppb in
October 1965. The September peaks occurred around noon and the October peaks
occurred shortly after 1:00 p.m., with the difference in time possibly being a
function of the temperature-dependence of PAN formation and persistence. In
Riverside, two peak concentrations were observed in composite diurnal data,
one of > 4 ppb around 10:00 a.m. and one ~1Q ppb between 4:00 and 6:00 p.m.
Seasonal data from Riverside for June 1966 to June 1967 showed that PAN concen-
trations were highest in September 1966 and in March and June 1967.
Total oxidants (by Mast meter) in these Los Angeles sites reached a peak
concentration (the same composite diurnal data as above) of close to 140 ppb
in September 1965 at about the same hour of day as the PAN peak. In October,
total oxidants peaked at nearly 200 ppb, again coinciding in time with peak
PAN concentrations. In Riverside, the morning PAN peak preceded the ozone
019SP/A 6-74 6/18/84
-------
peak (~110 ppb around noon) by almost 2 hours but the afternoon PAN peak
trailed the afternoon ozone peak (~160 ppb around 2:00 to 4:00 p.m.) by about
2 hours. The ozone/PAN ratio was lower from October until March than during
the rest of the year.
In the 1978 criteria document for ozone and other photochemical oxidants
(U.S. Environmental Protection Agency, 1978), additional PAN data for Los
Angeles as well as two other cities were presented. Lonneman et al. (1976)
measured PAN and total oxidants in Los Angeles in 1968. In 118 samples col-
lected for the period 10:00 a.m. to 4:00 p.m. during the study, the median PAN
concentration was 13 ppb and the average PAN concentration was 18 ppb. The
median total oxidant concentration (measured by UKI) was 97 ppb and the average
was 117 ppb. Thus, the median oxidant/PAN ratio was 7.5 and the average oxidant/
PAN ratio was 6.5.
Lonneman et al. (1976) conducted similar studies in Hoboken, New Jersey,
in 1970 and in St. Louis, Missouri, in 1973. Samples were measured during the
period 10:00 a.m. to 4:00 p.m. over the course of the study. In Hoboken, PAN
concentrations averaged 3.7 ppb. Ozone concentrations (measured by chemilumi-
nescence) averaged 90.5 ppb. In St. Louis, PAN averaged 6.4 ppb, ozone (meas-
ured by chemiluminescence) averaged 50.1 ppb, and total oxidants (by UKI)
averaged 74.3 ppb.
From these 1966 through 1973 urban data, it is clear that PAN concentrations
in urban areas are appreciably lower than those of ozone, even in the winter
in California, when PAN concentrations are proportionally higher than those of
ozone.
In addition to urban data, the 1978 criteria document (U.S. Environmental
Protection Agency, 1978) also included PAN concentrations from one nonurban-
agricultural area, Wilmington, Ohio (Lonneman et al., 1976). The maximum PAN
concentration observed in 1500 samples taken during August 1974 was 4.1 ppb.
The daily maximum PAN concentration rarely exceeded 3.0 ppb even though the
daily maximum ozone concentration frequently exceeded 80 ppb.
Conclusions reached in the 1978 document were (1) that PAN concentrations
are much lower in the ambient air of nonurban than of urban areas; and (2)
that ozone/PAN or total oxidant/PAN ratios vary with location, such ratios
being higher in nonurban areas than in urban (U.S. Environmental Protection
Agency, 1978). From data presented in the 1970 document, it may be concluded
(1) that oxidant/PAN ratios vary with season; (2) that PAN concentrations are
019SP/A 6-75 6/18/84
-------
lower than ozone concentrations in urban areas; and (3) that PAN and ozone
concentrations exhibit similar but not identical diurnal patterns.
The historical data presented above have been given in some detail because
the information about PAN conveyed by those data remains valid. Examination
of the more recent data presented in subsequent sections shows that the newer
data, for the most part, extend and corroborate the findings of the older
literature.
6.6.3 Ambient Air Concentrations of PAN and Its Homologues in Urban Areas
Additional studies on concentrations of PAN and its homologues in both
urban and nonurban areas are now available. Newer data have been summarized
here, where possible, and the individual studies or examples presented are
merely a few of many, chosen to represent current knowledge regarding ambient
air PAN concentrations and their patterns.
The existing literature on the concentrations in ambient air of the per-
oxyacyl nitrates, PAN and its higher homologues, has been compiled and examined
in two recent review articles (Temple and Taylor, 1983; Altshuller, 1983). In
the first, Temple and Taylor reviewed the concentrations of PAN in the ambient
air in Europe, Japan, and North America in the context of the phytotoxicity of
PAN. Altshuller, in the second, reviewed the published concentrations of PAN and
of PPN in ambient air, also within and outside the United States. In addition,
Altshuller analyzed the relationships to ozone of PAN and other photochemical
reaction products. The reader is referred to these reviews for detailed infor-
ration and for references therein. The review by Altshuller (1983) is especially
comprehensive.
Table 6-16 presents a summary of PAN concentrations observed in the
ambient air of urban areas of the United States. Data in this table include
the results of studies cited in section 6.6.2 as well as the results of newer
studies. This table was derived from the reviews of Altshuller (1983) and
Temple and Taylor (1983), as well as from a few additional sources (Jorgen et
al., 1978; Lewis et al., 1983). The data are summarized in the table by
region of the United States and by date, with the newer studies reported first
for each region.
Because of variations in diurnal patterns of PAN by location and season,
and because no national, uniform aerometric data base for PAN exists, few of
the data reported in Table 6-16 are really comparable. Thus, data in this
019SP/A 6-76 6/18/84
-------
TABLE 6-16. SUMMARY OF CONCENTRATIONS OF PEROXYACETYL NITRATE IN AMBIENT AIR IN URBAN AREAS OF THE UNITED STATES
Site
West
Riverside, CA
W. Los Angeles, CA
Claremont, CA
Claremont, CA
Claremont, CA
CTl
1
--j Riverside, CA
Riverside, CA
Riverside, CA
Riverside, CA
West Covina, CA
West Covina, CA
Pasadena, CA
Riverside, CA
Downtown
Los Angeles, CA
Time
of yr
All
June
Sept
Aug.
Oct.
Apr.
July
Oct
year
.-Oct.
-Sept.
, May
, Aug. ,
All year
Oct.
Aug.
Aug.
July
, Sept.
, Sept.
Aug. -Apr.
Sept
Nov.
.-
PAN concentrations, ppb ,
Yr
1980
1980
1980
1979
1978
1977
1977
1975 (May)-
1976 (Oct. )
1976
1977
1973
1973
1967-1968
1968
Hours
sampled
8 a.m. -8 p.m.
6:35 a.m. -1:35 p.m.
24 hr/day
Morning to late
evening
Late morning to
late evening
24 hr/day
Late morning to
evening
24 hr/day
Late morning to
early evening
23 hr/day
NAd
7 a.m. -4: 30 p.m.
24 hr/day
10 a.m. -4 p.m.
No. days
sampled
365
2
11
8
5
10
10
. 533 '
3
24
NA
3
273
-
Method3 Avg.
GC-ECD -C
LP-FTIR 7
GC-ECD 13
GC-ECD 4
LP-FTIR 6
GC-ECD 1.6
LP-FTIR 7
GC-ECD 3.6
LP-FTIR 9
GC-ECD 9
NA
LP-FTIR 30
GC-ECD
GC-ECD 8
b Monthly
Max. mean
41.6 4.9
16
47
11
11
5.7
18
32
18
20
46 8.8
53
58 4.6
68
D
Original reference
Temple and Taylor (1983)
Hanst et al .
Grosjean and
Tuazon et al .
Tuazon et al .
1981b)
Singh et al.
Tuazon et al .
(1982)
Kok (1981)
(1981a)
(1981a;
(1979)
(1980)
Pitts and Grosjean (1979)
Tuazon et al .
Spicer (1977)
Spicer (1974)
Hanst et al .
Taylor (1969)
(1978)
(1975)
Lonneman et al. (1976)
Source
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
Ref.
1
2
2
2
2
2
2
2
2
2
1
2
1
2
-------
TABLE 6-16. SUMMARY OF CONCENTRATIONS OF PEROXYACETYL NITRATE IN AMBIENT AIR IN URBAN AREAS OF THE UNITED STATES (continued)
PAN concentrations,
Site
Salt Lake
City, UT
Los Angeles, CA
Downtown
Los Angeles, CA
Southwest
Houston, TX
Houston, TX
en
^j Midwest
00 Dayton, OH
(Huber Heights, OH)
St. Louis, MO
St. Louis, MO
East
New Brunswick, NJ
New Brunswick, NJe
Hoboken, NJ
Time
of yr
July-
Sept.
Sept.-
Oct.
July-
Oct.
Oct.
July
July,
Aug.
June-
Aug.
Aug.
All year
Sept.-
Dec.
June, July
Yr
1966
1965
1960
1977
1976
1974
1973
1973
1978 (Sept.)
-1980 (May)
1978
1970
Hours No. days
sampled sampled
7 a.m. -3 p.m.
8 a.m.-l p.m. 35
9 a.m. -Noon 9
8 a.m. -8 p.m.
24 hr/day 22
24 hr/day 20
23 hr/day 26
8 to 24 hr/day 12
24 hr/day 400
(9600 1-hr values)
8 a.m. -6 p.m.
10 a.m. -4 p.m.
Method3
GC-ECD
GC-ECD
IR
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
Avg. Max.b
54
31 214
-20 70
15.6
0.4 11.5
0.7 10
1.8 19
6.3 25
0.5 10.6
10.5
9.9
ppb b
Monthly
mean Original reference
4.4 Tingey and Hill (1968)
38 (Sept. ) Mayrsohn and Brooks
40 (Oct.) (1965)
Renzetti and
Bryan (1961)
Jorgen et al . (1978)
Westberg et al . (1978)
Spicer et al . (1976)
Spicer (1974)
Lonneman et al . (1976)
Lewis et al . (1983)
2.7 Brennan (1979)
3.7 Lonneman et al. (1976)
Source
Ref. 1
Ref 1;
ref. 2
Ref. 2
Jorgen et
al. (1978)
Ref. 2
Ref. 1
Ref. 1;
ref. 2
Ref. 1
Lewis et al
(1983)
Ref. 2
Ref. 2
Reference 1 is Altshuller (1983); reference 2 is Temple and Taylor (1983).
cRef. 1 reported averages for the sampling period; reference 2 reported monthly means.
dNot available.
eSubset of data reported by Lewis et al. (1983), cited in table above.
-------
table lend themselves to general conclusions but not to the analysis of trends
over the past decade and a half or necessarily to analysis of between-city or
between-region similarities or differences. Concentrations of PAN in Los
Angeles in 1965 appear to have been a great deal higher (Mayrsohn and Brooks,
1965) than in 1980 (Hanst et al., 1982) for nearly the same time of day, but
the 1965 concentrations were measured in September and October and the 1980
concentrations were measured in June. Other data from California indicate
that September and October are more likely to be part of the smog season there
than June. A comparison of PAN concentrations in Riverside in 1980 with those
in 1967 and 1968 indicates little difference. Again, however, the sampling
periods were not identical relative to averaging time or time of year.
In their review, Temple and Taylor (1983) have presented monthly average
PAN concentrations for 1967-1968 and for 1980. These data are plotted in
Figure 6-34, derived from their tabular data. More similarities than differ-
ences are apparent.
In addition to compiling existing data on the concentrations of PAN in
ambient air, Altshuller (1983) also did the more difficult task of relating
PAN concentrations to ozone concentrations where data for both exist. It must
be borne in mind that sampling periods (years, time of year, number of meas-
urements) are not identical and in many cases are not even similar (see Table
6-16). Neither are the averaging times over which samples were collected and
calculated within respective studies identical or even necessarily similar.
Nevertheless, the data of Altshuller constitute a comprehensive review and
examination of the relationships among respective photochemical oxidants in
urban areas of the United States. For ease of presentation, PAN/03 ratios,
expressed as percentages, are given in Table 6-17, but Table 6-16 should be
consulted for information on sampling periods and averaging times.
The existence of peroxybenzoyl nitrate (PbzN) in ambient air of urban
areas was postulated in the 1978 criteria document for ozone and other oxidants
(U.S. Environmental Protection Agency, 1978). Although it has been reported
to occur in ambient air in Europe, PBzN has not been clearly identified in
ambient air in the United States. Hanst et al. (1982) estimated that 2 ppb
PBzN would be clearly discernible in FTIR measurements but reported no clear
absorption band for PBzN in their measurements during a smoggy period in
Los Angeles. They estimated an upper limit of 1 ppb PBzN during their 1980
study (the maximum ozone concentration was 272 ppb and the maximum PAN concen-
tration was 16 ppb during that period).
019SP/A 6-79 6/18/84
-------
0.07
0.06
I 0.05
a
0.04
^ 0.03
O
O
z
< 0.02
0.01
IIIIHIll 1967-1968
I I 1980
AUG SEPT OCT NOV DEC JAN FEB MAR APR
Figure 6-34. Comparison of monthly daylight average and
maximum PAN concentrations at Riverside, CA, for
1967-1968 and 1980.
Source: Derived from Temple and Taylor (1983)
6-80
-------
TABLE 6-17. RELATIONSHIP OF OZONE AND PEROXYACETYL NITRATE AT
URBAN AND SUBURBAN SITES IN THE UNITED STATES
Site/year
CALIFORNIA
Downtown Los Angeles, 1960
Downtown Los Angeles, 1965
Downtown Los Angeles, 1968
West Los Angeles, 1980
Pasadena, 1973
West Covina, 1977
Claremont, 1978
Claremont, 1979
Riverside, 1967-1968
Riverside, 1975-1976
Riverside, 1976
Riverside, 1977
Riverside, 1977
SOUTHWEST
Houston, TX, 1976
MIDWEST
St. Louis, MO, 1973
St. Louis, MO, 1973
Dayton, OH, 1974
(Huber Hts, OH)
EAST
Hoboken, NJ, 1970
New Brunswick, NJ 1978-1980
Avg.
8
NA
13
9
10
20
7
4
8
9
5
4
4
3
13
5
2
4
4
PAN/03, %
At 03 peak
7
7
NAa
6
8
12
6
4
NA3
5
4
4
NA3
3
NAa
5
1
NAa
2
Reference
Renzetti and Bryan (1961)
Mayrsohn and Brooks (1965)
Lonneman et al. (1976)
Hanst et al. (1982)
Hanst et al . (1975)
Spicer (1977)
Tuazon et al . (1981a, 1981b)
Tuazon et al . (1981a)
Taylor (1969)
Pitts and Grosjean (1979)
Tuazon et al . (1978)
Tuazon et al . (1980)
Singh et al. (1979)
Westberg et al. (1978)
Lonneman et al. (1976)
Spicer (1977)
Spicer et al . (1976)
Lonneman et al. (1976)
Brennan (1980)
Not available.
Source: Adapted from Altshuller (1983).
019SP/A
6-81
6/18/84
-------
The only homologue of PAN that has been unambiguously identified in
ambient air in the United States is peroxypropionyl nitrate (PPN). In his
review of existing literature, Altshuller (1983) compiled data on the concen-
trations of PPN in ambient air in urban areas. In addition, he calculated
ratios of the concentrations of PPN and PAN. His data, modified to express
ratios as percentages [(PPN/PAN) x 100], are presented as Table 6-18 (Altshuller,
1983).
As Altshuller has pointed out, average PPN concentrations are 10 to 30
percent of the average PAN concentrations in Table 6-18 except for San Jose
(8 percent) and Oakland (42 percent). The maximum PPN concentrations reported
are highly variable, however, ranging from 0.13 ppb in San Jose (August 1978)
to 6.0 ppb in Riverside (month and year as well as sampling period of day un-
known). Thus, the PPN/PAN ratio at maximum concentrations of PPN is highly
variable, as well. Among more recent data, the maximum PPN concentration was
5.0 ppb in St. Louis in August 1973. (Note, however, that the sampling period
in St. Louis was 10:00 a.m. to 3:30 p.m. Depending upon temperature and concen-
trations of precursors, a true PPN maximum may or may not have occurred by
3:30 p.m.)
A comparison of data in Table 6-19 obtained by the same investigators
(Singh et al., 1981) for three separate cities (Los Angeles, Oakland, and
Phoenix) helps demonstrate the variability of PPN and PAN concentrations
between locations. Table 6-19 is derived from Singh et al. (1981) and presents
PAN and PPN concentrations for the three cities, as well as PPN/PAN ratios (in
percent) calculated from their data.
6.6.4 Ambient Air Concentrations of PAN and Its Homologues in Nonurban Areas
Data on the concentrations of PAN and PPN in agricultural and other non-
urban areas of the United States are sparse. They include the measurements
done by Lonneman et al. (1976) in Wilmington, Ohio, in 1974, and by Westberg
et al. (1978) in Downington, Pennsylvania, in 1979, two agriculturally-oriented
areas. In the study by Lonneman et al. (1976), cited earlier in section
6.6.1, measurements made by GC-ECD from 10:00 a.m. to 4:00 p.m., local time,
showed a maximum concentration for the study period of 4.1 ppb. The average
daily maximum was 2.0 ppb. Westberg et al. (1978) measured PAN from 8:00 a.m.
to 6:00 p.m. and found a maximum concentration for the study period of 5.0
ppb; the average daily maximum was 2.2 ppb and the average for the entire
study period was <1 ppb. While the 4:00 p.m. cutoff used by Lonneman et al.
019SP/A 6-82 6/18/84
-------
TABLE 6-18. AMBIENT AIR MEASUREMENTS OF PEROXYPROPIONYL NITRATE CONCENTRATIONS
BY ELECTRON CAPTURE GAS CHROMATOGRAPHY AT URBAN SITES IN THE UNITED STATES
Site
Los Angeles, CA
Riverside, CA
Riverside, CA
Riverside, CA
San Jose, CA
Oakland, CA
Phoenix, AZ
en
i
00
w Denver, CO
Houston, TX
St. Louis, MO
St. Louis, MO
Chicago, IL
Pittsburgh, PA
Staten Island, NY
Period of
measurement/
no. of days (n)
April 1979 (13)
NA3 (1)
April -May
July 1980 (13)
August 1978 (7)
June-July 1979
(13)
April -May 1979
(14)
June 1980 (14)
May 1980 (12)
10 Aug. 1973
(1)
May- June, 1980
(9)
April -May, 1981
(13)
April 1981 (11)
March- April
1981 (11)
Period
of day
24
NAa
24
24
24
24
24
24
24
100
(LT
24
24
24
24
hr
hr
hr
hr
hr
hr
hr
hr
0-1530
B)
hr
hr
hr
hr
Peroxypropionyl
nitrate, ppb
Avg.
0.7
NAa
0.3
0.2
0.08
0.15
0.09
0.05
0.11
3.0
0.66
0.05
0.05
0.20
Max.
2.7
6
1.8
0.9
0.13
0.5
0.33
0.32
0.63
5.0
0.25
0.13
0.07
3.1
(PPN/PAN, percent)
Avg. Max.
15
NAa
21
16
8
28
12
10
14
17
23
12
17
27
16
12
32
16
10
42
9
3
25
20
28
8
10
80
Reference
Singh
Darley
Singh
Singh
Singh
Singh
Singh
Singh
Singh
et
et
et
et
et
et
et
et
et
Lonneman
Singh
Singh
Singh
Singh
et
et
et
et
al.
al
al.
al.
al.
al.
al.
al.
al.
et
al.
al.
al.
al.
(1981)
. (1963)
(1979)
(1981)
(1979)
(1981)
(1982)
(1982)
(1982)
al. (1976)
(1982)
(1982)
(1982)
(1982)
aNot available.
Local time.
Source: Altshuller (1983).
-------
I
oo
TABLE 6-19. CONCENTRATIONS OF PEROXYACETYL AND PEROXYPROPIONYL NITRATES
IN LOS ANGELES, OAKLAND, AND PHOENIX, 1979
(ppb)
Value
Mean
Std. dev.
Maximum
Minimum
PAN
4.977
4.483
16.820
0.030
Los Angeles
PPN PPN/PAN, % PAN
0.722
0.673
2.740
ND3
14 0.356
0.422
16 1.850
0.050
Oakland
PPN PPN/PAN, % PAN
0.149 42
0.118
0.500 27
ND3
0.779
0.767
3.720
NDa
Phoenix
PPN
0.093
0.077
0.330
ND3
PPN/PAN, %
12
-
9
-
aNot detectable. Lower limit of detection is -0.02 part per trillion (ppt) for PAN and -0.03 ppt for PPN.
Source: Singh and Sal as (1983).
-------
(1976) could possibly have resulted in missing some peak PAN concentrations,
especially in transported air masses, the data from that study are comparable
to those of the Westberg et al. (1978) study, in which the sampling period
extended to 6:00 p.m. (local time).
Data from two nonurban sites in Canada are of interest even though they
are outside the United States. Cherniak and Corkum (1981; in Temple and
Taylor, 1983) measured PAN at a nonurban site in Simcoe, Ontario, Canada for 6
months. Measurements made by GC-ECD showed monthly means of <2 ppb and a
maximum concentration during the study of 5.6 ppb. At a'remote site in the
Kananaskis Valley of Alberta, Canada, monthly mean concentrations were <1 ppb
for samples taken at half-hour intervals, 24 hr/day, for 110 days. The site,
located at the base of a mountain range and about 50 miles west of Calgary, is
thought to be free of manmade pollutants, including transported pollutants.
Concentrations of PAN have recently been reported by Singh and Sal as
(1983) for a Pacific marine site, Point Arena, California, at which earlier
measurements were also made and reported by Singh et al. (1979) (see Table
6-20). Data collected in August 1982 showed concentrations of PAN ranging
from 0.01 to 0.12 ppb during the 5-day study period. The average concentration
for the period was 0.032 ± 0.024 ppb. The site is thought to be free of
manmade pollutants. Winds were west-to-northwest 90 percent of the time and
northerly the rest of the time. Modeled trajectories confirmed that air
masses passing over the site were of a marine origin.
In his comprehensive review, Altshuller (1983) compiled data on PAN, PPN,
and 03 concentrations at nonurban sites in the United States. These data are
presented in Table 6-20.
6.6.5 Temporal Variations in Ambient Air Concentrations of Peroxyacetyl Nitrate
6.6.5.1 Diurnal Patterns. Concentration data obtained in the 1960s were
briefly discussed in section 6.6.1, where it was noted that the first criteria
document for photochemical oxidants (U.S. Department of Health, Education, and
Welfare, 1970) documented concentrations and patterns that remain valid now.
In that document, the general proximity in time of PAN and oxidant peaks was
established by data from Los Angeles and Riverside, California. Maximum PAN
concentrations, although varying from location to location, generally occur in
midday; i.e., late morning to mid-afternoon. Figures 6-35 and 6-36, taken
from the 1970 criteria document (U.S. Department of Health, Education, and
019SP/A 6-85 6/18/84
-------
TABLE 6-20. CONCENTRATIONS IN AMBIENT AIR OF PEROXYACETYL AND PEROXYPROPIONYL NITRATES AND OZONE
AT NONURBAN REMOTE SITES IN THE UNITED STATES
(ppb)
Site
Mill Valley, CA
Point Arena, CA
Badger Pass, CA
CTi
i
0° Reese River, NV
Jetmore, KA
Sheldon Wildlife
Reserve, TX
Wilmington, OH
Van Hi Seville, NJ
Reference
Singh et al
Singh et al
Singh et al
Singh et al
Singh et al
Westberg et
(1978)
Lonneman et
(1976)
Spicer and
(1981)
. (1979)
. (1979)
. (1979)
. (1979)
. (1979)
al.
al.
Sverdrup
Nature of site
Maritime
Clean-maritime
Remote-high
altitude
Remote- high
altitude
Rural -
continental
Rural-
continental
Rural -
continental
Rural-
continental
Period of
measurement and
no. of days (n)
Jan. 1977 (12)
Aug. -Sept. 1978 (7)
May 1977 (10)
May 1977 (7)
June 1978 (7)
October 1978 (9)
August 1974 (9)
July-Aug. 1979 (31)
Period
of day
24 hr
24 hr
24 hr
24 hr
24 hr
24 hr
24 hr
24 hr
Average
concentrations
PAN
0.30
0.08
0.13
0.11
0.25
0.64
NAb
0.50
PPN
0.04
ND3
0.05
0.04
NDa
ND8
ND3
NO3
03
38
NDa
46
39
31
47
NAb
36
Maximum
concentrations
PAN
0.83
0.28
0.22
0.26
0.52
2.8
4.1
6.5
PPN
0.11
ND3
0.09
0.09
NDa
NO3
NO3
NDa
03
0.55
NDa
54
56
53
148
107
161
Avg. PAN/
Avg. 03, %
0.8
NDa
0.3
0.3
0.8
1.4
NAb
1.4
Not determined.
Measured, but results not given in the reference.
Source: Altshuller (1983).
-------
a
a
CO
cc
z
111
u
z
o
u
1
O
X
o
z
<
LU
0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0
I I I
AVERAGES:
-i 19 WEEKDAYS,
-1 OCTOBER
-116 WEEKDAYS,
"' SEPTEMBER
10 11 12
• a.m.-
•*+*
1 2 3
_ p.m
HOUR OF DAY, PST
Figure 6-35. Variation of mean 1-hour oxidant
and PAN concentrations, by hour of day, in
downtown Los Angeles, 1965.
Source: U.S. Department of Health, Educa-
tion, and Welfare (1970)
6-87
-------
a
a
O
UJ
O
Z
O
o
a
x
o
Z
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
I I I I I I
OXIDANT
I I 1
0.010
0.008
0.006
0.004
0.002
0
a
a.
DC
I-
ai
O
Z
O
o
1
10 12
8
a.m.-
•p.m.
HOUR OF DAY, PST
Figure 6-36. Variation of mean 1-hour
average oxidant and PAN concentra-
tions by hour of day. Air Pollution
Research Center, Riverside, CA,
September 1966.
Source: U.S. Department of Health,
Education and Welfare (1970)
6-88
-------
Welfare), graphically present the diurnal patterns of PAN in Los Angeles in
1965 and in Riverside in 1966. The occurrence of the second PAN peak in
Riverside, which appears to trail a second total oxidant peak by an hour or
two, was ascribed to transport, as verified by the occurrence of maximum
oxidant concentrations at three receptor sites east of West Los Angeles (down-
town Los Angeles, Azusa, and Riverside), at times that corresponded with
respective distances from West Los Angeles.
Two examples drawn from recent data substantiate that the general diurnal
pattern (as it appears in composite diurnal data averaged over a week, a
month, or longer) remains the same as the pattern established by data obtained
in the mid-1960s.
Using FTIR spectroscopy, Tuazon and coworkers (Tuazon et al., 1978, 1980,
1981b) measured concentrations of PAN at Claremont and Riverside over a 5-year
period. Concentrations of PAN ranged from about 5 to 40 ppb over the course
of the study. The diurnal profiles for PAN and ozone at Claremont are shown
for 2 days of a multi-day smog episode in October 1978 in Figure 6-37 (Tuazon
et al., 1981b). Note the qualitative relationship of the two pollutants, in
which the peak concentrations of the two occur almost simultaneously. That
the relationship between PAN and ozone concentrations and behavior in the
atmosphere is neither constant nor monotonic is borne out by the slight time
differences in occurrence of their peak concentrations but especially by the
persistence of somewhat elevated PAN concentrations before return to "baseline"
levels. It appears that PAN concentrations, in this instance at least, closely
parallel the nitric acid (HN03) concentrations, persisting after ozone concen-
trations have subsided. The quantitative relationship between PAN and ozone
differed slightly at the peak concentrations of the two on the 2 days. On
October 12, the peak PAN concentration was close to 6 percent of the peak
ozone concentration; on October 13, the peak PAN concentration was nearly 8
percent of the peak ozone concentration.
6.6.5.2 Seasonal Patterns. Seasonal differences in PAN concentrations were
alluded to in section 6.6.3 and mean and maximum PAN concentrations were pre-
sented by month for 2 years (1967-1968 and 1980) for Riverside, California, in
Figure 6-34. That seasonal differences exist in PAN concentrations was docu-
mented in the 1970 criteria document for photochemical oxidants (U.S. Department
of Health, Education, and Welfare, 1970). Oxidant data for the same period
were obtained by continuous Mast meter measurements, 24 hr/day. Concentrations
019SP/A 6-89 6/18/84
-------
I
l£>
O
Q.
a
Z
O
g
1C
UJ
O
Z
O
O
HI
Z
O
N
O
0.50
1000 1400 1800
OCTOBER 12, 1978
2200 0200 0600 1000
TIME OF DAY, PDT
1400 1800 2200
i J
OCTOBER 13, 1978
a
a
to
O
5
cc
O
Z
o
o
1
Q
O
Z
X
Figure 6-37. Diurnal profiles of ozone and PAN at Claremont, CA,
October 12 and 13, 1978, 2 days of a multi-day smog episode.
Source: Tuazon et al. (1981)
-------
of PAN were measured in sequential samples analyzed by GC-ECD from 6:00 a.m.
to about 4:00 or 5:00 p.m. The data are not strictly comparable, since the
shorter, daylight averaging time for PAN would be expected to result in somewhat
higher mean concentrations of PAN than would be obtained across a 24-hour
averaging period. Nevertheless, the patterns, given in Figure 6-38, are of
interest and demonstrate that peak PAN concentrations can constitute a higher
percentage of the peak ozone concentrations during winter months than during
the rest of the year. This observation is still valid (Spicer et al. , 1983)
and has been attributed by Holdren et al. (1984) to greater stability in the
winter months because of cooler temperatures (Cox and Roffey, 1977). The
possibility that the somewhat greater NO emissions of the heating season
y\.
(winter months) may contribute to this phenomenon should not be overlooked.
Data from one additional study complement data already presented on diur-
nal and seasonal patterns. Lewis et al. (1983) measured PAN and ozone concen-
trations from September 25, 1978, to May 10, 1980. Average (10-hr and 24-hr)
and maximum concentrations of both pollutants are given in Table 6-21 by month
of the year (Lewis et al., 1983). Note that the highest monthly mean concen-
trations, both 10-hr and 24-hr, occurred during the smog season (August and
September) but that the next highest occurred in October and February, respec-
tively. Average diurnal profiles were obtained during this same study and are
shown, by month, in Figure 6-39 (Lewis et al., 1983).
6.6.6 Spatial Variations in Ambient Air Concentrations of Peroxyacetyl Nitrate
6.6.6.1 Urban-Rural Gradients and Transport of PAN. Whether PAN, like ozone,
can be transported from urban to rural areas is of consequence in assessing
potential exposures of crops in agricultural areas. As noted earlier, precursors
to PAN, especially NOp, are lower in nonurban than in urban areas, such that
little local formation is expected in nonurban areas. Available data on PAN
concentrations indicate clearly that they are lower in nonurban areas than in
urban (section 6.5.3). It should be noted, however, that data on concentrations
in agricultural areas are quite sparse, such that the possibility of transport
is important in assessing exposures of vegetation, especially since PAN is a
known phytotoxicant. Studies by Lonneman et al. (1976) and Nieboer and Van Ham
(1976) that were cited in the 1978 criteria document (U.S. Environmental
Protection Agency, 1978) reported the transport of PAN. The more recent study
by Nielsen et al. (1982) has confirmed that PAN can be present at relatively
019SP/A 6-91 6/18/84
-------
! I
I I I i ! I I
MONTHLY MEANS OF DAILY MAXIMUM
1-hour AVERAGE CONCENTRATIONS
_ MONTHLY MEANS OF 1-hour AVERAGE
CONCENTRATIONS
OXIDANT .PAN
-o o—
,»- PAN ^-«
I i I
LU
5
JUN. JUL AUG. SEP. OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUN.
MONTH I
- »«i - 1967 -
Figure 6-38. Monthly variation of oxidant (Mast meter, con-
tinuous 24-hr) concentrations and PAN {GC-ECD, sequential, 6:00
a.m. to 4:00-5:00 p.m.) concentrations. Air Pollution Research
Center, Riverside, CA, June 1966 - June 1967.
Source: U.S. Department of Health, Education, and Welfare
(1970)
6-92
-------
TABLE 6-21. PAN AND OZONE CONCENTRATIONS IN AMBIENT AIR, NEW BRUNSWICK, N.J.,
FOR SEPTEMBER 25, 1978, TO MAY 10, 1980
PAN concentration, ppb
Month
January
February
March
April
May
June
July
August
September
October
November
December
24-hr
average
0.12
0.61
0.36
0.45
0.23
0.09
0.26
1.17
1.04
0.93
0.25
0.57
10-hr
average
0.19
0.69
0.41
0.57
0.28
0.17
0.44
1.63
1.41
1.08
0.31
0.62
Hourly
maximum
1.3
4
1.3
2.5
I
0.8
3.5
10.6
7.5
5.8
3.5
2.5
0<5 concentration, ppb
24-hr
average
11.5
17.2
23
28.5
31.4
NA
37.5
37.4
22.4
15.8
11.6
9.7
10-hr
average
15.5
23.2
29.1
37.3
40.9
NA
57.6
55.9
33.9
22.6
15.8
12.8
Hourly
maximum
34
40
58
80
78
NA
130
145
110
68
40
35
These results are lower than expected; however, there was no evidence of
instrument malfunction.
Source: Lewis et al. (1983).
high concentrations in photochemically polluted air after long-range transport.
Variations in concentrations of PAN and other oxidants measured in Claremont,
California (Grosjean, 1983), are consistent with transport patterns. Recent
work of Singh and Salas (1983) has shown that significant nighttime PAN concen-
trations can occur aloft, at least in a relatively clean environment. It is
possible that the transport of PAN occurs aloft, as with ozone, and that it
can be transported long distances under the right conditions.
6.6.6.2 Intracity Variations. Few data on PAN concentrations at different
sites in the same city are available. One study is available for Houston,
Texas (Jorgen et al. , 1978), in which PAN was measured on October 26 and 27,
1977, at three separate sites, two within Houston and one just north of the
city. Diurnal concentration plots for the three sites are shown as Figures
6-40 through 6-42. Site 2 was in Houston, near the junction of routes 1-10
and 1-45. Site 3 was located about 11 miles south-southeast of site 2; site 1
was about 12 miles north and slightly east of site 2; and sites 1 and 3 were
about 22 miles apart.
019SP/A
6-93
6/18/84
-------
£
a
a
z
0.080
0.060
0.040
0.020
DC
H
jfj 0.025
O
Z
O
° 0.015
0.005
_ OZONE
_ PAN
-a.m.-
12
-p.m.
24
TIME OF DAY, hour
Figure 6-39. Average daily profile by
month (July 7 - October 10} for PAN
and ozone in New Brunswick, NJ,
1979. Numbers refer to months of the
year.
Source: Lewis et al. (1983)
6-94
-------
E
a
a.
z
o
LU
u
z
o
o
<
GL
0.025
0.020
0.015
0.010
0.005
SITE1
6:00 a.m. 10/26/77
TO 10:00 p.m. 10/27/77
0.200
0.150
0.100
0.050
o.
a
z"
O
LU
O
z
o
u
o
X
o
12
18
!.*-»•*
-p.m.
24
-»-M-
12
-a.m.-
TIME OF DAY, hour
18 22
-p.m.—*\
Figure 6-40. Diurnal plot of PAN and oxidant
concentrations at site just north of Houston,
October 26-27, 1977.
Source: Jorgen et al. (1977)
6-95
-------
0.025
a
a
Z 0.020
O
0.015
cc
O
g 0.010
O
< 0.005
I
I
OXIDANT
PAN
SITE 2
6:00 a.m. 10/26/77
TO 10:00 p.m. 10/27/77
12
18
•p.m.
24
>[«
I -
12
Q.
a
«
0.200 2
0.150
0.100
UJ
o
2
O
o
H-
0.050 Q
X
o
18 22
a.m.
TIME OF DAY, hour
p.m
H
Figure 6-41. Diurnal plot of PAN and oxidant
concentrations at site in Houston, near junction
of 1-10 and I-45, October 26-27, 1977.
Source: Jorgen et al. (1978)
6-96
-------
PAN CONCENTRATION, ppm
en
i
CO
O
c
O
(O
CD
3
(D
i-f
0)
(O
>J
00
Q) (Q
O •*
3 ®
cn o>
Qj £,
"* NJ
0) '
5>' 2
= 3
CO 03
§E.
ifS
2 o
8> •*
p* -o
z>
lz
W S
o a
? o
OS.
l&
8-3
W O
"• O
N) 3
O
N>O>
^j 3
p
O
3i
Q)
3
«
— H
o
•n
o
o •
•p
3
o>
N)
I I
I '>/\
\ I
p
8
OXIDANT CONCENTRATION, ppm
-------
6.6.6.3 Indoor-Outdoor Ratios of PAN Concentrations. No recent studies
appear in the literature on indoor concentrations of PAN or indoor-outdoor
ratios (I/O). In three school buildings in southern California, Thompson et
al. (1973) found I/O ratios (expressed here as percentages) of 89, 97, and 148
percent, respectively, in the absence of air conditioning. With air condi-
tioning, the I/O ratios were 75, 108, and 117 percent, respectively. Total
oxidants were nearly constant all day, remaining about 30 percent (in air
conditioning) of the average outside concentration. The higher I/O for PAN
than for oxidants was attributed to the greater breakdown of ozone ("oxidants")
through its reaction with surfaces. The lesser reactivity of PAN with surfaces
and the cooler temperatures are the probable causes of its greater persistence
indoors. Like ozone, however, it also decays indoors, but over an extended
period (Thompson et al., 1973).
6.7 CONCENTRATIONS OF OTHER PHOTOCHEMICAL OXIDANTS IN AMBIENT AIR
Concentrations of nitrogen dioxide (NO,,) and related nitrogenous oxidants
are presented in a recent criteria document on oxides of nitrogen (U.S.
Environmental Protection Agency, 1982) and are not presented here. In addition,
the comprehensive review by Altshuller (1983) also documents available informa-
tion on nitrogenous oxidants such as nitric acid (HNO-) and peroxynitric acid,
as well as on formic acid (HCOOH) and hydrogen peroxide (H?0?). The reader is
referred to these reviews for detailed information on these oxidants. Because
formic acid and hydrogen peroxide are appropriate concerns for this document,
however, the limited information on concentrations of these two pollutants is
summarized below.
Neither health nor welfare effects have ever been attributed to the pre-
sence of formic acid in ambient air. It has been found in polluted areas such
as the Los Angeles Basin at concentrations up to 20 ppb. Maximum concentrations
of HCOOH observed by Tuazon and coworkers, using FTIR (Tuazon et al., 1978;
1980; 1981b), in Claremont and Riverside were in the range 5 to 20 ppb in a
study covering 5 years. The ranges of concentrations of HCOOH measured by
Tuazon et al. were consistent with those found in a recent long-path FTIR
study by Hanst et al. (1982). The FTIR method offers a reliable assessment of
the ambient air concentrations of HCOOH and reported concentrations are believed
to be accurate.
019SP/A 6-98 6/18/84
-------
Data on HCOOH concentrations for 2 days in October 1978 are shown in
Figure 6-43 (Tuazon et al., 1981b). The diurnal pattern is similar to that of
related oxidants and of some of the other smog products.
The measurement of hydrogen peroxide (H-CL) in ambient air is fraught
with difficulties that remain unresolved. Ozone itself is known to be an
interference in virtually all of the past and current measurement methods for
H.O- (chapter 5). In his comprehensive review of non-ozone oxidants and other
smog constituents and photochemical products, Altshuller (1983) examined data
obtained in the South Coast Air Basin of California in the late 1970s and in
1980 for possible consistency in the interference of ozone in H?0? measurements.
Laboratory experiments (Heikes et al., 1982) have indicated that 1 ppb H-O^
would be generated per 100 ppb ozone. Altshuller's analysis shows that this
relationship does not hold in ambient air in the South Coast Air Basin once
H-O- levels exceed about 5 to 10 ppb, and Altshuller (1983) concluded that
variations in H?0_ measurements there remain unexplained.
Because of measurement problems, the true levels of H-O- in ambient air
are unknown, especially in polluted areas, where multiple interferences may
possibly occur. The FTIR method has been used to look for hydrogen peroxide
in ambient air, but concentrations, even in polluted areas, are apparently
below the limits of detection of the method; FTIR spectroscopy is known to be
capable of measuring H.O- with specificity if it is present above the limits
of detection (40 ppb at 1 km pathlength; see chapter 5).
Notwithstanding measurement difficulties, some ranges of H_0p concentra-
tions at urban and nonurban sites have been reported in the literature. These
are given in Table 6-22, along with the general type of measurement method
used to obtain the reported concentrations. It must be kept in mind, however,
that the reported concentrations are not thought to be accurate but to be
rough approximations only of H_02 levels in ambient air.
6.8 SUMMARY
In the context of this document, the concentrations of ozone and other
photochemical oxidants found in ambient air are important for:
1. Assessing potential exposure of individuals; communities; the general
population and those subpopulations in communities that may be
especially susceptible to adverse effects from these oxidants;
natural ecosystems, managed ecosystems such as crops; and nonbiologi-
cal materials such as polymers and paints.
019SP/A 6-99 6/18/84
-------
en
i
O
O
a
a
Z
O
O
z
O
O
O
O
a
a
ui
O
Z
O
O
ui
I
O
0.080
-OCTOBER 12, 1978
1000 1400 1800 2200
TIME OF DAY, PDT
0200 0600 1000
OCTOBER 13, 1978
1400 1800 2200
1000 1400 1800 2200
I* OCTOBER 12, 1978
0200 0600 1000
TIME OF DAY, PDT
1400 1800 2200
OCTOBER 13, 1978-
Figure 6-43. Diurnal profile of HCOOH, along with other oxidants and
smog constituents, on October 12 and 13, 1978, at Claremont, CA.
Q.
Q.
O
O
O
1
a
o
z
Source: Tuazon et al. (1981b).
-------
TABLE 6-22. CONCENTRATIONS OF HYDROGEN PEROXIDE IN AMBIENT AIR
AT URBAN AND NONURBAN SITES
Location
Hoboken, NJ
(urban)
Riverside, CA
(urban)
Riverside and
Claremont, CA
(urban)
Minneapolis, MN
(urban)
Boulder, CO
(urban)
Boulder, CO
(nonurban,
east of
Boulder)
Tucson, AZ
(nonurban,
54 km SE
of Tucson)
Tucson, AZ
(remote, near
Tucson)
Concns. , ppb;
Date comments
1970 < 40
1970 <180 (during
smog episode
with 650 ppb
oxidants)
July-Aug 100 max. (ozone
1977 also 100 and
increasing);
10 to 50 on
most days
NAa <6
NAa <0.5
Feb. 1978 0.2 to 3
NAa <7
NAa -1
Method
Titanium (IV)
sulfate/8-qui no-
li no!
Titanium (IV)
sul f ate/8-qui no-
linol
Luminol
chemi 1 umi nescence
Wet chemical
Wet chemical
Luminol
chemi luminescence
Luminol
chemi 1 umi nesence
Luminol
chemi 1 umi nescence
Reference
Bufalini et
al. (1972)
Bufalini et
al. (1972)
Kok et al.
(1978)
Heikes et al.
(1982)
Heikes et al.
(1982)
Kelly et al.
(1979)
Farmer and
Dawson (1982)
Farmer and
Dawson (1982)
Not available from Altshuller (1983).
See chapter 5 for method used by Heikes et al. (1982).
Source: Derived from data in Altshuller (1983).
019SP/A
6-101
6/18/84
-------
2. Assessing whether levels of ozone and other photochemical oxidants
in ambient air are within or near the range of concentrations shown
to produce adverse effects in health and welfare effects studies.
3. Determining whether levels of ozone and other photochemical oxidants
are as high indoors as in the ambient air, for the purpose of asses-
sing actual, as opposed to potential, exposures of individuals in
the general population or susceptible subpopulations.
4. Assessing whether non-ozone photochemical oxidants occur in ambient
air at levels within or near the range of concentrations shown to
produce potentially adverse effects in health and welfare effects
studies.
5. Assessing whether concentrations of ozone plus the other photochemical
oxidants together occur at levels sufficient to produce adverse
effects in the general or susceptible subpopulations, or in vegetation
and ecosystems.
6. Evaluating the relationship(s) between ozone and the other photochemi-
cal oxidants, in order to determine whether ozone can function as a
control surrogate in the event that these other, non-ozone photochem-
ical oxidants are found to produce adverse effects on public health
and welfare.
6.8.1 Ozone Concentrations in Urban Areas
The current ozone standard is expressed in terms of a 1-hour value not to
be exceeded on more than 1 day per year. Thus, the second-highest value is a
concentration of significance, since it determines compliance with the standard
and is, thereby, an indicator of exposures having potential health and welfare
significance.
In this chapter, the second-highest 1-hour ozone concentrations reported
in each of 4 years have been given for the 80 most populous Standard Metropoli-
tan Statistical Areas (SMSAs) of the United States, i.e., those with popula-
tions > 0.5 million. In Table 6-19, 1982 ozone concentrations for the subset
of SMSAs with populations > 1 million are given by geographic area, demarcated
according to United States Census divisions and regions (U.S. Department of
Commerce, 1982). The second-highest concentrations measured in 1982 in those
38 SMSAs having populations of at least 1 million ranged from 0.09 ppm in the
Ft. Lauderdale, Florida, and Seattle, Washington, areas to 0.32 ppm in the Los
Angeles and Riverside, California, areas. The second-highest 1-hour ozone
concentrations for 32 of the 38 SMSAs in Table 6-19 equal or exceed 0.12 ppm.
019SP/A 6-102 6/18/84
-------
TABLE 6-23. SECOND-HIGHEST 1-HR OZONE CONCENTRATIONS IN 1982
STANDARD METROPOLITAN STATISTICAL AREAS WITH POPULATIONS
> 1 MILLION, GIVEN BY CENSUS DIVISIONS AND REGIONS3
IN
Division
and region
SMSA
population,
SMSA millions
Second-highest
1982 03
concn. , ppra
NORTHEAST
New England
Boston, MA
Middle Atlantic Buffalo, NY
Nassau-Suffolk, NY
Newark, NJ
New York, NY/NJ
Philadelphia, PA/NJ
Pittsburgh, PA
SOUTH
>2
1 to <2
>2
I to <2
>2
>2
>2
0.16
0.11
0.13
0.17
0.17
0.18
0.14
South Atlantic
SOUTH
West South
Central
NORTH CENTRAL
East North
Central
West North
Central
Atlanta, GA >2
Baltimore, MD >2
Ft. Lauderdale-Hollywood, FL 1 to <2
Miami, FL 1 to <2
Tampa-St. Petersburg, FL 1 to <2
Washington, DC/MO/VA >2
Dallas-Ft. Worth, TX >2
Houston, TX >2
New Orleans, LA 1 to <2
San Antonio, TX 1 to <2
Chicago, IL >2
Detroit, MI >2
Cleveland, OH 1 to <2
Cincinnati, OH/KY/IN I to <2
Milwaukee, WI I to <2
Indianapolis, IN 1 to <2
Columbus, OH 1 to <2
St. Louis, MO/IL >2
Minneapolis-St. Paul, MN/WI >2
Kansas City, MO/KS 1 to <2
0.14
0.14
0.09
0.14
0.11
0.15
0.17
0.21
0.17
0.14
0.12
0.16
0.12
0.13
0.13
0.12
0.13
0.15
0.10
0.10
019SP/A
6-103
6/18/84
-------
TABLE 6-23 (cont'd). SECOND-HIGHEST 1-HR OZONE CONCENTRATIONS IN 1982 IN
STANDARD METROPOLITAN STATISTICAL AREAS WITH POPULATIONS > 1 MILLION
GIVEN BY CENSUS DIVISIONS AND REGIONS8
Division
and region SMSA
WEST
Mountain Denver-Boulder, CO
Phoenix, AZ
Pacific Los Angeles- Long Beach, CA
San Francisco-Oakland, CA
Anaheim-Santa Ana-
Garden Grove, CA
San Diego, CA
Seattle-Everett, WA
Rivers1de-San Bernardino-
Ontario, CA
San Jose, CA
Portland, OR/WA
Sacramento, CA
SMSA Second-highest
population, 1982 03
millions concn. , pptn
1 to <2
1 to <2
>2
>2
1 to <2
1 to <2
1 to <2
1 to <2
1 to <2
1 to <2
1 to <2
0.14
0.12
0.32
0.14
0.18
0.21
0.09
0.32
0.14
0.12
0.16
Standard Metropolitan Statistical Areas and geographic divisions and regions
as defined by Statistical Abstract of the United States (U.S. Department of
Commerce, 1982).
Source: U.S. Environmental Protection Agency, SAROAD data file for 1982.
The data clearly show, as well, that the highest 1-hour ozone concentrations
in the United States occur in the northeast (New England and Middle Atlantic
States), in the Gulf Coast area (West South Central states), and on the west
coast (Pacific states). Second-highest 1-hour concentrations in the SMSAs
within each of these three areas average 0.15, 0.17, and 0.19 ppm, respectively.
It should be emphasized that these three areas of the United States are subject
to those meteorological and climatological factors that are conducive to local
oxidant formation, or transport, or both. It should also be emphasized that
11 of the 16 SMSAs in the country with populations > 2 million are located in
these areas.
Sources of oxidant precursors are strongly correlated with population
(chapter 3). In accord with this relationship, three population groups within
the 80 largest SMSAs (Table 6-6) had the following median values for their
019SP/A
6-104
6/18/84
-------
collective second-highest 1-hour ozone concentrations in both 1981 and 1982:
populations > 2 million, 0.15 ppm 0-; populations of 1 to 2 million, 0.13 ppm
03; and populations of 0.5 to 1 million, 0.12 ppm 0~.
Among all stations reporting valid ozone data (> 75 percent of possible
hourly values per year) in 1979, 1980, and 1981 (collectively, 906 station-
years) in the United States, the median second-highest 1-hour ozone value was
0.12 ppm, and 5 percent of these stations reported second-highest 1-hour
values > 0.28 ppm (Figure 6-9).
A pattern of concern in assessing human physiological and vegetational
responses to ozone is the occurrence of repeated or prolonged periods, or
both, when the ozone concentrations are close to or equal or exceed levels
known to elicit responses. In addition, the number of days of respite between
such multiple-day periods of high ozone is of possible consequence. Data
presented in this chapter (Figures 6-23 through 6-26 and accompanying text)
show that the probabilities of prolonged exposures to (consecutive days) or
respites from (consecutive days) specified concentrations are location-specific.
In Pasadena, CA, a high-ozone area, there is a 42 percent probability (based
on 1979 through 1981 aerometric data) that an ozone concentration of 0.18
ppm, once reached, is likely to persist for 3 days or longer. Other, lower-
ozone areas show lower probabilities of such multi-day high ozone concentra-
tions. These and other data presented demonstrate the occurrence, at least in
some urban areas, of multiple-day exposures to relatively high concentrations
of ozone.
6.8.2. Trends in Urban and Nationwide Ozone Concentrations
Discussion in chapter 5 pointed out and substantiated that aerometric
data obtained by potassium iodide methods in earlier years are essentially
concentrations of ozone rather than "total oxidants." Comparison of concen-
trations of "total oxidants" in major urban areas for 1974 and 1975 in Table
6-2 (U.S. Environmental Protection Agency, 1978) with ozone data in Table 6-6
for those same urban areas in 1979 through 1982 (U.S. Environmental Protection
Agency, SAROAD file) shows that the more recent ozone concentrations are in
the same general range for many cities, have declined in some, and are somewhat
higher in others.
Trends in nationwide ozone concentrations, gauged by annual averages at
two subsets of stations reporting data from 1974 through 1981, show declines
019SP/A 6-105 6/18/84
-------
of 15 to 20 percent. These trend data represent urban areas almost exclusively.
Interpretation of this trend is complicated by four potentially significant
influences: (1) a change in calibration procedure (1979); (2) improved data-
quality audits (1979); (3) possible shifts in underlying meteorological patterns;
and (4) changes in precursor emission rates. When adjustments for the first
two factors are made, a portion of the decrease is real. The exact portion of
the decline that is attributable to the calibration change can not be determined
without minute examination of aerometric data records from each monitoring
station, since some monitoring stations began using the UV calibration procedure
as early as 1975, some changed to UV calibration in 1979 (but not all in the
same month of 1979), and some used the interim BAKI calibration procedure
permitted by EPA for up to 18 months after promulgation of the UV calibration
procedure (Hunt and Curran, 1982; also chapter 5).
6.8.3. Ozone Concentrations in Nonurban Areas
Nonurban areas are not routinely monitored for ozone concentrations.
Consequently, the aerometric data base for nonurban areas is considerably less
substantial than for urban areas. Data on maximum 1-hour concentrations and
arithmetic mean 1-hour concentrations reveal that maximum (peak) 1-hour concen-
trations at nonurban sites classified as rural (SURE study, Martinez and
Singh, 1979; NAPBN studies, Evans et al., 1983) may exceed the concentrations
observed at nonurban sites classified as suburban (SURE study, Martinez and
Singh, 1979). For example, maximum 1-hour ozone concentrations measured in
1980 at Kisatchie National Forest (NF), Louisiana; Custer NF, Montana; and
Green Mt. NF, Vermont, were 0.105, 0.070, and 0.115 ppm, respectively. Arith-
metic mean concentrations for 1980 were 0.028, 0.037, and 0.032 ppm at the
respective sites. For four nonurban (rural) sites in the SURE study, maximum
1-hour ozone concentrations were 0.106, 0.107, 0.117, and 0.153; and mean
1-hour concentrations ranged from 0.021 to 0.035 ppm. At the five nonurban
(suburban) sites of the SURE study, maximum concentrations were 0.077, 0.099,
0.099, 0.080, and 0.118 ppm, respectively. The mean 1-hour concentrations at
those sites were 0.023, 0.030, 0.025, 0.020, and 0.025 ppm, respectively.
Comparison of these data with data for nonurban and remote locations
during the 1973-1976 period show that mean concentrations at these various
nonurban locations are not dissimilar. Ranges of concentrations and the
maximum 1-hour concentrations at the NAPBN and SURE sites show the probable
019SP/A 6-106 6/18/84
-------
influence, however, of ozone transported from urban areas. In one documented
case, for example, a 1-hour peak ozone concentration of 0.125 ppm at an NAPBN
site in Mark Twain National Forest, Missouri, was measured during passage of
an air mass whose trajectory was calculated to have included Detroit, Cincinnati,
and Louisville in the preceding hours (Evans et a!., 1983).
These data corroborate the conclusion given in the 1978 criteria document
(U.S. Environmental Protection Agency, 1978) regarding urban-nonurban and
urban-suburban gradients; i.e., nonurban areas may sustain higher ozone con-
centrations than those found in urban areas. Reasons for this phenomenon
include induction and transport times, as well as the possible additive effects
of plumes from suburban or smaller areas as air masses pass over them downwind
from urban areas. Generally, however, lower ozone concentrations occur in
nonurban areas, as the data in this chapter indicate.
6.8.4. Patterns in Ozone Concentrations
Since the photochemical reactions of precursors that result in ozone for-
mation are driven by sunlight, as well as emissions, the patterns of ozone
occurrence in ambient air depend on daily and seasonal variations in sunlight
intensity. The typical diurnal pattern of ozone in ambient air has a minimum
ozone level around sunrise (near zero in most urban areas), increasing through
the morning to a peak concentration in early afternoon, and decreasing toward
minimal levels again in the evening. Obviously, meteorology is a controlling
factor; if strong winds disperse the precursors or heavy clouds intercept the
sunlight, high ozone levels will not develop. Another important variation on
the basic diurnal pattern appears in some localities as a secondary peak in
addition to the early afternoon peak. This secondary peak may occur any time
from midafternoon to the middle of the night and is attributed to ozone trans-
ported from an upwind area where high ozone levels have occurred earlier in the
day. As documented in this chapter, secondary peak concentrations may be higher
than concentrations resulting from the photochemical reactions of locally emit-
ted precursors (Martinez and Singh, 1979). At one nonurban site in Massachusetts,
for example, primary peak concentrations of about 0.11, 0.14, and 0.14 occurred
at noon, from noon to about 4:00 p.m., and at noon, respectively, on 3 succes-
sive days of high ozone levels (Martinez and Singh, 1979). Secondary peaks at
the same site for those same 3 days were about 0.150, 0.157, and 0.130 ppm at
about 6:00 p.m., 8:00 p.m., and 8:00 p.m., respectively (Martinez and Singh,
1979).
019SP/A 6-107 6/18/84
-------
Because weather patterns, ambient temperatures, and the intensity and
wavelengths of sunlight all play important roles in oxidant formation, strong
seasonal as well as diurnal patterns exist. The highest ozone levels occur in
the summer and fall (second and third quarters), when sunlight reaching the
lower troposphere is most intense and stagnant meteorological conditions
augment the potential for ozone formation and accumulation. Average summer
afternoon levels can be from 150 to 250 percent of the average winter afternoon
levels. Minor variations in the smog season occur with location, however. In
addition, it is possible for the maximum and second-highest 1-hour ozone
concentration to occur outside the two quarters of highest average ozone
concentrations (cf. data for Tucson, Arizona, and data for the California
sites given in this chapter). Exceptions to seasonal patterns are important
considerations with regard to the protection of crops from ozone damage,
especially since respective crops have different growing seasons in terms of
length, time of year, and areas of the country in which they are grown.
As data in this chapter for different averaging times clearly demonstrate,
averaging smooths out and submerges the occurrence of peak concentrations.
This is an obvious and familiar statistical phenomenon. It is pointed out,
however, because it has direct relevance to the protection of public health
and welfare. Averaging times must correspond to or be related in a consistent
manner to the pattern of exposure that elicits untoward responses.
Certain spatial variations in ozone concentrations occur that are general-
ly of little consequence in exposure assessment. For example, ozone concentra-
tions increase with increasing altitude (e.g., Viezee and Singh, 1981). The
gradients are of no known consequence for inhabited elevations. They could
potentially be of some consequence for high-altitude flights unless compensated
for by adequate ventilation/filtration systems. Likewise, ozone concentrations
exhibit hemispheric asymmetry (Logan et al., 1981), with concentrations highest
in the northern hemisphere. Aerometric data sufficiently describe concentra-
tions in the latitudes of the United States such that the fact of asymmetry
has no practical consequences.
Spatial variations on a smaller scale assume more importance relative to
exposure assessment. Indoor-outdoor gradients in ozone concentrations are
known to occur even in structures ventilated by fresh air rather than air
conditioning (e.g., Sabersky et al., 1973; Thompson et al., 1973). Ozone
reacts with surfaces inside buildings, so that decay occurs fairly rapidly.
019SP/A 6-108 6/18/84
-------
Ratios of indoor-to-outdoor (I/O) ozone concentrations are quite variable,
however, since the presence or absence of air conditioning, air infiltration
or exchange rates, interior air circulation rates, and the composition of
interior surfaces all affect indoor ozone concentrations. Ratios (I/O) in the
literature thus vary from 80 ± 10 percent (Sabersky et al., 1973) in an air-
conditioned office building (but with 100 percent outside air intake) to 10 to
25 percent in air-conditioned residences (Berk et al., 1981).
On a larger scale, within-city variations in ozone concentrations can
occur, despite the commonly accepted maxim that ozone is a regional pollutant.
Data in this chapter show relatively homogeneous ozone concentrations in New
Haven, Connecticut (U.S. Environmental Protection Agency, SAROAD files), which
is a moderately large city downwind of a reasonably well-mixed urban plume
(Cleveland et al., 1976). In a large metropolis such as New York City, however,
appreciable gradients in ozone concentration can exist from one side of the
city to the other (Smith, 1981). Such gradients must be taken into considera-
tion in exposure assessments.
6.8.5 Concentrations and Patterns of Other Photochemical Oxidants
6.8.5.1 Concentrations. No aerometric data are routinely obtained by Federal,
state, or local air pollution agencies for any photochemical oxidants other than
nitrogen dioxide and ozone. The concentrations presented in this chapter for
non-ozone oxidants were all obtained in special field investigations. The
limitations in the number of locations and areas of the country represented in
the information presented simply reflect the relative paucity of data in the
published literature.
The four non-ozone photochemical oxidants for which concentration data
have been presented are formic acid, peroxyacetyl nitrate (PAN), peroxypropionyl
nitrate (PPN), and hydrogen peroxide. Peroxybenzoylnitrate has not been
clearly identified in ambient air in the United States.
Recent data indicate the presence in urban atmospheres of only trace
amounts of formic acid (< 15 ppb, measured by FTIR). Estimates in the earlier
literature (1950s) of 600 to 700 ppb of formic acid in smoggy atmospheres were
erroneous because of faulty measurement methodology (Hanst et al., 1982).
The measurement methods (IR and GC-ECD) for PAN and PPN are specific and
highly sensitive, and have been in use in air pollution research for nearly
two decades. Thus, the more recent literature on the concentrations of PAN
019SP/A 6-109 6/18/84
-------
and PPN confirm and extend, but do not contradict, earlier findings reported
in the two previous criteria documents for ozone and other photochemical
oxidants (U.S. Department of Health, Education, and Welfare, 1970; U.S. Environ-
mental Protection Agency, 1978).
Table 6-16 summarized the concentrations of PAN reported in the literature
from 1960 through the present. The highest concentrations reported over this
extended period were those found in the 1960s in the Los Angeles area: 70 ppb
(1960), 214 ppb (1965); and 68 ppb (1968) (Renzetti and Bryan, 1961; Mayrsohn
and Brooks, 1965; Lonneman et a!., 1976; respectively).
The highest concentrations of PAN measured and reported in the past
5 years were 42 ppb at Riverside, California, in 1980 (Temple and Taylor,
1983) and 47 ppb at Claremont, California, also in 1980 (Grosjean and Kok,
1981). These are clearly the maximum concentrations of PAN reported for
California and for the entire country in this period. Other recently mea-
sured PAN concentrations in the Los Angeles Basin were in the range of 10 to
20 ppb. Average concentrations of PAN in the Los Angeles Basin in the past
5 years ranged from 4 to 13 ppb (Tuazon et a!., 1981a; Grosjean and Kok, 1981;
respectively). Only one published report covering PAN concentrations outside
California in the past 5 years is that of Lewis et al. (1983) for New Brunswick,
New Jersey. The average PAN concentration was 0.5 ppb and the maximum was 11 ppb
during a study done from September 1978 through May 1980. Studies outside
California from the early 1970s through 1978 showed average PAN concentrations
ranging from 0.4 ppb in Houston, Texas, in 1976 (Westberg et al., 1978) to
6.3 ppb in St. Louis, Missouri, in 1973 (Lonneman et al., 1976). Maximum PAN
concentrations outside California for the same period ranged from 10 ppb in
Dayton, Ohio, in 1974 (Spicer et al. , 1976) to 25 ppb in St. Louis (Lonneman
et al., 1976).
Table 6-18 summarized the findings of reports of PPN concentrations from
1963 through the present. The highest PPN concentration reported in these
studies was 6 ppb in Riverside, California (Darley et al., 1963). The next
highest reported PPN concentration was 5 ppb at St. Louis, Missouri, in 1973
(Lonneman et al., 1976). Among more recent data, maximum PPN concentrations
at respective sites ranged from 0.07 ppb in Pittsburgh, Pennsylvania, in 1981
(Singh et al., 1982) to 3.1 ppb at Staten Island, New York (Singh et al., 1982).
California concentrations fell within this range. Average PPN concentrations at
the respective sites for the more recent data ranged from 0.05 ppb at Denver and
Pittsburgh to 0.7 ppb at Los Angeles in 1979 (Singh et al., 1981).
019SP/A 6-110 6/18/84
-------
Altshuller (1983) has succinctly summarized the nonurban concentrations
of PAN and PPN by pointing out that they overlap the lower end of the range of
urban concentrations at sites outside California. At remote locations, PAN
and PPN concentrations are lower than even the lowest of the urban concentra-
tions '(by a factor of 3 to 4).
In urban areas, hydrogen peroxide (H_0?) concentrations have been reported
to range from < 0.5 ppb in Boulder, Colorado (Heikes et a!., 1982) to < 180 ppb
in Riverside, California (Bufalini et al., 1972). In nonurban areas, reported
concentrations ranged from 0.2 ppb near Boulder, Colorado, in 1978 (Kelly
et al., 1979) to _< 7 ppb 54 km southeast of Tucson, Arizona (Farmer and Dawson,
1982). These nonurban data were obtained by the 1umino! chemiluminescence
technique (see chapter 5). The urban data were obtained by a variety of
methods, including the luminol chemiluminescence, the titanium (IV) sulfate
8-quinolinol, and other wet chemical methods (see chapter 5).
The higher concentrations of FLO- reported in the literature must be
regarded as especially problematic, since FTIR measurements of ambient air
have not demonstrated unequivocally the presence of H_02 even in the high-
oxidant atmosphere of the Los Angeles area. The limit of detection for a
1-km-pathlength FTIR system is around 0.04 ppm (chapter 5); FTIR is capable of
measuring concentrations of hLOp if it is present above the limit of detection.
6.8.5.2 Patterns. The patterns of formic acid (HCOOH), PAN, PPN, and HJ)^ may
be summarized fairly succinctly. They bear qualitative but not quantitative
resemblance to the patterns already summarized for ozone concentrations.
Qualitatively, diurnal patterns are similar, with peak concentrations of each
of these occurring in close proximity to the time of the ozone peak. The
correspondence in time of day is not exact, but is close. As the work of
Tuazon et al. (1981) at Claremont, California, demonstrates (see Figures 6-37
and 6-43), ozone concentrations return to baseline levels faster than the
concentrations of PAN, HCOOH, or H?0? (PPN was not measured). The diurnal
patterns of PAN were reported in earlier criteria documents. Newer data
merely confirm those patterns.
Seasonally, winter concentrations (third and fourth quarters) of PAN are
lower than summer concentrations (second and third quarters). The winter
concentrations of PAN are proportionally higher, however, than the winter
019SP/A 6-111 6/18/84
-------
concentrations of ozone. Data are not readily available on the seasonal
patterns of the other non-ozone oxidants.
Indoor-outdoor data on PAN are limited to one report (Thompson et al.,
1973), which confirms the pattern to be expected from the known chemistry of
PAN; that is, it persists longer indoors than ozone. Data are lacking for the
other non-ozone oxidants.
6.8.6 Relationship Between Ozone and Other Photochemical Oxidants
The relationship between ozone concentrations and the concentrations of
PAN, PPN, HpO^, and HCOOH is important only if these non-ozone oxidants are
shown to produce adverse health or welfare effects, singly, in combination with
each other, or in various combinations with ozone. If only ozone is shown to
produce adverse health or welfare effects, then only ozone needs to be con-
trolled. If any or all these other four oxidants is shown to produce adverse
health or welfare effects, then it, or they, will also have to be controlled.
Since ozone and all four of the other oxidants arise from reactions among the
same organic and inorganic precursors, an obvious question is whether the con-
trol of ozone will also result in the control of the other four oxidants.
Chapters 7 through 9 document what is known about the welfare effects of
PAN. No data are available regarding the possible welfare effects of HCOOH,
HJK, or PPN. Chapters 10 through 13 document what is known about the health
effects of PAN and H?0~. Formic acid is not covered because of extremely
limited aerometric data and no health effects data pertinent to the trace
quantities of formic acid measured in the ambient air. No health effects data
are available for PPN. One report that PBzN is a potent lachrymator is not
discussed in the health effects chapters since no reliable data indicate its
presence in ambient air, even in high-oxidant areas. The health effects data
reported in chapter 10 on H202 show that all levels tested to date are orders
of magnitude above even the highest concentrations reported for ambient air;
and, as noted above, the highest concentrations are not strongly documented.
Thus, the brief discussion below focuses on the relationship between ozone and
PAN concentrations in ambient air.
The most straightforward evidence of the lack of a quantitative, monotonic
relationship between ozone and the other photochemical oxidants is the range
019SP/A 6-112 6/18/84
-------
of PAN-to-0., and, indirectly, of PAN-to-PPN ratios presented in the review of
Altshuller (1983) and summarized in this chapter. Ratios of PAN concentrations
to ozone concentrations are summarized in Table 6-17, derived from Altshuller
(1983). The correspondence of PAN and ozone concentrations is not exact but
is similar for most locations at which both pollutants have been measured in
the same study. Disparities between locations point up the lack of a consis-
tent quantitative relationship. Likewise, disparities between the ratio of
the average concentrations of the two pollutants and the ratio of their concen-
trations when ozone is at its maximum level also point up the lack of a mono-
tonically quantitative relationship.
Certain other information presented in this chapter bears out the lack of
a monotonic relationship between ozone and PAN. Not only are ozone-PAN rela-
tionships not consistent between different urban areas, but they are not
consistent in urban versus nonurban areas, in summer versus winter, in indoor
versus outdoor environments, or even, as the ratio data show, in location,
timing, or magnitude of diurnal peak concentrations within the same city.
Data obtained in Houston by Jorgen et al. (1978) show variations in peak
concentrations of PAN among three separate monitoring sites in the same city.
Temple and Taylor (1983) have shown that PAN concentrations are a greater
percentage of ozone concentrations in winter than in the remainder of the year
in California (see chapter 6). Lonneman et al. (1976) demonstrated that
PAN-to-(L ratios are considerably lower in nonurban than in urban areas.
Thompson et al. (1973), in what is apparently the only published report on
indoor concentrations of PAN, showed that PAN persists longer than ozone
indoors. (This is to be expected from its lower reactivity with surfaces and
its enhanced stability at cooler-than-ambient temperatures such as found in
air-conditioned buildings.) Tuazon et al. (1981b) demonstrated that PAN persists
in ambient air longer than ozone, its persistence paralleling that of nitric
acid, at least in the locality studied (Claremont, CA). Reactivity data
presented in the 1978 criteria document for ozone and other photochemical
oxidants indicated that all precursors that give rise to PAN also give rise to
ozone. The data also showed, however, that not all precursors giving rise to
ozone also give rise to PAN. Not all that give rise to both are equally
reactive toward both, however; and therefore some precursors preferentially
give rise, on the basis of units of product per unit of reactant, to more of
one product than the other (U.S. Environmental Protection Agency, 1978).
019SP/A 6-113 6/18/84
-------
Altshuller (1983) prepared for the U.S. Environmental Protection Agency a
comprehensive review and analysis of concentrations and relationships for
ozone and other smog components, including PAN. The smog components he reviewed
relative to ozone included aldehydes, aerosols, and nitric acid, as well as
the non-ozone oxidants covered in this chapter. It must be emphasized that
Altshuller examined the issue of whether ozone could serve as an abatement
surrogate for all photochemical products, not just the subset of concern in
this document. His conclusion was that "the ambient air measurements indicate
that ozone may serve directionally, but cannot be expected to serve quantita-
tively, as a surrogate for the other products" (Altshuller, 1983). He found a
greater correspondence between aldehydes and their organic precursors than be-
tween aldehydes and ozone. The correspondence between ozone and PAN concentra-
tions (as well as PPN, H^Op, and HCOOH) is greater by far than the ozone-aldehyde
relationship.
In summary, the significance for public health or welfare of the imposition
of an additional oxidant burden from non-ozone oxidant rests on the answers to
three basic questions:
1. Do PAN, PPN, H-Op, or HCOOH, singly or in combination, elicit
adverse of potefftfally adverse responses?
2. Do any or all of these non-ozone oxidants act additively or synergis-
tically in combination with ozone to elicit adverse or potentially
adverse responses? Do any or all act antagonistically with ozone?
3. Do the time course magnitude of the concentrations of these non-
ozone oxidants parallel the time course and magnitude of ozone
concentrations in the ambient air?
Given the information on health and welfare effects presented in subsequent
chapters, coupled with the aerometric data presented in this chapter, the
relationship between ozone and PAN concentrations is the specific relationship
of most concern in this document.
019SP/A 6-114 6/18/84
-------
6.9 REFERENCES FOR CHAPTER 6
Aldaz, L. (1967) Surface air radioactivity and ozone at Amundsen-Scott Station
(90°), Antarctica. Nature (London) 215: 722-723.
Altshuller, A. P. (1975) Evaluation of oxidant results at CAMP sites in the
United States. J. Air Pollut. Control Assoc. 25: 19-24.
Altshuller, A. P. (1983) Measurements of the products of atmospheric photo-
chemical reactions in laboratory studies and in ambient air—relation-
ships between ozone and other products. Atmos. Environ. 17: 2383-2427.
Berk, J. V.; Young, R. A.; Brown, S. R.; Hollowell, C. D. (1981) Impact of
energy-conserving retrofits on indoor air quality in residential housing.
Presented at the 74th Annual Meeting of the Air Pollution Control Assoc-
ciation; June; Philadelphia, PA. Berkeley, CA: University of California,
Lawrence Berkeley Laboratory; report no. LBL-12189. Cited in: Yocom, J. E
(1982).
Berk, J. V.; Young, R.; Hollowell, C. D.; et al. (1980) The effects of energy-
efficient ventilation rates on indoor air quality in an Ohio elementary
school. Berkeley, CA: University of California, Lawrence Berkeley
Laboratory; report no. LBL-10223, UC~95d, EEB-Vent 80-9. Cited in:
Yocum, J. E. (1982)
Brennan, E. (1980) PAN concentrations in ambient air in New Brunswick, NJ.
Final Report. Prepared under Grant 805827 for U.S. Environmental Protec-
tion Agency, Research Triangle Park, NC.
Bufalini, J. J.; Gay, B. W., Jr.; Brubaker, K.L. (1972) Hydrogen peroxide
formation from formaldehyde photooxidation and its presence in urban
atmospheres. Environ. Sci. Techno!. 6: 816-821.
Canadian Department of Forestry and Rural Development. (1967) Unpublished
data. Commonwealth of Canada, Ottawa. Cited in: U.S. Department of
Health, Education and Welfare (1970).
Cherniak, E. A.; Corkum, R. (1981) Continuous monitoring of PAN in the polluted
troposphere. Brock University, St. Catherines, Ontario, Canada. Cited
in: Temple and Taylor (1983).
Cleveland, W. S.; Graedel, T, E.; Kleiner, B. (1976a) Photochemical air pollu-
tion: transport from the New York City area into Connecticut and Massa-
chusetts. Science (Washington, DC) 191: 179-181.
Cleveland, W. S.; Guarino, R.; Kleiner, B.; McRae, J. E.; Warner, J. L. (1976b)
The analysis of the ozone problem in the northeast United States. In:
Specialty Conference on: Ozone/Oxidants — Interactions with the Total
Environment. Pittsburgh, PA: Air Pollution Control Association; pp.
109-120.
Coffey, P. E.; Stasiuk, W. N. (1975a) Evidence of atmospheric transport of
ozone into urban areas. Environ. Sci. Techno!. 9: 59-62.
Coffey, P. E. ; Stasiuk, W. N. (1975b) Urban ozone: its local and extraregional
components. Presented at the 79th Annual National Meeting, American
Institute of Chemical Engineers; March; Houston, Texas. Paper No. 60C.
019AA/B 6-115 6/18/84
-------
Coffey, P. E.; Stasiuk, W. N.; Mohnen, V. (1977) Ozone in rural and urban
areas of New York State. In: Dimitriades, B., ed. International Confer-
ence on Photochemical Oxidant Pollution and Its Control: Proceedings.
Volume 1. Research Triangle Park, NC: U.S. Environmental Protection
Agency; pp. 89-96; EPA report no. EPA-600/3-77-001a.
Cox, R. A.; Roffey, M. J. (1977). Thermal decomposition of peroxyacetyl nitrate
in the presence of nitric oxide. Environ. Sci. Techno!. 11: 900-906.
Darley, E. F.; Kettner, K. A.; Stevens, E. R. (1963) Analysis of peroxyacyl
nitrates by gas chromatography with electron capture detection. Anal.
Chem. 35: 589-591.
Evans, G.; Finkelstein, P.; Martin, B.; Possiel, N.; Graves, M. (1982) The
National Air Pollution Background Network, 1976-1980. Research Triangle
Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/4-82-058.
Farmer, J. C.; Dawson, G. A. (1982) Condensation sampling of soluble atmos-
pheric trace gases. J. Geophys. Res. 87: 8931-8942.
Garland, J. A. (1977) The dry deposition of sulfur dioxide to land and water
surfaces. Proc. R. Soc. London Ser. A 354: 245-268.
Garland, J. A.; Penkett, S. A. (1976) Absorption of peroxyacetyl nitrate and
ozone by natural surfaces. Atmos. Environ. 10: 11-27.
Grosjean, D. (1981) Critical evaluation and comparison of measurement methods
for nitrogenous compounds in the atmosphere. Final report, A 706-05 for
Coordinating Research Council. Environmental Research and Technology.
Westlake Village, CA.
Grosjean, D. (1983) Distribution of atmospheric nitrogenous pollutants at a
Los Angeles area smog receptor site. Environ, Sci. Techno!. 17: 13-19.
Hanst, P. L.; Wilson, W. E.; Patterson, R. K.; Gay, B. W. Jr.; Chaney, L. W.
(1975) A spectroscopic study of California smog. Research Triangle Park,
NC: U.S. Environmental Protection Agency; EPA report no. EPA-650/4-75-006.
Hanst, P. L.; Wong, N. W.; Bragin, J. (1982) A long path infrared study of Los
Angeles smog. Atmos. Environ. 16: 969-981.
Heikes, B. G.; Lazrus, A. L.; Kok, G. L.; Kunen, S. M.; Gandrud, B. W.; Gitlin,
S. N.; Sperry, P. D. (1982) Evidence of aqueous phase hydrogen peroxide
synthesis in the troposphere. J. Geophys. Res. 87: 3045-3051.
Heuss, J. M.; Glasson, W. A.; (1968) Hydrocarbon reactivity and eye irritation.
Environ. Sci. Technol. 2: 1109-1116.
Hill, A. C.; Chamberlain, E. M. (1976) The removal of water-soluble gases from
the atmosphere by vegetation. Proceedings of Symposium on Atmosphere-
Surface Exchange of Particles and Gases. Richland, WA: ERDA Symposium
Series; pp. 153-170. Available from National Technical Information
Service, Springfield, VA. Cited in: McMahon, T. A.; Oenison, P. J.
Review paper: Empirical atmospheric deposition parameters-a survey.
Atmos. Environ. 13: 571-585.
019AA/B 6-116 6/18/84
-------
Holdren, M. W.; Spicer, C. W.; Hales, J. M. Peroxyacetyl nitrate solubility
and decomposition rate in water. Atmos. Environ. (In press).
Hunt, W. F.; Curran, T. C. (1982) National and regional trends in ambient
ozone measurements, 1975-1981. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Office of Air Quality Planning and Standards.
Isaksen, I. S. A.; Hov, 0.; Hesstvetlt, E. (1978) Ozone generation over rural
areas. Environ. Sci. Techno!. 12: 1279-1284.
Jorgen, R. T.; Meyer, R. A.; Hughes, R. A. (1978) Routine peroxyacetyl nitrate
(PAN) monitoring applied to the Houston Area Oxidant Study. Presented at
71st Annual Meeting of the Air Pollution Control Association; June;
Houston, TX. Paper No. 78-50.1.
Junge, C. E. Air Chemistry and Radioactivity. Vol. 4. In: Mieghem, J., ed.
International Geophysics Series. New York, NY: Academic Press.
Kelly, T. V.; Stedman, D. H.; Kok, G. L. (1979) Measurements of H202 and HNO-
in rural air. Geophys. Res. Lett. 6: 375-378.
Kok, G. L.; Darnall, K. R.; Winer, A. M.; Pitts, J. N., Jr.; Gay, B. W., Jr.;
(1978) Ambient air measurements of hydrogen peroxide in the California
South Coast Air Basin. Environ. Sci. Technol. 12: 1077-1080.
Kress, L. W.; Miller, J. E. (1983) Impact of ozone on soybean yield. J.
Environ. Qual. 12: 276-281.
Lefohn, A. S. (1982) The identification and description of ozone monitoring
sites which represent high yield reduction potential for agricultural
crops. Prepared for U.S. Environmental Protection Agency, Corvallis, OR.
A.S.L. and Assoc. Helena, MT.
Lewis, T. E.; Brennan, E.; Lonneman, W. A. (1983) PAN concentrations in ambient
air in New Jersey. J. Air Pollut. Control Assoc. 33: 885-887.
Logan, J. A.; Prather, M. J.; Wofsy, S. C.; McElroy, M. B. (1981) Tropospheric
chemistry: a global perspective. J. Geophys. Res. 86: 7210-7254.
Lonneman, W. A.; Bufalini, J. J.; Seila, R. L. (1976) PAN and oxidant measure-
ment in ambient atmosphere. Environ. Sci. Technol. 10: 347-380.
Martinez, J. R.; Singh, H. B. (1979) Survey of the role of NO in nonurban
ozone formation. Prepared by SRI International for U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA report no.
EPA-450/4-79-035.
Mayrsohn, H.; Brooks, C. (1965) The analysis of PAN by electron capture gas
chromatography. Presented at the Western Regional Meeting of the American
Chemical Society, Los Angeles, CA.
McKee, H. C. (1961) Atmospheric ozone in northern Greenland. J. Air Pollut.
Control Assoc. 11: 562-565.
019AA/B 6-117 6/18/84
-------
Moschandreas, D. J.; Stark, J. W. C.; McFadden, J. E.; Morse, S. S. (1978)
Indoor air pollution in the residential environment. Vol. I. Data
collection, analysis, and interpretation. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-600/7-78-229a.
Moschandreas, D. J.; Zabransky, J.; Petton, D. J. (1981) Comparison of indoor
and outdoor air quality. Prepared by Geomet, Inc., for Electric Power
Research Institute, Palo Alto, CA, on Research Project 1309. Report No.
EA-1733.
Mueller, F. X.; Loeb, L. ; Mapes, W. H. (1973) Decomposition rates of ozone in
living areas. Environ. Sci. Techno!. 7: 342-346.
Nieboer, H.; van Ham, J. (1976) Peroxyacetyl nitrate (PAN) in relation to
ozone and some meteorological parameters at Delft in the Netherlands.
Atmos. Environ. 10: 115-120.
Nielsen, T.; Samuelson, U.; Grennfelt, P.; Thomsen, E. L. (1981) Peroxyacetyl
nitrate in long-range transported polluted air. Nature (London) 293:
553-555.
Odishaw, H. (1961) International Geophysical Year. Science (Washington, DC)
129: 14-25.
Pitts, J. N., Jr.; Grosjean, D. (1979) Detailed characterization of gaseous
and size-resolved particulate pollutants at a south coast air basin
smog receptor site: levels and modes of formation of sulfite, nitrate
and organic particulates and their implications for control strate-
gies. U.S. Department of Commerce, National Technical Information
Service, PB-301 294, Springfield, VA 22161.
Pratt, G. C. ; Hendrickson, R. C.; Chevone, B. I.; Christopherson, D. A.;
O'Brien, M. V,; Krupa, S. V. (1983) Ozone and oxides of nitrogen in the
rural upper-midwestern U.S.A. Atmos. Environ. 17: 2013-2023.
Renzetti, N. A.; Bryan, R. J. (1961) Atmospheric sampling for aldehydes and
eye irritation in Los Angeles smog - 1960. J. Air Pollut. Control Assoc.
11: 421-424.
Sabersky, R. H. ; Sinema, D. A.; Shair, F. H. (1973) Concentrations, decay
rates, and removal of ozone and their relation to establishing clean
indoor air. Environ. Sci. Technol. 1: 347-353.
Schuck, E. A.; Pitts, J. N., Jr.; Swan, J. K. (1966) Relationships between
certain meteorological factors and photochemical smog. Int. J. Air Water
Pollut. 10: 689-711.
Seiler, W.; Fishman, J. (1981) The distribution of carbon monoxide and ozone
in the free troposphere. J. Geophys. Res. 86: 7255-7265.
Shair, F. H.; Heitner, K. L. (1974) Theoretical model for relating indoor
pollutant concentrations to those outside. Environ. Sci. Technol. 8:
444.
019AA/B 6-118 6/18/84
-------
Singh, H. B.; Hanst, P. W. (1981) Peroxyacetyl nitrate (PAN) in the unpolluted
atmosphere: an important reservoir for nitrogen oxides. Geophys. Res.
Lett. 8: 941-944.
Singh, H. B.; Salas, L. J. (1983) Methodology for the analysis of peroxyacetyl
nitrate (PAN) in the unpolluted atmosphere. Atmos. Environ. 17: 1507-1516.
Singh, H. B.; Salas, L. J.; Shigeishi, H.; Smith, A. J.; Scribner, E.; Cavanagh,
L. J. (1979) Atmospheric distributions, sources, and sinks of selected
halocarbons, hydrocarbons, SFfi and N~0. Research Triangle Park, NC:
U.S. Environmental Protection Agencyf EPA report no. EPA-600/3-79-107.
Singh, H. B.; Salas, L. J.; Smith, A. J.; Shigeishi, H. (1981) Measurements of
some potentially hazardous organic chemicals in urban atmospheres.
Atmos. Environ. 15: 601-612.
Singh, H. B., Salas, L. J.; Stiles, R.; Shigeishi, H. (1982) Measurements of
hazardous organic chemicals in the ambient atmosphere. Report on EPA
Cooperative Agreement 805990. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Environmental Sciences Research Laboratory.
Smith, W. J. (1981) New York State Air Monitoring Data Report for the Northeast
Corridor Regional Modeling Project. Albany, NY: New York State Department
of Environment and Conservation.
Spicer, C. W. (1977) The fate of nitrogen oxides in the atmosphere. In:
Pitts, J. N., Jr.; Metcalf, R. L.; Lloyd, A. C.; eds. Advances in Envi-
ronmental Science and Technology, Vol. 7. New York, NY: John Wiley and
Sons; pp. 163-261.
Spicer, C. W.; Gemma, J. L; Joseph, D. W.; Sticksel, P. R.; Ward, G. F. (1976)
The transport of oxidant beyond urban areas. Prepared for U.S. Environ-
mental Protection Agency, Research Triangle Park, NC. Publication No.
EPA-600/3-76-018.
Spicer, C. W.; Holdren, M. W.; Keigley, G. W. (1983) The ubiquity of peroxy-
acetyl nitrate in the lower atmosphere. Atmos. Environ. 17: 1055-1058.
Taylor, 0. C. (1969) Importance of peroxyacetyl nitrate (PAN) as a phytotoxic
air pollutant. J. Air. Pollut. Control Assoc. 19: 347-351.
Temple, P. J.; Taylor, 0. C. (1983) World-wide ambient measurements of peroxy-
acetyl nitrate (PAN) and implications for plant injury. Atmos. Environ.
17: 1583-1587.
Thompson, C. R.; Hensel, E. G.; Kats, G. (1973) Outdoor-indoor levels of six
air pollutants. J. Air Pollut. Control Assoc. 19: 347-351.
Tingey, D. T.; Hill, A. C. (1968) The occurrence of photochemical phytotoxi-
cants in the Salt Lake Valley. Utah Acad. Sci. Proc. 44: 387-395.
Tuazon, E. C.; Graham, A. M.; Winer, A. M.; Easton, R. R. ; Pitts, J. N., Jr.
(1978) A kilometer pathlength Fourier-transform infrared system for the
study of trace pollutants in ambient and synthetic atmospheres. Atmos.
Environ. 12: 865-875.
019AA/B 6-119 6/18/84
-------
Tuazon, E. C.; Winer, A. M.; Graham, R, A.; Pitts, J. N., Jr. (1980) Atmos-
pheric measurements of trace pollutants by kilometer-pathlength FT-IR
spectroscopy. In: Pitts, J. N.; Metcalf, R. L., eds. Advances in
Environmental Sciences and Technology. Vol. 10. New York, NY: John
Wiley and Sons; pp. 254-300.
Tuazon, E. C.; Winer, A. M.; Graham, R. A.; Pitts, J. N., Jr. (1981a) Atmos-
pheric measurements of trace pollutants: long path Fourier transform
infrared spectroscopy. Prepared for U.S. Environmental Protection Agency,
Environmental Sciences Research Laboratory, Research Triangle Park, NC.
Tuazon, E. C. ; Winer, A. M.; Pitts, J. N., Jr. (1981b) Trace pollutant concen-
trations in a multiday smog episode in the California South Coast Air
Basin by long path length Fourier-transform infrared spectrometry.
Environ. Sci. Techno!. 15: 1232-1237.
U.S. Department of Health, Education, and Welfare (1970) Air quality criteria
for photochemical oxidants. Washington, DC: National Air Pollution
Control Administration; publication No. AP-63.
U.S. Department of Interior (1967) Unpublished data. Southeastern Forest
Experiment Station. Asheville, NC. Cited in: U.S. Department of Health,
Education, and Welfare (1970).
U.S. Environmental Protection Agency (1978) Air quality criteria for ozone and
other photochemical oxidants. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; EPA report no. EPA-600/8-78-004.
U.S. Environmental Protection Agency. SAROAD data file for 1976. Office of
Air Quality Planning and Standards, Research Triangle Park, NC 27711.
U.S. Environmental Protection Agency. SAROAD data file for 1979. Office of
Air Quality Planning and Standards, Research Triangle Park, NC 27711.
U.S. Environmental Protection Agency. SAROAD data file for 1980. Office of
Air Quality Planning and Standards, Research Triangle Park, NC 27711.
U.S. Environmental Protection Agency. SAROAD data file for 1981. Office of
Air Quality Planning and Standards, Research Triangle Park, NC 27711.
>
U.S. Environmental Protection Agency (1982) Air Quality Criteria for Oxides
of Nitrogen. Research Triangle Park, NC. Publication No.
EPA-600/8-82-026.
Westberg, H.; Allwine, K.; Robinson, R. (1978) Measurement of light hydrocarbon
and oxidant transport: Houston study, 1976. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Environmental Sciences Research
Laboratory; EPA report no. EPA-600/3-78-662.
Wolff, G. T.; Lioy, P. T.; Wight, G. D.; Meyers, R. F.; Cederwall, R. T.
(1977) An investigation of long-range transport of ozone across the
midwestern and eastern United States. In: Diraitriades, B., ed. Inter-
national Conference on Photochemical Oxidant Pollution and Its Control.
Proceedings, Volume 1. Research Triangle Park, NC: U.S. environmental
Protection Agency; pp. 307-317; EPA report no. EPA-600/3-77-001a.
019AA/B 6-120 6/18/84
-------
Yocom, J. E. (1982) Indoor-outdoor air quality relationships: a critical
review. J. Air Pollut. Control Assoc. 32: 500-520.
Zagraniski, R. T.; Leaderer, B. P.; Stolwijk, J. A. J. (1979) Ambient sulfates,
photochemical oxidants, and acute adverse health effects: an epidemiolog-
ical study. Environ. Res. 19: 306-320.
019AA/B 6-121 6/18/84
------- |