AIR QUALITY CRITERIA
       FOR
    m
    m


    1

    I
    i^Ji
    I
    i
siiii=i!gi
       NITROGEN OXIDES
    «g

    I
    I
            ENVIRONMENTAL PROTECTION AGENCY

-------
   AP84
  AIR QUALITY CRITERIA
            FOR
    NITROGEN OXIDES
ENVIRONMENTAL PROTECTION AGENCY
      Air Pollution Control Office
        Washington, D. C.
           January 1971

-------
             Air Pollution Control Office Publication No. AP-84
For sale by the Superintendent of Documents, U.S. GoTerument Printing Office, Washington, D.C., 20402 - Price $1.60

-------
                                      PREFACE
  Air quality criteria tell us what science has
thus far been able to measure of the obvious
as well as the insidious effects of air pollution
on  man  and his environment.  Such criteria
provide the most realistic basis  that we pres-
ently have for determining to what point pol-
lution  levels must  be reduced  if we  are to
protect the public health and welfare.
  The criteria we can issue  at the  present
time do not tell us all that we would  like to
know;  but taking all of man's previous experi-
ence in evaluating  environmental hazards as a
guide, we can  conclude that improved knowl-
edge will  show that there are identifiable
health  and welfare hazards associated with air
pollution  levels that were previously thought
to be innocuous. As our scientific knowledge
grows, air quality  criteria will have to be re-
viewed and, in  all probability,  revised.  The
Congress has made it clear, however, that we
are expected, without delay, to make  the
most effective use of the knowledge we now
have.
  The 1967 amendments  to the Clean  Air
Act require that the Administrator, Environ-
mental Protection  Agency  ". . .  from time to
time, but as soon  as practicable, develop  and
issue to the States such criteria  of air quality
as in his  judgment may be requisite for  the
protection  of  the  public  health  and  wel-
fare . . .  Such  criteria  shall . .  . reflect  the
latest scientific knowledge useful in indicating
the kind and extent of all  identifiable  effects
on  health and welfare which may be expected
from  the presence  of  an  air  pollution
agent.. ."
   Under  the Act, the issuance  of air quality
criteria is a vital step in a program designed to
assist the  States in taking responsible tech-
nological, social, and political action to pro-
tect the public from the adverse effects of air
pollution.
  Briefly, the Act calls for the Administrator,
Environmental Protection Agency  to define
the broad atmospheric areas of the Nation in
which climate, meteorology, and topography,
all of which influence the capacity of air to
dilute and  disperse pollution,  are generally
homogeneous.
  Further, the Act requires the Administrator
to  define those  geographical regions in the
country  where air pollution is  a problem —
whether  interstate or  intrastate.  These air
quality control regions will be designated on
the basis of meteorological,  social, and politi-
cal  factors which suggest that a group of com-
munities should be treated  as a unit for set-
ting  limitations  on  concentrations  of
atmospheric  pollutants.  Concurrently, the
Administrator is  required to issue air quality
criteria for  those  pollutants he believes may
be  harmful  to health or welfare, and to pub-
lish related  information  on  the  techniques
which can be employed to control the sources
of those  pollutants.
  Once  these steps have  been taken for any
region, and  for any pollutant or combination
of  pollutants, then the  State  or  States re-
sponsible for  the  designated  region are  on
notice to develop ambient air quality stand-
ards  applicable to the region for  the pollu-
tants involved, and to develop plans of action
for implementing the standards.
  The Air  Pollution Control Office of EPA
will  review,  evaluate,  and  approve  these
standards and plans; and,  once they are ap-
proved,  the States will be  expected to take
action to  control pollution  sources in the
manner outlined in their plans.
  At the direction of the Administrator of
the Environmental Protection Agency, the Air
Pollution Control  Office has established ap-
propriate programs to  carry  out the several
                                             in

-------
Federal responsibilities specified in the legisla-
tion. Air Quality Criteria information has now
been published on several pollutants: particu-
late matter, sulfur oxides, carbon monoxide,
photochemical oxidants, and hydrocarbons.
  This  publication, Air  Quality Criteria for
Nitrogen Oxides, is the result of extensive and
dedicated effort on the part of many persons
— so  many that it is  not practical to name
each of them.
  In  accordance with the Clean Air Act, a
National Air Quality  Criteria Advisory Com-
mittee was established, having a membership
broadly representative of industry, univers-
ities,  conservation interests,  and all levels of
government. The committee provided invalu-
able advice on policies and procedures under
which to  issue  criteria,  and provided major
assistance in drafting this  document.

   Expert consultants were  retained to draft
portions  of this  document; other segments
were  drafted  by  staff members of the Air
Pollution  Control Office of EPA. After the
initial drafting,  there followed a sequence of
review and revision by the committee, as well
as by individual reviewers specially selected
for their  competence and  expertise in the
many fields of science and technology related
to the problems of atmospheric pollution by
hydrocarbons.  These  efforts, without which
this document could not have been completed
successfully,  are  acknowledged individually
on the following pages.
  As also required by the 1967 amendments
to the  Clean Air Act, appropriate Federal
departments and  agencies, also listed on the
following  pages,  were  consulted  prior  to
issuing this criteria document. A Federal con-
sultation  committee,  comprising  members
designated  by the heads of 17 departments
and  agencies, reviewed  the  document, and
met with staff personnel of the Air  Pollution
Control Office to  discuss their comments.

  The Administration is pleased to  acknowl-
edge the efforts of each of the persons specif-
ically named, as well as  the many not named,
who contributed  to  the publication of this
volume. Their participation  does not neces-
sarily imply their complete endorsement of all
the conclusions presented in the document,
however, for in the last analysis APCO alone
retains full responsibility for its content.
                     JOHN T. MIDDLETON
                       Acting Commissioner
                 Air Pollution Control Office
            Environmental Protection Agency
                                            IV

-------
               NATIONAL  AIR QUALITY  CRITERIA
                       ADVISORY  COMMITTEE
                                     Chairman
                             Dr. Delbert S. Earth, Director
                           Bureau of Criteria and Standards
                             Air Pollution Control Office
                           Environmental Protection Agency
 Dr. Seymour Calvert*
 Dean
 School of Engineering
 University of California at Riverside
 Riverside, California

 Dr. Adrian Ramond Chamberlain*
 President
 Colorado State University
 Fort Collins, Colorado

 Dr. James P. Dunn
 Staff Medical Director
 Western Electric Company, Inc.
 New York, New York

 Dr. David M. Gates*
 Director
 Missouri Botanical Gardens
 St. Louis, Missouri

 Dr. Neil V. Hakala
 President, Esso Research and
 Engineering Company
 Linden, New Jersey

 Dr. Ian T. Higgins
 Professor, School of Public Health
 Department of Community Health Services
 The University of Michigan
 Ann Arbor, Michigan

 Dr. Donald A. Jensen*
 Director, Automotive  Emissions Office
 Engineering Staff
 Ford Motor Company
Dearborn, Michigan
Dr. Herbert E. Klarman*
Department of Environmental Medicine
and Community Health
Downstate Medical Center
State University of New York
Brooklyn, New York
Frank J. Massey, Jr.
Professor, "School of Public Health
University of California
Los Angeles, California
Dr. George L.  McNew
Director, Boyce Thompson
Institute for Plant Research, Inc.
Yonkers, New York

Mr. Alexander Rihm, Jr.
Assistant Commissioner
Division of Air Resources
Department of Health
State of New York
Albany, New York
Dr. Frederick Sargent II
Provost
Western Washington State College
Bellingham, Washington
Dr. Vincent J. Schaefer
Director, Atmospheric Sciences
Research Center
State University of New York
Schenectady, New York
Dr. Warren Winkelstein, Jr.
Professor and  Head
Division of Epidemiology
School of Public Health
University of California at Berkeley
Berkeley, California
"Term expired June 30, 1970.

-------
                  CONTRIBUTORS AND  REVIEWERS
Dr. David M. Anderson
Assistant Manager, Environmental
Quality Control
Bethlehem Steel Corporation
Bethlehem, Pennsylvania

Mr. Roynal W. Andrews
Chief Environmental Engineer
Aluminum Company of America
Pittsburgh, Pennsylvania

Dr. Mario C. Battigelli
Department of Medicine
School of Medicine
University of North Carolina
Chapel Hill, North Carolina

Mr. Francis E.  Blacet
Department of Chemistry
University of California at
Los Angeles
Los Angeles, California

Dr. Fred W. Bowditch
Director, Emission Control
General Motors Technical Center
Warren, Michigan

Dr. Leslie A. Chambers
Director, Institute of Environmental
Health
School of Public Health of Houston
University of Texas
Houston, Texas

Mr. Robert L.  Chass
Chief, Deputy  Air Pollution Control
Officer
Los Angeles County Air Pollution
Control District
Los Angeles, California
Dr. R. S. Danchik
Analytical Chemistry Division
Alcoa Research Laboratories
New Kensington, Pennsylvania
Dr. Richard Ehrlich
Director, Life Sciences Research
Illinois Institute of Technology
Research Institute
Chicago, Illinois
Dr. W. L. Faith
Consulting Chemical Engineer
San Marino, California
Dr. Hans Falk
Associate Director for Laboratory
Research
National Institute of Environmental Health
Sciences
Research Triangle Park, North Carolina
Dr. Gustave Freeman, Jr.
Life Sciences Division
Stanford Research Institute
Menlo Park, California


Mr. James R. Garvey
President and Director of Research
Bituminous Coal Research, Inc.
Pittsburgh, Pennsylvania


Mr. J. L. Gilliland
Technical Director
Ideal Cement Company
Denver, Colorado
                                            VI

-------
Dr. John R. Goldsmith
Chief, Environmental Epidemiology
Unit
California Department of Public
Health
Berkeley, California

Mr. Walter J. Hamming
Chief Air Pollution Analyst
Los Angeles County Air Pollution
Control District
Los Angeles, California

Mr. Charles M. Heinen
Chief Engineer, Materials
Laboratories
Product Planning & Development Staff
Chrysler Corporation
Detroit, Michigan

Dr. Charles H. Hine
Clinical Professor of Occupational
& Environmental Medicine
University of California Medical
School
San Francisco, California

Dr. Kenneth D. Johnson
Staff Engineer, Air Quality
Program
Manufacturing Chemists' Association
Washington, D. C.

Dr. Charles G. Kramer
Associate Medical Director,
Midland Division
The Dow Chemical Company
Midland, Michigan

Dr. Lawrence D. Kornreich
Director, Chemistry Division
The Research Corporation of New England
Hartford, Connecticut

Dr. Paul Kotin
Director, National Institute of
Environmental Health Sciences
Research Triangle, North Carolina
Dr. Philip A. Leighton
Professor Emeritus
Stanford University


Mr. Arthur Levy
Chief, Air Water and Solid Waste
Chemistry Division
Batelle Memorial Institute
Columbus, Ohio

Dr. Joseph A. Lieberman
Assistant Administrator for Research
and Development
Environmental Health Service
Department of Health, Education, and
Welfare
Washington, D. C.

Dr. H. N.  MacFarland
Professor and Director, Centre
of Research on Environmental Quality
York University
Downsview, Ontario, Canada

Dr. Herbert C. McKee
Assistant Director, Department of
Chemistry and Chemical Engineering
Southwest Research Institute
Houston, Texas

Dr. Elmer Robinson
Chairman, Environmental Research
Department
Stanford Research Institute
Menlo Park, California

Dr. Lawrence M. Roslinski
Professional Associate
Advisory  Center on Toxicology
National Research Council

 Mr. Francis Scofield
 Vice President, Technical Affairs
 National Paint, Varnish, and
 Lacquer Association, Inc.
 Washington, D. C.
                                            VII

-------
Dr. R. W. Scott
Coordinator for Conservation
Technology
Esso Research and Engineering
Company
Linden, New Jersey

Mr. William J. Stanley
Director, Chicago Department of
Air Pollution Control
Chicago, Illinois
Dr. Edgar R. Stephens
Research Chemist, Statewide Air
Pollution Research Center
University of California at Riverside
Riverside, California
Mr. Morton Sterling
Air Pollution Control Division
Wayne County Department of Health
Detroit, Michigan

Dr. Herbert E. Stokinger
Chief, Laboratory of Toxicology
and Pathology
Bureau of Occupational  Safety and
Health
Environmental Control Administration
Department of Health, Education,
and Welfare
Cincinnati, Ohio

Dr. E. Suriani
Research Scientist, Division of
Research
Cottrell Environmental Systems, Inc.
Bound Brook, New Jersey

Dr. O. Clifton Taylor
Associate Director, Statewide Air
Pollution Research Center
University of California at Riverside
Riverside, California
Dr. Ralph C. Wands
Director, Advisory Center on
Toxicology
National Research Council
Washington, D. C.


Dr. Peter O. Warner
Laboratory Supervisory, Air Pollution
Control Division
Wayne County Department of Health
Detroit, Michigan
Dr. Lowell G. Wayne
Consultant, Air Pollution Chemistry
Los Angeles, California
Dr. Eugene E. Weaver
Product Development Group
Ford Motor Company
Detroit, Michigan
Mr. Melvin Weisburd
Manager, Environmental Health Systems
System Development Corporation
Santa Monica, California


Dr. Bernard Weinstock
Manager and Senior Scientist, Fuel
Sciences Department
Ford Motor Company
Dearborn, Michigan
Dr. John E. Yocom
Director, Engineering and
Technical Programs
The Travelers Research Corporation
Hartford, Connecticut
                                            Vlll

-------
       FEDERAL AGENCY LIAISON  REPRESENTATIVES
Department of Agriculture
Dr. Theodore C. Byerly
Assistant Director of Science and
   Education
Office of the Secretary

Department of Commerce
Dr. James R. Hibbs
Advisor on Environmental Quality

Department of Defense
Lt. Col. Herbert E. Bell
Staff Assistant, BCS
Office of the Assistant Secretary
   of Defense

Department of Housing and Urban
   Development
Mr. Samuel C. Jackson
Assistant Secretary for Metropolitan
   Development

Department of Interior
Mr. William L. Crentz
Director, Energy
Bureau of Mines

Department of Justice
Mr. Walter Kiechel, Jr.
Assistant Chief, General Litigation
   Section
Land & Natural Resources Division

Department of Labor
Mr. Robert D. Gidel
Deputy Director, Bureau of Labor
   Standards
Railway Labor Building

Post Office Department
Mr. W. Norman Meyers
Environmental Coordinator
Bureau of Facilities
Department of Transportation
Mr. William H. Close
Chief, Research Division
Office of Noise Abatement

Department of Treasury
Mr. Gerald M. Brannon
Director, Office of Tax Analysis

U. S. Atomic Energy Commission
Dr. Martin B. Biles
Director of Division of Operational
   Safety

Federal Power Commission
Mr. George Tomlinson
Acting Chief, Bureau of Power

General Services Administration
Mr. Thomas E. Crocker
Director of Repair & Improvement
   Division
Public Buildings Service

National Aeronautics & Space
   Administration
Maj. General R. H. Curtin, USAF (Ret.)
National Aeronautics & Space
   Administration

National Science  Foundation
Dr. Eugene W. Bierly
Program Director for Meteorology
Division of Environmental Sciences

Tennessee Valley Authority
Dr. F. E. Gartrell
Assistant Director of Health

Veterans Administration
Mr. Gerald M. Hollander
Director of Architecture & Engineering
                                           IX

-------

-------
                      TABLE OF CONTENTS

Chapter                                                               Page

LIST OF FIGURES	xv
LIST OF TABLES  	xvii
1.  INTRODUCTION   	1-1
2.  PROPERTIES OF NITROGEN  OXIDES AND PHYSICAL EFFECTS  ON LIGHT
   TRANSMISSION	2-1
   A.  INTRODUCTION  	2-1
   B.  OCCURRENCE   	2-1
   C.  PROPERTIES OF NITROGEN OXIDES  	2-1
      1.  Nitric Oxide (NO)  	2-1
      2.  Nitrogen Dioxide (NC>2)  	2-2
      3.  Nitrous Oxide (^O)   	2-4
      4.  Nitrogen Sesquioxide (^03)  	2-4
      5.  Nitrogen Tetroxide (^04)  	2-4
      6.  Nitrogen Pentoxide (^05)  	2-4
      7.  Other Oxides of Nitrogen	2-4
   D.  EFFECTS ON ATMOSPHERIC LIGHT TRANSMISSION   	2-4
   E.  SUMMARY	2-5
   F.  REFERENCES	2-6
3.  SOURCES AND CONTROL OF ATMOSPHERIC NITROGEN OXIDES  	3-1
   A.  INTRODUCTION  	3-1
   B.  NATURAL SOURCES   	3-1
   C.  MAN-MADE SOURCES	3-1
   D.  PRINCIPLES OF FORMATION AND CONTROL	3-3
      1.  Formation   	3-3
      2.  Control  	3-4
   E.  SUMMARY	3A
   F.  REFERENCES	3-5
4.  CHEMICAL INTERACTIONS OF NITROGEN OXIDES IN THE ATMOSPHERE   ...  4-1
   A.  INTRODUCTION  	4-1
   B.  OBSERVATIONAL MODEL FOR HC-NOX -OXIDANT RELATIONSHIPS	  4-3
      1.  Early Morning HC Relationship to Maximum Daily Oxidant	4-3
      2.  Early Morning NOX Relationship to Maximum Daily Oxidant	4-4
      3.  HC-NOx-Oxidant Relationship	4-10

     4. Limitations and Interpretation of Observational Approach	4-13

   C.  ADDITIONAL REACTIONS AND PRODUCTS OF NOX	4-18
   D.  FUTURE RESEARCH   	4-18
   E.  SUMMARY	4-19
   H.  REFERENCES	4-20

                                   xi

-------
5.  METHODS FOR MEASUREMENT OF NITROGEN OXIDES  	5-1
   A.  INTRODUCTION  	5-1
   B.  NITROGEN DIOXIDE  	5-1
      1.  Manual Methods  	5-1
      2.  Continuous Method  	5-2
      3.  Miscellaneous Methods  	5-2
      4.  Calibration  	5-3
   C.  NITRIC OXIDE  	5-3
      1.  Oxidation to NO2  	5-3
      2.  Chemiluminescent Method  	5-4
      3.  Calibration  	5-5
   D.  TOTAL NITROGEN OXIDES  	5-5
      1.  Oxidation of NO to NO2  	5-5
      2.  Chemiluminescent Method  	5-5
   E.  SUMMARY	5-6
   F.  REFERENCES	5-6
6.  ATMOSPHERIC LEVELS OF NITROGEN OXIDES  	6-1
   A.  INTRODUCTION  	6-1
   B.  TEMPORAL VARIATIONS IN NOX CONCENTRATIONS	6-1
      1.  Diurnal Patterns  	6-1
      2.  Seasonal Patterns   	6-3
      3.  Annual Trends  	6-4
      4.  Model for Relating Temporal Maxima  	6-5
   C.  EFFECT OF METEOROLOGICAL FACTORS  	6-7
   D.  OBSERVED URBAN NOX CONCENTRATIONS  	6-8
   E.  EFFECTS OF MEASUREMENT SYSTEMS ON DATA   	6-32
   F.  SUMMARY	6-37
   G.  REFERENCES  	6-37
7.  EFFECTS OF NITROGEN OXIDES ON MATERIALS 	7-1
   A.  INTRODUCTION  	7-1
   B.  EFFECTS ON TEXTILE DYES  	7-1
      1.  Acetate Rayon Fading   	7-1
      2.  Cotton and Viscose Rayon Fading (Cellulosics) 	7-3
      3.  Yellowing of Whites	7-3
   C.  EFFECTS ON TEXTILE FIBERS  	7-4
      1.  Cellulosic Fibers  	7-4
      2.  Synthetic Fibers  	7-5
   D.  EFFECTS ON NICKEL-BRASS ALLOYS	7-5
      1.  Stress Corrosion  	7-5
      2.  Protection	7-6
   E.  FUTURE RESEARCH NEEDS   	7-6
   F.  SUMMARY	7-6
   G.  REFERENCES	7-7
8.  EFFECTS OF NITROGEN OXIDES ON VEGETATION  	8-1
   A.  INTRODUCTION  	8-1
   B.  ACUTE INJURY	8-2
   C.  CHRONIC INJURY  	8-2
                                    Xll

-------
   D.  PHYSIOLOGICAL EFFECTS  	8-3
      1.  Observed Responses 	8-3
      2.  Biochemical Mechanisms  	8-3
   E.  FACTORS AFFECTING RESPONSE TO NITROGEN DIOXIDE	8-3
   F.  DOSE-INJURY RELATIONSHIP BETWEEN NITROGEN DIOXIDE AIR
      POLLUTION AND VEGETATION  	8^
   G.  NEED FOR FUTURE RESEARCH	8-4
   H.  SUMMARY	8-11
   I.  REFERENCES	8-11
9.  TOXICOLOGICAL EFFECTS OF NITROGEN OXIDES 	9-1
   A.  INTRODUCTION   	9-1
   B.  NITRIC OXIDE  	9-1
      1.  Methemoglobin Increase  	9-1
      2.  Enzyme Inhibition   	9-2
   C.  NITROGEN DIOXIDE  	9-2
      1.  Effects in Animals	9-2
      2.  Effects in Man   	9-14
   D.  OTHER OXIDES OF NITROGEN	9-17
   E.  FUTURE RESEARCH  NEEDS  	9-19
   F.  SUMMARY	9-19
   G.  REFERENCES 	9-28
10. EPIDEMIOLOGICAL APPRAISAL OF NITROGEN OXIDES  	10-1
   A.  INTRODUCTION   	10-1
   B.  EPIDEMIOLOGICAL EVIDENCE OF LONG-TERM EFFECTS   	10-1
      1.  The Cleveland Clinic Study - An Accidential High Exposure	10-1
      2.  The Chattanooga Studies - Ambient NO2 and Respiratory Illness	10-1
      3.  Czechoslovakian Study - Ambient NO2 and SO2 on Peripheral Blood	10-7
      4.  Occupational Exposure    	10-7
   C.  IMPLICATIONS OF CHATTANOOGA STUDIES	10-7
   D.  FUTURE RESEARCH  NEEDS  	10-8
   E.  SUMMARY	10-8
   F.  REFERENCES	10-9
11. SUMMARY AND CONCLUSIONS	n-1
   A.  INTRODUCTION   	1M
   B.  PROPERTIES  OF NITROGEN OXIDES AND PHYSICAL EFFECTS ON LIGHT
      TRANSMISSION   	H_l
   C.  SOURCES AND CONTROL OF ATMOSPHERIC NITROGEN OXIDES	H-2
   D.  CHEMICAL INTERACTIONS OF NITROGEN OXIDES IN THE ATMOSPHERE  . \\-2
   E.  METHODS FOR MEASUREMENT OF NITROGEN OXIDES	11-2
   F.  ATMOSPHERIC LEVELS OF NITROGEN OXIDES  	11-3
   G.  EFFECTS OF NITROGEN OXIDES ON MATERIALS	11-3
   H.  EFFECTS OF NITROGEN OXIDES ON VEGETATION	11-4
   I.  TOXICOLOGICAL EFFECTS OF NITROGEN OXIDES   	11-4
      1.  Nitric Oxide  	11-4
      2.  Nitrogen Dioxide  	11-4
   J.  EPIDEMIOLOGICAL APPRAISAL OF NITROGEN OXIDES	11-8
                                  X11I

-------
K. AREAS FOR FUTURE RESEARCH	H_8
   1. Environmental Aspects of Oxides of Nitrogen  	11-8
   2. Effects on Vegetation and Materials  	11-9
   3. Toxicity of Oxides of Nitrogen	11-9
   4. Epidemiology of Oxides of Nitrogen	11-9
L. CONCLUSIONS  	11-9
   1. Nitric Oxide  	11-9
   2. Nitrogen Dioxide  	11-10
   3. Other Nitrogen Oxide Effects  	  11-11
M. RESUME  	11-11
N. REFERENCES   	11-12
   APPENDIX  	A-l
   SUBJECT INDEX   	1-1
                                     xiv

-------
                             LIST  OF  FIGURES


Figure                                                                            Page

 2-1     Atmospheric NC>2 Photolytic Cycle   	      2-3
 2-2     NC>2 Absorption Spectrum	      2-4
 2-3     Transmittance of Visible Light at Different NC>2 Concentrations and View-
         ing Distances	     2-6
 2-4     Aerosol Attenuation  of Spectral Luminance  of Horizon  Sky and Distant
         Objects at Different NO2 and Aerosol Concentrations  	     2-7
 4-1     Maximum Daily 1-Hour-Average Oxidant Concentrations as a Function of 6-
         to 9-a.m. Averages of Total Hydrocarbon Concentrations at CAMP Stations,
         June through September, 1966  through 1968, and in Los Angeles, May
         through October 1967  	     4-4
 4-2     Maximum Daily  1-Hour-Average  Oxidants as a Function of  6- to 9-a.m.
         Averages  of Nonmethane Hydrocarbons at CAMP Stations, June through
         September,  1966 through 1968,and in Los Angeles,-May through October
         1967   	     4-5
 4-3     Maximum Daily 1-Hour-Average  Oxidant Concentrations as a Function of 6-
         to 9-a.m. Averages of  Total  Nitrogen Oxides in  Washington, D. C., June
         through September, 1966 through 1968  	     4-6
 4-4     Maximum Daily 1-Hour-Average  Oxidant Concentrations as a Function of 6-
         to 9-a.m. Average Total Nitrogen Oxides in Philadelphia, June through Sep-
         tember, 1965 through 1968  	     4-7
 4-5     Maximum Daily 1-Hour-Average  Oxidant Concentrations as a Function of 6-
         to 9-a.m. Average Total Nitrogen Oxides in Denver, June  through Septem-
         ber, 1965 through 1968	     4-8
 4-6     Upper  Limit of Maximum Daily  1-Hour-Average Oxidant Concentrations, at
         Calculated Nonmethane Hydrocarbon Concentrations  of  1.5 ppm C, as a
         Function of Average  Total Nitrogen Oxides from 6 to  9 a.m. at Three Los
         Angeles Stations,  May through  October 1967   	     4-9
 4-7     Hydrocarbon-Oxidant Envelopes  Superimposed on  Maximum Daily 1-Hour-
         Average Oxidant  Concentrations as  a Function of 6- to 9-a.m. Average of
         Total Nitrogen Oxides in Pasadena, California, May through October 1967.  .     4-10
 4-8     Approximate Isopleths  for Selected  Upper-Limit Maximum Daily  1-Hour-
         Average Oxidant  Concentrations, as a Function of  the 6- to 9-a.m. Averages
         of Nonmethane Hydrocarbons and Total Nitrogen Oxides in Philadelphia,
         Washington, D. C., and Denver, June through August,  1966 through 1968.  .     4-11
 4-9     Oxidant Isopleths from  Laboratory Experiments, Showing Effect of Varying
         Initial Precursor, Hydrocarbon  (Propylene) and Nitric Oxide, Concentrations
         on Maximum Ozone Concentrations   	     4-13
 4-10    Maximum Daily  1-Hour-Average  Oxidants as a Function of  6- to 9-a.m.
         Averages of Total Nitrogen Oxides for 2.0- to 2.4- ppm C Total Hydrocarbon

                                          xv

-------
        in Washington, D. C., Philadelphia,  and Denver, June through September,
        1965 through 1968  	     4-15
4-11    Diurnal  Variation of Mean Carbon  Monoxide Levels on Weekdays, Satur-
        days, and Sundays in Chicago, 1962 through 1964  	     4-17
5-1      Permeation Tube System	     5-4
5-2     Oxidation Schemes for Analyzing NO and NOX	     5-5
6-1      Average Daily 1-Hour Concentrations of Selected Pollutants in Los Angeles,
        California, July 19, 1965	     6-2
6-2     Diurnal  Variation in Monthly Mean 1-Hour-Average NO2 Concentrations
        From Three Urban Stations  	     6-3
6-3     Weekday and Weekend  1-Hour-Average  NO Levels in Chicago, 1962 through
        1964   	     6-3
6-4     Monthly Mean NO Concentrations at Four Urban Sites	     6-4
6-5     Monthly Mean NO2 Concentrations at Four Urban Sites	     6-5
6-6     Average and  Standard Deviation  of Daily Maximum Nitrogen Oxides and
        Nitrogen Dioxide Concentrations for Seven Locations in Los Angeles Basin. .     6-5
6-7a    Monthly Means of Daily Maximum NOX Concentrations at Los Angeles Civic
        Center, 1957 through 1961   	     6-6
6-7b    Trend Curves Fitted to Monthly  Means of Daily Maximum NOX Concentra-
        tions Corrected for Seasonal Variation,  Los Angeles Civic Center, Stations 1
        and 58, 1957 through 1961   	     6-6
6-8     Average Monthly  NO2  Concentrations in Four New Jersey Communities,
        1965 through 1968  	     6-7
6-9     Diurnal Variation of NO Levels During 1963 Stagnation in Washington, D.C.,
        October 15 through 19,  1963  	     6-8
6-10    Diurnal  Variation of NO2  Levels During Stagnation in Washington,  D. C.,
        October 15 through  19,  1963  	     6-9
6-11    Diurnal  Variation  of Oxidant Levels During Stagnation in Washington, D. C.,
        October 15 through  19,  1963  	     6-9
6-12    Frequency Distribution of 3-Hour-Average Concentrations of NOX  at Los
        Angeles CAMP Station, December 1, 1963, to December  1, 1964   	     6-32
9-1     Effect of NO2 on Respiratory Rate and  Tidal Volume of Guinea Pigs   ....     9-5
                                        xvi

-------
                              LIST OF TABLES

Table                                                                             Page

 1-1      Factors to be Considered in Developing Air Quality Criteria   	     1-2
 2-1      Physical Properties of Nitrogen Oxides	     2-5
 3-1      Summary of Nationwide Nitrogen Oxides Emissions, 1968   	     3-2
 3-2      Nitrogen Oxide Emission Sources for 22 Selected Cities   	     3-3
 3-3      Summary  of Annual  Nationwide  Emissions  of Nitrogen Oxides, 1966
         through 1968  	     3-3
 4-1      Changes in Maximum  Daily 1-Hour Average Oxidant  Concentration as a
         Function 6-to 9-a.m. Average HC and NOX Concentrations	     4-12
 6-1      Nitric Oxide Concentration at CAMP  Sites, by Averaging Time and Fre-
         quency, 1962 through 1968  	     6-10
 6-2      Nitrogen Dioxide  Concentration at CAMP Sites, by Averaging Time and
         Frequency, 1962 through 1968   	     6-12
 6-3      Nitrogen Oxides Concentration at CAMP Sites, by Averaging Time and Fre-
         quency, 1962 through 1968  	     6-14
 6-4      Nitric Oxide Concentration in California, by Averaging Time and Frequency,
         1963 through 1967  	     6-16
 6-5      Nitrogen Dioxide Concentration in California, by Averaging  Time and
         Frequency, 1963 through 1967   	     6-19
 6-6      Nitrogen Oxides Concentration in California, by Averaging Time and Fre-
         quency, 1963 through 1967  	     6-22
 6-7      Nitric Oxide Concentration in Los Angeles County, by Averaging Time and
         Frequency, 1963 through 1967   	     6-26
 6-8      Nitrogen Dioxide Concentration in Los Angeles County, by Averaging Time
         and Frequency, 1963 through 1967   	     6-28
 6-9      Nitrogen Oxides Concentration in Los Angeles County, by Averaging Time
         and Frequency, 1963 through 1967   	     6-30
 6-10    Average 24-Hour  NO2 Concentration  During  1967,  1968,  and 1969  at
         National Air Sampling Network Sites	     6-33
 6-11    Ratio of NASN to CAMP Average Yearly NO2 Measurements   	     6-37
 7-1      Breaking-Strength of Cotton Yarn Samples Exposed to Air and Sunlight in
         Berkeley, California	     7-4
 7-2      Air Pollutant Levels and Weather Measurements    	     7.4
 8-1      Chronic Injury and Physiological Effects of NO and NO2  	     8-5
 8-2      Threshold Susceptibility of Plants to Acute  Injury from Nitrogen Dioxide   .     g_6
 8-3      Projected NO2 Exposures for 5 Percent Injury Levels on Selected Vegetation     3.7
 84      Relative Plant Sensitivity to NO2  Injury   	     g_7
 8-5      Nitrogen Dioxide Dose-Response Equations	     g.jO
 9-1      Comparison of Lethal Levels of Acute Exposure of Male Rats to NO2  ....     9.2
 9-2      Time Necessary to Produce Death  in Animals Exposed to High Concen-
         trations of NO2   	     9-3

                                        xvii

-------
 9-3      Changes in Pulmonary Function in Animals Exposed to Sublethal Doses of
         NO2   	    9-4
 9-4      Pathological  Changes in Lungs  of Mice Exposed Daily  to 940 Mg/m^ (0.5
         ppm) NO2 on Various Schedules  	    9-8
 9-5      Pathological Changes in Lungs of Rats Continuously Exposed to NO2  ....    9-9
 9-6      Recognition  of NO2  Odor by Healthy  Young Adult Men After Varied
         Exposures in an Experimental Chamber  	    9-14
 9-7      Respiratory Effects of Exposure to Nitrous Fumes	    9-18
 9-8      Summary of Studies of Toxicological Effects of Nitric Oxide in Animals and
         Bacteria	    9-21
 9-9      Summary of Studies of Toxicological Effects in Animals of Short-Term NO2
         Exposure	    9-22
 9-10    Summary of Studies of Toxicological Effects in Animals of Long-Term NO2
         Exposure  	    9-25
10-1      Arithmetic Mean and 90th Percentile Concentrations of Pollutants Sampled
         for 24 Hours at Various Sites   	   1Q-2
10-2      Analysis of  Variance of Effects on Height-Adjusted  FEV0.75 of Sex of
         Child, Month of Testing, and Study Area	   10-3
10-3      Analysis of Variance of Effects on Height-Adjusted FEVg 75 of NO2 Con-
         centrations on Day of Test, School Within High NO2 Area, Sex of Child, and
         Month of Test 	   10-4
10-4      Average Biweekly Respiratory  Illness Rates Per 100  for Each  Family Seg-
         ment According to Exposure to  Oxides of Nitrogen  	   10-4
10-5      Percent  Relative  Excess of Respiratory Illness Among Family  Segments in
         Exposed Versus Control Areas During 24-Week Study  	   10-5
10-6      Distribution  of Children Reporting One or More Episodes of Bronchitis, by
         Length of Exposure	   10-6
11-1      Summary  of  Toxicological  NO2  Effects	   11-5
                                        xvm

-------
                                    CHAPTER  1.

                                 INTRODUCTION
   Pursuant to  authority  delegated  to  the
Commissioner, Air Pollution Control Office*
of the Environmental Protection Agency, Air
Quality Criteria for Nitrogen Oxides is issued
in accordance with Section  107(b)  of the
Clean Air Act (42 U.S.C. 1857-18571).
   Air quality criteria are an expression of the
scientific  knowledge  of the relationship be-
tween various concentrations of pollutants in
the air  and their adverse effects on man and
his environment. They are issued to assist the
States in developing air quality standards. Air
quality  criteria  are descriptive; that is, they
describe the effects that have  been observed
to occur when the ambient air level of a  pol-
lutant  has  reached or  exceeded  a specific
figure for a specific time period. In developing
criteria, many factors  have to be considered.
The  chemical and physical characteristics of
the pollutants and the techniques available for
measuring these characteristics must be con-
sidered,  along with exposure  time, relative
humidity, and other conditions of the envi-
ronment.  The criteria must consider the con-
tribution of all such variables to the effects of
air  pollution  on human health, agriculture,
materials, visibility, and  climate. Further, the
individual characteristics of the receptor must
be taken into account. Table 1-1 is a listing of
the major factors that need to be considered
in developing criteria, t
*Formerly the National Air Pollution Control Administra-
 tioa
f Adapted from:
  S. Calvert. Statement for air quality criteria hearings held
by the Subcommittee on Air and Water Pollution of the
U.S. Senate  Committee on Public Works, July 30, 1968,
published in "Hearings Before the Subcommittee on Air and
Water Pollution of the Committee on Public Works, United
States Senate (Air Pollution-1968, Part 2.)"
   Air quality standards are prescriptive. They
prescribe pollutant exposures that a political
jurisdiction determines  should  not  be  ex-
ceeded in a specified geographic  area, and are
used as  one  of  several factors  in  designing
legally enforceable pollutant  emission stand-
ards.
   Several oxides of nitrogen are found in the
atmosphere, but  only nitric oxide (NO)  and
nitrogen dioxide (NO2) are important as air
pollutants.  This  document focuses on them
and uses  the symbol NOX to  represent the
composite of the two. The others seldom occur
in appreciable quantities and then, only under
special  conditions.  A complex relationship
that exists among  the oxides  of nitrogen,
photochemical oxidants, and hydrocarbons is
discussed here primarily in terms of the nitro-
gen oxides. Photochemical  oxidants  and
hydrocarbons are  discussed  in  companion
documents: AP-63,  Air Quality Criteria for
Photochemical  Oxidants,  and  AP-64,  Air
Quality Criteria for Hydrocarbons.
   The Air Pollution Control  Office (APCO)
 of EPA  is  currently advocating  the
use of the  mass  per unit  volume measure to
express  atmospheric  concentrations of  air
pollutants  (e.g.,  micrograms per cubic meter,
jug/m^). In most  instances, gaseous pollutants
have been  reported on a volume-ratio basis,
that is, parts per million (ppm); almost all
data for atmospheric oxides of nitrogen were
originally recorded as ppm. Conversion from
volume (ppm) to mass (jug/m-^) units requires
a  knowledge  of  the gas density, which varies
with the  temperature and pressure of meas-
urement. In this  document 25°  C (77° F) has
been taken as standard temperature, and  760
mm Hg (atmospheric pressure at sea level) as
standard pressure.  It  should  be remembered
                                             1-1

-------
        Table 1-1. FACTORS TO BE CONSIDERED IN DEVELOPING AIR QUALITY CRITERIA
                       Factor
           Components
             Properties of pollution
Adsorbed gases
Coexisting pollutants
Kinetics of formation
Residence time
             Properties of pollutant
Concentration
Chemistry
Physical state
             Methods of measurement
Chemical analysis
  Colorimetric
  Coulometric
  Spectroscopic
  Gas chromatographic
Sampling techniques
             Exposure parameters
Time
Temperature
Pressure
Humidity
             Characteristics of receptor
Physical
Susceptibility
State of health
Rate, site, and agent of transfer to
             Responses (effects)

               Health
               Human welfare



               Vegetation


               Material



               Atmospheric
Type of illness
  Diagnosable
  Latent
Increase in general susceptibility
Victim
  Animals
  Man


Economics
Sociology
Epidemiology

Crops
Ornamental plants

Objectionable surface decomposition
Corrosion
Deterioration

Radiation
Temperature
Diffusion
Dispersion
1-2

-------
that ambient concentrations of nitrogen ox-
ides expressed in  mass units (/xg/m^) will
therefore  vary  depending on the ratios and
densities of the specific oxides involved. For
purposes of consistency and readability, all
references  to  nitrogen  oxides in this docu-
ment are expressed in terms of NC>2 mass per
unit volume, unless otherwise specified. Simi-
larly, references  to hydrocarbon (HC) and
oxidant concentrations will be in terms of
mass of methane and ozone (03) per  unit
volume, respectively.
  This publication reviews the chemical and
physical characteristics of the nitrogen oxides
and  considers  the  relative  merit of  various
analytical  methods for measuring them in the
atmosphere. It also discusses their effects on
visibility, vegetation, and materials; their toxi-
cological effects on animals  and on man; and
epidemiological studies that  assess the general
population  dose  response  and the specific
response of children to nitrogen oxides.
  In general, the terminology employed
follows  usage  recommended in the publica-
tions style guide of the American Chemical
Society. A conversion table  for mass and vol-
ume units of  measurement,  and a  subject
index are provided.
   This  document is not intended as a com-
plete, detailed literature review, and it does
not  cite every  published  article relating  to
nitrogen oxides in the ambient atmosphere.
The literature has been reviewed thoroughly,
however, for information related to the devel-
opment of  criteria through April 1970, and
the document not  only summarizes the cur-
rent scientific knowledge of nitrogen  oxides
air pollution, but  also points  up the major
deficiencies  in that knowledge  and  the needs
for further research. The  results and conclu-
sions of foreign investigations were evaluated
for their possible application to the air pollu-
tion problem in the United States and rele-
vant ones are cited.

   The technological and economic aspects  of
air pollution control are considered in com-
panion  volumes to criteria documents.  The
best methods and techniques for controlling
the sources  of nitrogen oxide emissions,  as
well as the costs of applying these techniques,
are described in: AP-67, Control Techniques
for Nitrogen Oxide Emissions from Stationary
Sources,  and AP-66, Control Techniques for
Carbon Monoxide,   Nitrogen  Oxide,  and
Hydrocarbon Emissions from Mobile Sources.
                                                                                       1-3

-------

-------
                                 CHAPTER  2.

           PROPERTIES OF NITROGEN  OXIDES AND
       PHYSICAL EFFECTS ON LIGHT TRANSMISSION
A.   INTRODUCTION
  Of the various oxides of nitrogen, the most
important as  air pollutants  are nitric oxide
(NO) and  nitrogen dioxide (NO2).  Other
known  oxides  of  nitrogen include  nitrous
oxide (N2O), nitrogen  sesquioxide (^03),
nitrogen tetroxide  (^04),  and  nitrogen
pentoxide (^05), but only N2O is present in
the atmosphere in appreciable concentrations.
  The  term NOX  is used to represent the
composite  atmospheric concentration  of
nitric oxide and nitrogen dioxide throughout
this document.

B.   OCCURRENCE
  Nitric  oxide  (NO)  and a  comparatively
small amount  of NO2 are formed under high-
temperature conditions such as those that ac-
company the burning of fossil fuels. They are
emitted to the atmosphere from automobile
exhausts, furnace stacks, incinerators,  and
vents  from   certain  chemical   processes.
Although vast  quantities are produced  by
natural biological reactions, the resultant con-
centrations  of  atmospheric  NOX  are  low
because these  reactions  take place over wide
areas.  Most of  the  NOX  produced  tech-
nologically is  in the form of NO,  which is
subsequently oxidized  in  the atmosphere to
the more toxic and irritant NO2- Normally, at
low-NO concentrations of 1.2 mg/m^ (1 ppm)
or less, the direct reaction with oxygen of the
air proceeds slowly. The oxidation of NO to
NO2 is speeded up enormously, however, by
photochemical processes  involving reactive
hydrocarbons.
  Nitrous oxide (N2O)  is not considered an
air  contaminant. There is  no evidence to
suggest  N2O  is  involved in photochemical
reactions in the lower atmosphere. Formation
of N2O is largely due to the decomposition of
nitrogen compounds by soil bacteria and to
reactions  between  nitrogen  and  atomic
oxygen or ozone in the upper atmosphere.1
The mean atmospheric concentration of N2O
is about 449 ng/m^ (0.25 ppm).2 The oxides,
N2O4, N2O3, and ^05,  are not found in
appreciable concentrations  under  urban
atmospheric  conditions.  Nevertheless,  it is
feasible  that even in small concentrations,
N2©3 and ^05  could be involved in the re-
actions  leading  to photochemical  air  pol-
lution.3

C.  PROPERTIES OF NITROGEN
    OXIDES
1.  Nitric Oxide (NO)
  Nitric oxide is a  colorless, odorless  gas,
slightly  soluble  in water. It is the primary
product formed during high-temperature com-
bustion  processes  when atmospheric oxygen
and   nitrogen  combine according to  the
following endothermic reaction:
N2
                2ND
(1)
This reaction  continues to an  equilibrium
level, dependent  upon variables  such as the
temperature of the flame, the  concentration
of each gas, and  the movement of the gases
through  different  zones of  temperatures,
pressures,  and concentrations. When rapid
cooling  follows combustion, time is inade-
quate for equilibrium to develop, and so NO
persists in the flame products.
                                         2-1

-------
2.   Nitrogen Dioxide (N(>2)
  Nitrogen dioxide is a reddish-orange-brown
gas  with a  characteristic  pungent  odor.
Though its boiling point is  21.2°C, the low
partial pressure  of NO2 in the atmosphere
restricts it to  the gas  phase at usual atmos-
pheric temperatures.  NO2 is  corrosive and
highly oxidizing,  and may be physiologically
irritating  and  toxic.  Although  NO2 reacts
with  water  to  form nitric  acid and either
nitrous acid or NO:
   2  N02  + H20 ->  HN03 +  HN02
or                                     (2)
   3  NO2 + H20 -» 2HN03  + NO,


the reaction of NC>2 with H2O has not been
shown to have any  significance in polluted
air.3
   Although NC>2 exists  in equilibrium with
its dimer, nitrogen tetroxide (^04):
                                       (3)
the  fraction  present  in  dimer  form  is
negligible at atmospheric concentrations  of
   NC>2 is a product of the termolecular re-
action of nitric oxide with oxygen:
            2 NO + O2 -> 2 NO2
(4)
Since two molecules of NO are involved, the
NO2-formation rate varies with  the square of
the NO concentration. Doubling the NO con-
centration  increases the  reaction  rate  four
times.  A  generally accepted mechanism  to
explain this NO 2 formation was suggested by
Trautz4:
                                       (5)
(6)
           NO3 + NO -> 2 N02
molecule  (reaction   5a),  which  has  two
possible fates, i.e., decomposition back to NO
and O2 (reaction 5b), or combination  with
another NO  molecule (reaction 6) to produce
two molecules of stable NO2- Because of the
very short half-life  of NO3 the overall re-
action  on a  time-sequence basis  may be
viewed as the collision of three molecules. It
is technically more correct, however, to view
NO2  formation  as  the  result  of two bi-
molecular reactions, i.e., reaction (5) followed
by  reaction (6).  The  overall effect  of high
temperatures is to decrease the equilibrium
concentration of NO3, and thus decrease the
rate  of reaction  (6) and  the  formation  of
N02.
  The rate constants applicable to  reactions
(5)  and (6) permit  significant  formation  of
NO 2 only when the NO concentration is in
the range of 100 ppm and  above, and when
sufficient  concentrations of QI are present.
Both NO  and NO2 are formed at combustion
temperatures  above  1093°C (2000°F), but
the amount of NO2 formed is usually well
under 0.5 percent of the total NOX. Exhaust
gases ejected into the ambient air are thought
to be diluted by a factor of 100 in a matter of
seconds. During this  short interval, a fraction
(approximately 10 percent) of the  NO does
react  to  form NO2.  Ejection  into the air
promotes conversion of NO to NO2 because
the diluent  air increases the availability  of
oxygen.5  Once the  NO  has been diluted to
1230 ME/m  (1 ppm) or below, its direct reac-
tion with  O2, as in reactions (5) and (6), does
not contribute significantly to  NO2  forma-
tion.

  It is possible to explain NO 2 formation by
an  alternative  mechanism.  NO2 formation
may also result from reaction of ©2 with the
dimer of the NO molecule.
                                       (7)
 Thus, NO in the presence of ©2 produces the
 highly reactive and unstable nitrogen trioxide
        At present, it is not possible to prove which
        of the  proposed  schemes  more correctly
2-2

-------
describes the mechanism of NC>2 formation in
the atmosphere.
   During  daylight  hours,  atmospheric NO
may be quantitatively converted to NC>2 by
reactions that  involve absorption of sunlight
energy  by  NC>2 and  subsequent  interaction
with  certain  hydrocarbons  and oxygen.  A
detailed description of  the nitrogen dioxide
photolytic  cycle and  photochemical  inter-
action  with  hydrocarbons is given  in two
companion  documents:  AP-63, Air  Quality
Criteria for Photochemical Oxidants,6 and
AP-64,  Air  Quality   Criteria for  Hydro-
carbons. 7
   Schematically, the  NC>2 photolytic  cycle
has the  form shown in Figure 2-1. The  wave-
lengths of light responsible for NC>2 photoly-
sis are  the  ultraviolet portions  of sunlight
   energy reaching the earth's surface, i.e., 3000
   to 4000 A. The NC>2 absorption spectrum is
   shown in Figure 2-2. Wavelengths longer than
   4000 A are efficiently absorbed  by NC>2, but
   such absorption  does  not  add  sufficient
   energy to break bonds. Ultraviolet absorption,
   on  the  other hand, adds enough  energy to
   break one of the N-O bonds in the NC>2 mole-
   cule. The released oxygen atom reacts rapidly
   with atmospheric  oxygen to  form  ozone,
   which  in turn,  reacts  rapidly with  the  NO
   formed in the breakup of the NO2 molecule.
   The result  in the atmosphere  is  the rapid
   cycling  of the NO2- With the possible excep-
   tion of minor secondary  reactions,  Figure 2-1
   describes the NO2 photolytic cycle.  If there
   were no competing reactions, the  photolytic
   cycle would have no overall net effect; that is,
                                    ULTRAVIOLET
                                      SUNLIGHT
                                        ENERGY
             PRODUCT 02
\
                                                                  AIR  02
                       Figure 2-1.  Atmospheric N02 photolytic cycle.
                                                                                    2-3

-------
200
z
0

t 150
ct:
0 _
to 6
CD <•>
< 1
U- "
0 o 100
H-^
z ju
UJ O

L) —

t 50
UJ
O
2  have been suggested as intermediates
during  the formation of NO2 when NO reacts
with O2.
  A summary of the physical properties of
the oxides  of nitrogen is  presented in Table
2-1.

D.   EFFECTS ON ATMOSPHERIC LIGHT
     TRANSMISSION
  Visibility reduction is one of the most com-
mon manifestations of urban air pollution. It
is caused by the scattering and absorption of
light by particles or gases in the atmosphere
and depends in a complicated way on the con-
centration  and  properties  of the gases  and
particles  present. This subject is discussed in
AP-49, Air Quality  Criteria for Paniculate
Matter?
  NO2 is intensely  colored and absorbs light
over the entire visible spectrum, but primarily
in the  shorter wavelengths, violet, blue,  and
green. In the atmosphere it reduces the bright-
ness and contrast  of  distant objects,  and
causes  the  horizon sky  and white objects to
2-4

-------
                    Table 2-1. PHYSICAL PROPERTIES OF NITROGEN OXIDES8
Oxide
NO
N02
N20
N2°3
N204
N205
Molecular
weight
30.01
46.01
44.01
76.01
92.02
108.01
Melting
point, °C
-163.6
-11.2
-90.8
-102
c
30
Boiling
point, °C
-151.8
21.2
-88.5
3.5b
c
47 b
Density,
g/liter
1.3204

1.977
-
c
-
Specific
gravity3

1 .4494^
—
1 .4772
c
1.64218
             Superscript:  temperature of liquid, °C. Subscript:  temperature of water, °C.
             "Decomposes.

             '-Value uncertain, dimer of NO2.
appear pale yellow to reddish-brown. A token
amount of light is attenuated by  the molec-
ular scattering effect of NO2-
   The  additional  presence  of  particulate
matter tends to mask the coloration effect of
NO2, but the two combined markedly reduce
the  visibility,  contrast,  and brightness of
distant   objects.  Particulate  matter  and
aerosols  are present  in  the  atmosphere as
primary  contaminants  from  urban sources
such  as industrial combustion and vehicular
transportation, and from natural sources such
as the sea, soil, and fog. They are also formed
through   photochemical  reactions  and  are
considered  to  be  the major cause  of the
reduced  visibility  associated  with  photo-
chemical smog.
   Figure  2-3 shows the calculated percent of
visible light  transmitted  as a  function of
wavelength  for different  atmospheric  NC>2
concentrations  and viewing distances  in an
aerosol-free  atmosphere.  Each curve is based
on  the  concentration (ppm)-distance (mile)
product; thus, 3 ppm-mile could  be 1 ppm at
3 miles, or 3 ppm  at  1 mile, or any combina-
tion  of  concentration and distance whose
product is 3. A more normal picture of  con-
ditions actually existing in the  atmosphere,
with both NC>2 and aerosols present, is shown
in Figure  2-4. These data are given in terms of
calculated  attenuation coefficients.10  Here,
the effect of visible light absorption  is ac-
companied by  transmitted light  attenuation
due  to the light  scattering  by the aerosol.
The  photochemical system involves NOX and
HC  in  the formation  of visibility-reducing
aerosols. Light  scattering  associated with the
presence of aerosols is  the primary cause of
visibility reduction  in  photochemical smog;
absorption of light  by  NC>2 makes a minor
contribution.
E.   SUMMARY
   By convention, the term  NOX represents
the sum of NO and NC>2, the only significant
nitrogen oxide air pollutants.  They are chiefly
emitted from combustion processes in which
the nitrogen and oxygen in  the  combustion
air are subjected to temperatures  in excess of
1093° C (2000° F). The major oxide in com-
bustion emissions is NO. The rapid cooling of
the gases in the combustion chamber prevents
dissociation back to nitrogen and oxygen. A
small fraction of NO is converted to NO2 by
reaction with oxygen  during the exhaust dilu-
tion  process;  however,  the  major pathway
leading to formation of NO2 from NO is the
photochemical interaction between NOX and
hydrocarbons.

   NO 2  causes  reduction  of visibility  and
coloration  of the  horizon   sky  in degrees
dependent  on its concentration, the viewing
                                                                                      2-5

-------
               g
               i/>
               CO
               in
               Z
               o:
                    3500
                                                                              7000
         4000     4500     5000     5500     6000     6500
        VIOLET  BLUE     GREEN  YELLOW ORANGE    RED
                        WAVELENGTH, A

Figure 2-3.  Transmittance of visible  light at different NO2 con-
            centrations and viewing distances.^
distance, and the accompanying aerosol con-
centration.  The presence  of  photochemical
aerosol or  other particulate matter suppresses
the coloration effect of NC>2 and increases the
visibility-reduction effect.

F.   REFERENCES

 1. Altshuller, A.  P. Thermodynamic Consideration
    in the Interactions of Nitrogen Oxides and Oxy-
                                  Acids in  the Atmosphere. J. Air Pollut. Contr.
                                  Assoc. 5:97-100, September 1956.
                                2. Goody, R. M. and C. D. Walshaw. The Origin of
                                  Atmospheric  Nitrous  Oxide.  Quart.  J.  Roy.
                                  Meteorol. Soc. 542:496-500, October 1953.
                                3. Leighton, P. A. Photochemistry of Air Pollution.
                                  N.Y.  and London,  Academic Press, 1961.  p.
                                  184-200.
                                4. Trautz, M. Reaktionskinetik and Additivitat der
                                  Inneren  Atomwarmen  bei Idealen  Gasen.  Z.
                                  Elektrochem. 22 (5/6): 104-107, March 1916.
 2-6

-------
            1 0

            08


            06

            05

            04
       i-
       £
            oa
       <
       3
            0.1

           008


           006
 _ t   yA.    j	 ~- »•••••••••• • •••••••••••^•"••||-l
 OjJppmNOj	|	     L^*^l(!Z$ttnEi
           .     _ -p.t***^	'^fff    .
         _*Q1 ^.
-------

-------
                                 CHAPTER 3.
                        SOURCES AND  CONTROL
             OF ATMOSPHERIC NITROGEN  OXIDES
A.   INTRODUCTION
  The major component of worldwide atmos-
pheric nitrogen oxides  (NOX) is biologically
produced nitric oxide (NO).1 Natural sources
produce about 50 X 107 tons of NO per year;
man-made sources emit  5 X 107 tons per year
of  NOX,  which  include  NO  and nitrogen
dioxide  (N02).
  In  general,  NOx concentrations in urban
atmospheres are  10 to 100 times higher than
those in  nonurban  atmospheres.  Problems
associated with NOX consequently  reflect the
importance  of  technological  sources  over
natural sources. Recent estimates indicate that
total  emission rates of  NOX  from  man-made
sources  in the United  States have increased
from  approximately 1.7 X 107 tons in  1966
to 2.1 X 107  tons in 1968,2 nearly one-half
of the world production.

B.   NATURAL SOURCES
  On a  global basis the major proportion of
NOX  is  NO  produced  by bacterial action.
Relatively little  is known about the global
circulation of atmospheric NOX, and only a
few  data  on  background concentration of
NOX  are available  from scattered  nonurban
areas.
  Lodge and Pate3 found average values of
NO2 ranging from 1.8 Mg/m3 (0.9 ppb) during
the  dry  season to 7.1 jug/™3 (3.6 ppb) during
rainy  periods in  Panama.  Individual samples
were  as  high as  12 jug/m3  (6  ppb). Junge4
reported NO2 concentrations averaging 1.8
Mg/m3 (0.9 ppb) in Florida and 2.5  Mg/m3
(1.3 ppb) in Hawaii.
  At  Pikes Peak in the Rocky Mountain area
of Colorado, Hamilton5 reported an average
NO2  concentration of  8.0 jug/m3  (4.1 ppb)
and an  average NO level of 3.3 jug/m3 (2.7
ppb).  Ripperton et al.6 found an average NO2
concentration of 9.0 /xg/rn3 (4.6 ppb) in the
Appalachian area of North  Carolina.  From
these  data,  Robinson  and  Robbins1  have
concluded that North American continental
average  levels of N©2 are about 8 Mg/m3 (4
ppb) and those of NO are about 2 jug/m3 (2
ppb).2  They charted a tentative  simplified
global  nitrogen  circulation  pattern. Based
upon  the estimated global background level
and the annual  emission rates, the average
residence time of NO2 is about 3 days and for
NO, about 4 days. Residence times reflect the
action  of  natural scavenging  processes
including photochemical reactions. Although
scientists tend to emphasize the role of photo-
chemical processes  in  urban  areas,  these
processes occur during daylight in  any area
where  the  ambient  air contains  NOX and
reactive  hydrocarbons.
C.  MAN-MADE SOURCES
  Fuel  combustion  is the major  cause  of
technology-associated  NOX  emissions.  In
1968  coal, oil, natural gas, and motor-vehicle
fuel combustion accounted for over 18 out of
an estimated  20.6 million  tons of man-made
NOX in  the  United States. Of the 10 million
tons  generated  by  stationary  combustion
sources, power plants emitted 4 million tons;
industries, 4.8 million tons; and home and
office  heating plants,  the  remaining  1.2
million tons. Natural-gas-burning  sources
made  the largest contribution of any fuel in
the  stationary  source group. An estimated 8
million tons was emitted from transportation
sources,  7 million tons of which  was from
motor vehicles.  Industrial  processes,  solid
waste  disposal,  and  other miscellaneous
                                         3-1

-------
                      Table 3-1. SUMMARY OF NATIONWIDE NITROGEN
                                 OXIDES EMISSIONS, 19682
Source category
Transportation
Motor vehicles
Gasoline
Diesel
Aircraft3
Railroads
Vessels
Nonhighway
Fuel combustion in
stationary sources
Coal
Fuel oil
Natural gasc
Wood
Industrial processes
Solid waste disposal
Miscellaneous
Forest fires
Structural fires
Coal refuse
Agricultural

Emissions
10" tons/year
8.1
7.2
6.6
0.6
Nb
0.4
0.2
0.3

10.0
4.0
1.0
4.8
0.2
0.2
0.6
1.7
1.2
N
0.2
0.3
20.6
percent
39.3
34.9
32.0
2.9
N
1.9
1.0
1.5

48.5
19.4
4.8
23.3
1.0
1.0
2.9
8.3
5.8
N
1.0
1.5
100.0
              aEmissions below 3000 feet.

              t>N=Not reported. Estimated less than 0.05 x 106 tons/year.

              Includes LPG and kerosene.
sources  accounted for about 2.5 million tons
of NOX. The  distribution of NOX emissions
by  major source  categories  is  indicated in
Table 3-1.
  Relatively  small quantities  of NOX  are
emitted  from  noncombustion industrial
processes, mainly the manufacturing and use
of nitric acid.7 Even .though total quantities
may be small, high concentrations of NOX can
be  emitted  from some of  these  chemical
processes. Electroplating, engraving, welding,
metal cleaning, and explosive detonation can
also  be responsible for  industrial NOX emis-
sions. The same is true with regard to the
manufacture  and  use  of  liquid-NO2-based
rocket propellants.
  NOX  emissions  from 22  cities  are sum-
marized by source category in Table 3-2.
  Over  60 percent of the total NOX emissions
occur in highly populated areas; 60 percent of
stationary source emissions and 45 percent of
motor-vehicle  emissions  occur  in urban
areas.2
3-2

-------
    Table 3-2.  NITROGEN OXIDE EMISSION
     SOURCES FOR 22 SELECTED CITIES8
Source category
Transportation
Motor vehicles
Other
Fuel combustion in
stationary sources
Power plants
Industrial
Domestic
Process losses
Refuse disposal
Emissions, percent of total
Average
42.6
36.3
6.3

50.7
23.0
23.8
3.9
5.2
1.5
Range
23-74


10-79

1-21
0.1-5.8
  Estimated  NOX  emissions increased  3.9
million tons from 1966 to 1968 (Table 3-3).
This increase was largely due  to:  (1)  2.9-
million-ton apparent increase from the burn-
ing  of natural  gas—apparent because these
sources  have  existed  but have  not  been
included in previous emission estimates (most
 of the increase attributed to the inclusion of
 fuel used to power the transmission of fluids
 through  pipelines); (2) 0.5-million-ton actual
 increase  in emissions  from  utility  power
 plants, mainly from increased use of coal; and
 (3) 0.6-million-ton increase in emissions from
 motor  vehicles  because  of  the  increased
 number and use of vehicles and as a result of
 CO and HC exhaust controls, which have led
 to increased NOX emissions.
D.   PRINCIPLES OF FORMATION AND
     CONTROL
1.   Formation
  Nitric  oxide  is  formed during fossil-fuel
combustion according to the reaction:
The equilibrium level depends upon variables
such as  flame  temperature,  pressure, con-
                 Table 3-3. SUMMARY OF ANNUAL NATIONWIDE EMISSIONS OF
                         NITROGEN OXIDES, 1966 THROUGH 19682
                                        (106 tons)

Source category
Transportation
Motor vehicles
Other
Fuel combustion
Coal
Fuel oil
Natural gas
Wood
LPG and kerosene
Industrial processes
Solid waste disposal
Miscellaneous
Man-made
Forest fires
Total

1966
7.6
6.6
1.0
6.7
4.0
0.9
1.6
0.2
_
0.2
0.5
1.7
0.5
1.2
16.7

1967 .
7.6
6.7
0.9
9.5
3.8
1.0
4.2
0.2
0.3
0.2
0.6
1.7
0.5
1.2
19.6

1968
8.1
7.2
0.9
10.0
4.0
1.0
4.5
0.2
0.3
0.2
0.6
1.7
0.5
1.2
20.6
Change from
1966 to 1968
+0.5
+0.6
-0.1
+3.3
Na
+0.1
+2.9b
N
+0.3b
N
+0.1
N
N
N
+3.9
              aN = Negligible.

              "Apparent change. (Emission estimates added in 1968 from sources not in-
               cluded in 1966).
                                                                                   3-3

-------
centration of each gas, retention time in zones
of  different  temperatures,   and  rates  of
cooling. The NO thus formed can react with
more oxygen to form NO 2- Equilibrium con-
centrations of  the NOX mixture constituents
are  a  function of the  variables encountered
during  combustion  and the  subsequent
extraction of heat from the gases. The  for-
mation of NO and achievement  of equilibrium
is  favored  by  high  temperatures  whereas
oxidation to NO2 is an exothermic, second-
order reaction favored at lower temperatures.
Consequently, high combustion temperatures,
rapid  cooling,  and instantaneous dilution of
exhaust gases promote the emission of high
concentrations of NO and low concentrations
of NO2-


2.   Control
  Techniques for controlling NOX from large
oil- and  gas-fired boilers  include two-stage
combustion, low-excess-air  firing, and furnace
modifications.  These approaches involve
reduction  of  peak  gas  temperatures  and
changes in the time-temperature history of
the combustion  gases. Feasibility of these
control  methods  has been  commercially
demonstrated;  research is now under way to
adapt  them or others to the reduction of NOX
emissions from coal-fired boilers.
  Control techniques directed at reducing the
temperature in the  combustion  chamber of
the  internal  combustion engine—such  as
exhaust-gas  recirculation,  higher-
than-stoichiometric air-fuel ratios, water injec-
tion,   spark  retardation  during  high-
speed-acceleration,   and low-compression
ratios  —  have  proved  to be effective  in  re-
ducing NOX emissions. It  should be noted,
however, that some of these  methods may
lead to increased HC and CO emissions while
reducing NOX-
  Catalytic methods  have been used  to
control NOX emissions  from chemical process
industries, notably nitric acid manufacturing.
One of the major problems is reduction of
catalyst life by contamination  with other
chemical  compounds.   Another control
technique employed in the chemical indus-
tries is caustic scrubbing.  Still  other tech-
niques are under active investigation. Control
of  emissions  from  stationary  sources  is
described  in  AP-67, Control Techniques for
Nitrogen  Oxide Emissions from Stationary
Sources.9
  Catalytic control approaches applicable to
motor-vehicle-exhaust  emissions  are  the
subject  of  extensive  research  by  catalyst
manufacturers  and the  motor  vehicle  in-
dustry.  Simultaneously controlling NO,  CO,
and HC is a  complex,  inter-related problem.
One  approach  to NOX  emission  control
involves catalytic  reaction of CO with NO to
form elemental nitrogen and carbon dioxide
(CO2). This approach is complicated by cur-
rent  motor-vehicle-exhaust  control  methods
aimed at the elimination of CO.
  More elaborate  discussion of  the  tech-
nology  applicable to  control of emissions
from  mobile  sources is presented  in AP-66,
Control Techniques for  Carbon  Monoxide,
Nitrogen Oxide, and Hydrocarbon Emissions
from Mobile Sources.] °
E.   SUMMARY
   The quantity of NOX generated by natural
sources  greatly  exceeds  the man-made
quantity on a worldwide basis. Man's contri-
bution is  a cause for concern, however, be-
cause the emissions are concentrated in urban
areas.  Fuel combustion, for transportation,
power, and heating, is  the  major source of
man-made  NOX  air  pollution.  Chemical
processing  is responsible for high, but local-
ized emissions.
   Control  of nitrogen  oxide  emissions has
been directed at both combustion sources and
chemical  processes.  For  stationary  com-
bustion sources,  the control  principle  has
been based  on  reducing either the  flame
temperature  or the availability of oxygen,
both of which prevent NO formation. Similar
principles of control are applicable to motor
vehicles. Catalytic principles, which have been
applied to reduce NOX from chemical proc-
3-4

-------
esses, may  also be applicable to the control of
NOX in motor-vehicle exhaust.
F.   REFERENCES
 1.  Robinson, E. and R. C. Robbins. Gaseous Atmos-
    pheric  Pollutants from  Urban and  Natural
    Sources. J. Air Pollut. Contr. Ass. 20: 303-306,
    May 1970.
 2.  Nationwide  Inventory of Air  Pollutant Emis-
    sions,  1968. National Air Pollution Control Ad-
    ministration.  Raleigh,  N.C.  Publication  No.
    AP-73, August 1970. p. 14-16.
 3.  Lodge, James P., Jr. and J. B. Pate. Atmospheric
    Gases  and Particulate in Panama. Science 153:
    408-410, July 22, 1966.
 4.  Junge,  C.   E.  Recent  Investigations  in  Air
    Chemistry. Tellus 8: 127-139, May 1956.
 5.  Hamilton, H. L. et al. An Atmospheric Physics
    and Chemistry  Study on Pikes  Peak in Support
    of  Pulmonary   Edema Research.  Research
    Triangle  Institute for Army Research Office. Re-
    search  Triangle  Park, N.C. Contract  No.
    DA-HC19-67-C-0029. 1968.
 6.  Ripperton,  L.  A.,  L.  Kornreich, and  J. J. B.
    Worth.  Nitrogen  Dioxide and  Nitric Oxide in
    Non-Urban  Air. J.  Air Pollut.  Contr. Ass.  20:
    589-592, September 1970.
 7.  Atmospheric Emissions  from  Nitric Acid
    Manufacturing  Processes.  Manufacturing
    Chemists' Association and  U.S.  Department of
    Health, Education,  and Welfare.  Cincinnati.
    Public Health Service Publication No. 999-AP-27.
    1966.
 8.  Mason,  D.  V., G.  Ozolins, and C.  B. Morita.
    Sources and Air Pollutant Emission Patterns in
    Major  Metropolitan   Areas.  (62nd  Annual
    Meeting, Air Pollution Control Association. June
    1969.) Paper 69-101.
 9.  Control Techniques  for  Nitrogen Oxides from
    Stationary  Sources.  National Air Pollution Con-
    trol Administration. Washington, D.C.  Publica-
    tion No. AP-67. March 1970.
10.  Control  Techniques  for  Carbon   Monoxide,
    Nitrogen  Oxide,  and  Hydrocarbon  Emissions
    from Mobile  Sources. National Air Pollution
    Control Administration. Washington, D.C. March
    1970.
                                                                                                3-5

-------

-------
                                   CHAPTER 4.
     CHEMICAL  INTERACTIONS OF NITROGEN OXIDES
                           IN THE ATMOSPHERE
A.   INTRODUCTION
  This  chapter  examines  the  chemical
relationship  of  nitrogen  oxides  (NOX) to
photochemical oxidants (OX) and to other
products of atmospheric reactions, a relation-
ship that encompasses several  interrelated
factors.l
  The most important  aspect  of this  rela-
tionship with respect to air pollution is associ-
ated with the  generation of oxidants.  The
generation of oxidants is a complex function
of  the  interaction  of certain hydrocarbons
and hydrocarbon oxidation products with the
nitrogen  dioxide photolytic cycle (See chap-
ter  2). Nitrogen dioxide (NO2), which initi-
ates the  process, is  an  efficient absorber of
both the ultraviolet and visible  portions of
solar radiation (chapter 2, Figure  2-2).  The
ultraviolet portion of the radiation has suffi-
cient  energy  to  disrupt bonds  in  NO2  and
form two new species, nitric oxide  (NO) and
atomic oxygen (O). The O atoms then react
with air oxygen (O2) to produce ozone (03),
the  principal oxidant in ambient air. The O
atoms and to a lesser extent the 03 react with
atmospheric  hydrocarbons,  especially  the
olefins and certain  aromatics to form other
pollutant products.
  The  NO formed in the initial NO2 photo-
lysis acts as a  natural ozone regulator  and
reforms oxygen  and nitrogen dioxide  in the
reaction:
         NO+ 03  -*  NO2 + O2

Since this reaction  occurs almost  instanta-
neously, appreciable concentrations of 03 do
not appear until NO has virtually disappeared
from  the ambient atmosphere.  Hydrocarbon
interaction with  the  NO2 photolytic cycle
leads to the disappearance of NO with subse-
quent formation of NO2, and to the accumu-
lation of hydrocarbon oxidation products and
residual ozone (chapter 2. Figure 2-1).
  The  rates of formation and the steady-state
concentrations of oxidants are a function of
light intensity, the concentrations of HC and
NOX,  the  HC-NOX ratio,  and  temperature.
The  nature of the HC  molecule  is an ad-
ditional factor since individual hydrocarbons
differ in their  rate of interaction with the
NO2 photolytic cycle. Some hydrocarbons,
such as methane and benzene, do not interact
to form photochemical oxidants, at least not
in synthetic systems.  Others, certain olefins
and substituted aromatics, for example, inter-
act  readily.2  The  onset  of the  reactions  is
regulated by NO2 concentrations.  Minimal
NO2 concentrations delay, whereas higher
concentrations expedite the cycle.
  This description of  atmospheric reactions
has been greatly simplified and mentions only
the major known reactions that account for
the formation  of photochemical  oxidants.
There are other reactions known to occur in
the photochemical system; however,  their
roles are considered less important. Certainly,
their involvement  with NOX is less clearly
understood, and discussion of them is omitted
in this  document.  They are covered, in part,
in AP-63,  Air Quality  Criteria for Photo-
chemical Oxidants.1
  Subsequent,  more detailed  research may
yet show that  some of  what  are now con-
sidered minor reactions have essential roles in
the regulation  of atmospheric phenomena.
These  so-called  minor  reactions may  also
become very  important as the  nature  of
                                          4-1

-------
ambient atmospheres change in response to
HC -  and NOX -  control programs.  This
chapter does  not  discuss products  of the
photochemical  reaction system,  other  than
oxidants, in detail.
   Earlier  publications1 >2   have  dealt  with
photochemical  oxidant relationships  to eye
irritants and  visibility-reducing aerosols, and
they suggest that the magnitude of effects
varies directly  with the oxidant formed, at
least in the controlled, experimental system.
Such an association, however, although typi-
cal of  present  photochemical  smog systems,
must  be  recognized  as a generalization to
which  there  are specific  exceptions.  Such
exceptions sustain  the need for continuous
atmospheric studies in order to reassess the
relationships between pollutant  and effects,
particularly as  they  may  be  affected  by
precursor-control strategies.

   The major known variables that contribute
to oxidant formation have been investigated
extensively under controlled laboratory  con-
ditions.3-7 On  the basis of such experimenta-
tion, it is possible to predict  the effect of
varied  inputs on oxidant concentration and
rates of formation. Extrapolation of labora-
tory results to ambient atmospheres, however,
involves  certain problems:  (1) The  mixing
volume in laboratory simulations is a constant
and favors a direct relationship between input
concentrations  of  the precursors and  the
maximum  concentration  of  resultant   oxi-
dant,  which  occurs several hours later. In
contrast,   the    concentrations   of   both
precursors  and   oxidants   in   ambient
atmospheres  are dominated  by  dispersion
factors, which  in turn alter the input/output
relationship.  (2)   The  definition of the
variables is much more precise in laboratory
simulations  than  in routine  atmospheric
monitoring. In  the laboratory, for example,
the identity and concentration of each hydro-
carbon in the system can be determined by
gas  chromatographic techniques.  Routine
atmospheric monitoring techniques have been
largely restricted to  measurements of total HC
concentrations and, only recently, to measure-
ments of total HC minus non-reactive methane.
In addition, laboratory simulations have been
largely carried  out at concentrations higher
than those observed in ambient atmospheres.
(3) All  laboratory  simulations are, to some
degree, affected by the experimental system.
Thus, interaction of the gases with contamina-
tion on the walls of the testing chamber may
affect the observed oxidant results.
   In defining the OX-NOX relationship, the
approach  was the same  as  the one used to
define the relationship of ambient HC con-
centrations to maximum daily oxidant values
in an earlier publication.8 The same patterns
of chemical reactions and time-concentration
relationships observed in the laboratory were
assumed to  apply  in  the atmospheric system
and to affect oxidant levels in the same way.
The approach, which  can be termed an obser-
vational hypothesis,  is based on  several  as-
sumptions that are covered, in limited detail,
in  the  companion   document, AP-64,  Air
Quality Criteria for Hydrocarbons?
   The basic  assumption  in the observational
approach is that early morning HC and NOX
levels are indicators of the oxidant levels that
will occur later in the day. Specifically, the 6-
to 9-a.m. HC and NOX levels are compared to
the daily 1 -hour maximum oxidant levels that
normally occur between  10 a.m. and 2 p.m.
Because the  time-concentration relationships
are influenced by atmospheric dispersion fac-
tors, a specific  ambient  combination of HC
and  NOX will not always result in a specific
maximum oxidant value. The hypothesis pro-
vides rather  for a wide  range of maximum
daily oxidant values to be associated with any
given early-morning  HC-NOX combination.
The precise oxidant concentration within that
range will be regulated by  the atmospheric
stability factors  for the day. For any concen-
tration of precursors  the hypothesis predicts
maximum oxidant  values near zero,  if sub-
stantial atmospheric instability follows the 6-
to 9-a.m.  period. Such instability  causes the
precursors as well as the products to disperse
before they can interact and yield any appre-
4-2

-------
ciable  amount  of oxidants. Conversely, the
hypothesis  allows for elevated oxidant values
when the dispersion  factors are more restric-
tive or when the atmosphere is stable follow-
ing the 6-  to  9-a.m. period.  In those cases,
precursors are trapped long enough for forma-
tion  and accumulation  of oxidants to occur.
Between maximum  and  minimum atmos-
pheric stability is a large range of dispersion
factor combinations,  which  lead to intermedi-
ate maximum  daily  oxidant concentrations
under the terms of the model.
   Information from laboratory studies, how-
ever,  suggests  that the maximum attainable
oxidant concentration will  be limited by the
available precursors regardless of how stable
the atmosphere becomes. It  is this upper-limit
oxidant concentration that  the observational
approach attempts to  define.

B.   OBSERVATIONAL MODEL FOR
     HC NOX - OX RELATIONSHIPS
   The  following  sections  examine  the
functional relationships observed between HC
and  OX, NOX and OX,  and the combined
relationship among HC, NOx, and OX. Since
OX concentrations are  also a  direct  function
of light intensity, the discussion concentrates
on those portions of the year in which sun-
light intensities are at or near maximum, thus
emphasizing the days with  maximum  poten-
tial for OX formation,  that is, the months of
May  through October.

1.   Early Morning HC  Relationship to
     Maximum Daily  Oxidant
   The generation of a model based on the
observational   approach is  limited  by  the
number of days for which pollutant measure-
ments  are available.  Oxidant  values for any
given combination of HC and NOX precursors
can range from zero to  the  upper limit. Since
the upper limit is attained  on only  about 1
percent of  the days  in a year, many days'
measurements are needed in order to provide
reasonable assurance that an upper-limit point
has actually been observed. The small number
of data points  in the vicinity of the upper-
limit line shown in Figure 4-1 for the relation-
ship between total  hydrocarbons  and the
maximum daily OX  observed  in several
United States cities, illustrates this point. (In
this figure and throughout the chapter, hydro-
carbons are measured  in parts  per million
carbon, ppm C).

  The problem with the  data in Figure 4-1 is
that total HC are  reported.  This includes a
large fraction of non-reactive methane, which
tends  to mask the effect  of  the smaller
fraction of HC known to be important to OX
formation. A recent publication 2 presents a
limited data  base  that  relates  the  more
important,  nonmethane  HC  values to
maximum  daily  oxidant  concentrations.
Figure 4-2 shows the 125 days of data. Again,
one  becomes acutely aware of the limited
number of points that define the upper-limit
oxidant  line.  Since they  are limited in
number,  these points presumably define the
minimum upper-limit-oxidant line, and more
extensive  observations  would  undoubtedly
include still higher oxidant concentrations.  It
seems reasonable to conclude, then, that the
upper-limit curve in Figure  4-2, is the most
valid relationship at this time.

  The curve can be  used to predict the maxi-
mum  1 -hour-average oxidant concentration
from  a measured 6- to 9-a.m. - average HC
concentration. Likewise, the minimal level of
HC that will produce a given OX level can also
be predicted. For example,  a reference  con-
centration  of 200  Mg/m3 (0.1  ppm) OX  is
associated with  200  Mg/m^ (0.3 ppm C) 6- to
9-a.m. - average nonmethane HC.
   Oxidant values below  140 Mg/m^  (0.07
ppm) are omitted  in  Figures 4-1  and 4-2
because several factors indicate these values
may be subject to measurement errors; how-
ever, they have no bearing on the upper-limit
values. The HC values in Figure 4-2 are not
reported below 200 /ug/m^ (0.3 ppm) even
though  they  may  have  a  bearing  on the
upper-limit oxidant  level,  because  they are
subject  to even greater measurement error.2
                                                                                    4-3

-------
          0.30
         0.25
          0.20
       E
       a.
       x
       o
          0.15
          0.10
          0.05
• DENVER
• CINCINNATI
A LOS ANGELES
O PHILADELPHIA
A WASHINGTON
              /•  AA mi
            1 • ABA A<

              • AOAAA
       O  B •OAOOB ABO A»OB  »O A
       OOA BB A • OB A    A     ••   A
o AOAO* BOOOAAC*OAB   AOAO  AB  A
 AA BAA ABAAOA*  B«*A«B     A*A
ABAAOA ABA ABA4MA0 A A*«  A  •  A—
      AABABAA ABA*B»  ADA  B  09AB
AOAAABABABA AA  *BA*«AB*B  «*A A
                                    2           3
                               TOTAL HYDROCARBONS, ppm C
          Figure 4-1. Maximum  daily  1-hour-average oxidant  concentrations as  a
                     function of  6-to 9-a.m.  averages  of  total hydrocarbon  con-
                     centrations  at CAMP stations, June  through September,  1966
                     through 1968 and in Los Angeles, May through October
2.    Early Morning NOX Relationship to
     Maximum Daily Oxidant

   An  observational approach similar to the
one described for the HC-OX relationship was
applied to NOX levels8  with one important
difference  in  regard  to definition  of the
specific precursors.  It  was not  possible to
                             define the reactive hydrocarbon mixture con-
                             tributing to OX concentrations2 precisely, be-
                             cause present  air-monitoring systems provide
                             only  gross measurements of ambient hydro-
                             carbons (i.e.,  concentrations of total HC or
                             total  HC less  methane).  In contrast, it is be-
                             lieved that only two species of NOX (NO and
                             NO 2)  are necessary to define the NOX-OX
4-4

-------
         0.30
         0.25
         0.20
       CL
       Q.
     APPROXIMATE UPPER-LIMIT
_   OBSERVED OXIDANT

                       ^j
  PHILADELPHIA

     PHILADELPHIA
    WASHINGTON
         0.15
       X
       o
         0.10
         0.05
                                             WASHINGTON A
                                 LOS ANGELES

                       LOS ANGELES     A
                              A      DENVER
                              * LOS ANGELES
                                  A  A PHILADELPHIA
                                  LOS ANGELES
                                  A
WASHINGTON  A A


WASHINGTON  */
     A > A A   A  A A
     A A A   A
A  At A      A A      A      AA
A  A   A Af     A
       >AA   A        A  A  A
       A A* A / A £ A A A   AA
                                                                  A
                                                                  A A
                        0.5         1.0          1.5
                                 NONMETHANE HC, ppm C
                                             2.0
                                         2.5
          Figure 4-2.  Maximum daily 1-hour-average oxidants as a function of 6-to 9-
                      a.m. averages of nonmethane hydrocarbons at CAMP stations,
                      June through September, 1966 through 1968, Los Angeles, May
                      through October 1967,  (Based on references 2 and 8).
relationship, and both of these are measured
routinely. The initial ratio of NO2 to NO will
influence the rate of formation of oxidant. In
terms of the oxidant maximum, however, the
primary consideration is the total NOx con-
centration, since NO is rapidly converted to
N02 during the photolytic reactions.
  Measurement problems  do  exist,  particu-
larly in regard  to  low NOX concentrations.
Unfortunately,  the NOX  measuring instru-
                                 ments, like the  HC and oxidant instruments,
                                 are primarily designed  to  measure values in
                                 the 0.1- to 1-ppm range, whereas much of the
                                 area of interest is below 0.1 ppm.
                                   The daily-maximum 1-hour-average oxidant
                                 levels and  the  6-  to 9-a.m. - average NOX
                                 concentrations9   were recorded  from  June
                                 through September,  1965  through 1968, at
                                 three  Continuous  Air  Monitoring Program
                                 (CAMP) stations in the downtown areas of
                                                                                    4-5

-------
Washington,  D.C., Philadelphia, and Denver.
These are shown in Figures 4-3, 4-4, and 4-5.
Careful evaluation of the data base has shown
that about 1  percent of the oxidant values fall
near the upper limit.8
  If we accept all data points as equally valid,
a reference value of 200 jug/m^ (0.10 ppm)
OX would be associated with as little as 20
jug/m3 (0.01  ppm)   NOX- Because of  the
analytical uncertainties in the low concentra-
tion region  a more  rational  approach is to
locate an NOx level  below which OX can be
expected to exceed the reference concentra-
tion, i.e., 200 Mg/m3 (0.10 ppm) oxidant, on
1 percent of the days. Figures 4-3, 4-4, and
4-5 show only seven occasions when the  oxi-
   dant equalled or exceeded 0.10 ppm and the
   NOX was  less than 0.04 ppm. These events
   occurred twice per smog  season  in Denver,
   once in four smog seasons in Philadelphia, and
   not at all during three smog seasons in Wash-
   ington, D. C.  This frequency represents about
   1 percent of the combined data base.
     Although the  curves that define the maxi-
   mum oxidant-forming potential in  Figures 4-1
   through 4-5 were drawn to include all the data
   points, they were not based  on a statistical
   approach. The data were too limited in number
   to justify a statistical analysis; however, strict
   adherence to the limits set by  the few applica-
   ble data points would associate an even lower
   value  of NOX with 200 Mg/m3 (0.10 ppm)
   oxidant.  Current analytical   uncertainties in
            0.25
            0.20
            0.15
         2
         Q

         o
            0.05
                    APPROXIMATE UPPER-LIMIT OBSERVED OXIDANT
0.05
          0.10
  0.15
NOX; ppm
                                                        0.20
0.25
                                                                            0.30
          Figure 4-3.  Maximum  daily  1-hour-average oxidant concentrations as a
                     function of 6-to 9-a.m. averages of total nitrogen  oxides in
                     Washington,  D.C.,  June  through September, 1966 through
                     1968.
4-6

-------
           0.35
           0.30
           0.25
        !  0.20
        CL
        1-
        z
           0.15
           0.10
           0.05
                APPROXIMATE UPPER-LIMIT
                OBSERVED OXIDANT
                       0.05
                                 0.10
  0.15
N0x,ppm
0.20
0.25
0.30
          Figure 4-4.  Maximum daily 1-hour-average oxjdant concentrations as  a
                     function of 6-to 9-a.m. average total nitrogen oxides in Phila-
                     delphia, June through September, 1965 through 1968.
the  low-range  measurements  make such  a
conclusion unwise. From the relationships in
Figures 4-3, 4-4, and 4-5, it appears that the
6- to 9-a.m.-average NOX levels must be kept
below 80 jug/m^ (Q.04 ppm) in order to pre-
vent the maximum daily 1-hour OX concen-
tration from reaching 200 /xg/ni^ (0.1  ppm)
or more.
  The reference concentration of 200
OX used  here  was selected on  the basis of
convenience  and  does  not  represent  the
lowest health-related value (130 ng/rn^ OX)
published  in  document AP-63,  Air Quality
     Criteria for Photochemical Oxidants.
       Data from three Los Angeles locations fre-
     quently affected by elevated oxidant levels
     are  shown  in  Figure 4-6 for a calculated2
     nonme thane HC level of 1 mg/m^ (1.5 ppm
     C).2  At two of these stations a 6- to 9-a.m.
     value of 80 jug/m^ (0.04 ppm) NOX and at a
     third, 90 ng/rn^ (0.05 ppm)  NOX, is associ-
     ated  with a  daily-maximum 1-hour-average
     oxidant concentration of 200 Mg/m3 (0.10
     ppm). These Los Angeles results are similar to
     those  obtained  in  Washington,  D.   C.,
     Philadelphia, and Denver.
                                                                                     4-7

-------
            0.30
            0.25
            0.20
         0.
         X
         o
            0.15
            0.10
            0.05
                 APPROXIMATE UPPER-LIMIT OBSERVED OXIDANT
                                                       •    •
                          0.05
0.10
0.15
0.20
0.25
                                           NO*
        Figure 4-5.  Maximum daily  1-hour-average  oxidant concentrations as  a
                    function of 6-to9-a.m. averages of total  nitrogen oxides in Den-
                    ver, June through September, 1965 through 1968.
   For these reasons the NOX-OX relationship
cited is considered the most reasonable  that
can be made at this time. It will be noted that
the  relationship  states only that the 6- to
9-a.m. - average NOX must be below 80 MS/m
(0.04 ppm) to prevent oxidant levels greater
than  200 jug/m^ (0.1 ppm) from occurring
later  in the day  more frequently than 1 per-
        cent of the time. It does not attempt to speci-
        fy the exact NOX concentration. Whether this
        or an even more stringent limitation of NOX is
        required, can be assessed  only after  further
        observations have  been  made  of ambient
        atmospheres and the manner in which ambi-
        ent HC-control affects ambient oxidant levels.
        Laboratory results  indicate that  HC  control,
4-8

-------
    0.30
    0.20
 Z
 <
 Q

 X
 o
    0.10
                                                                0.20
Figure 4-6.  Upper limit  of  maximum daily  1-hour  average oxidant con-
           centrations, calculated nonmethane hydrocarbon concentration
           of 1.5 ppm C, as a function of average total nitrogen  oxides
           from 6-to 9-a.m. at three Los Angeles stations, May through
           October 1967.
                                                                           4-9

-------
even in the  absence  of NOX  control, will
definitely  lead to reductions  in ambient oxi-
dant levels.
3.   HC-NOX • Oxidant Relationships
  The  previous  sections  dealt  with  the
relationships of  each  individual precursor to
maximum oxidant levels, without considering
combined  effects. An  initial  attempt  to
explore the  possible  combined  effects is
shown in Figure 4-7 for Pasadena, California.
In this figure,  upper-limit envelopes of HC
levels as a  function of oxidant and NOX are
superimposed on  a  graph  showing  the
relationships  of maximum oxidant levels to
NOX concentrations.
   The interdependent  relationship  among
HC,  NOX,  and oxidant concentrations in
Figure  4-7 is  in  complete agreement with
            0.40
            0.30
          Q.
            0,20
          X
          o
            0.10
                         I         I
                  APPROXIMATE UPPER-LIMIT
                 OBSERVED OXIDANTS
                                    HYDROCARBON ENVELOPES

                                       	—.-1.5 ppm C
                                       - — — 2.0 ppm C
                                       ——"• 2.5 ppm C

                                   I          I	I
                                  0.10
                                           NOX, ppm
     0.20
0.30
          Figure 4-7.  Hydrocarbon-oxidant envelopes superimposed  on maximum
                      daily 1-hour-average oxidant concentrations as a function of 6-
                      to  9-a.m. average  of total nitrogen oxides in Pasadena, Cali-
                      fornia, May through October 1967.
 4-10

-------
laboratory  simulations.3'6  Ambient  data
available are too  limited to  clearly define
these combinational effects. Furthermore, the
HC values in Figure 4-7 represent calculated,2
rather  than  measured,  nonmethane  hydro-
carbon  concentrations. As in  the  previously
discussed HC-oxidant studies, data were taken
at CAMP stations on the  days  when  non-
methane hydrocarbons were being measured.
A base from these,  involving 125  days of
validated  data,  was  used  to construct the
daily-maximum-oxidant isopleth  lines  as  a
function of the early-morning HC and NOX
values, which are  shown in Figure 4-8. The
dotted lines in Figure 4-8 are the levels of HC
and NOX associated with the reference 1-hour
maximum  oxidant   concentration  of  200
     ^ (0.1 ppm) in the previous HC-oxidant
        o
        CD
        DC
        <
        u
        o
             1.2
             1.0
            0.8
            0.6
            0.2
                                         PREDICTED LIMITS FOR 0.10 ppm OXIDANT
                                  0.10                 0.20
                                          NOX, ppm
                             0.30
          Figure 4-8. Approximate isopleths for selected upper-limit maximum daily
                     1-hour-average oxidant concentrations, as a function the 6-to 9-
                     a.m. averages of nonmethane hydrocarbons and total nitrogen
                     oxides in Philadelphia, Washington,  D.C., and Denver, June
                     through August, 1966 through 1968.
                                                                                    4-11

-------
and  NOx-oxidant  analyses (see  Figures 4-2
through 4-5).
  Since concentrations  of HC and  NOX are
apparently interdependent, it is not appropri-
ate to make a separate analysis of oxidant for
each, individually. This fact is illustrated in
the   contrast between Figures 4-3,  4-4, and
4-5,  which treat the NOx-oxidant relationship
as if it were independent of  HC level, and
Figure 4-7, which shows the interdependence.
Any  relationship of one precursor to oxidant
determines,  to  a  degree,  the relationship  of
the other precursor to oxidant.
  The data in  Figures 4-3, 4-4,  and 4-5 are
summarized in  Table  4-1.   Although  the
precursor effects  are  not  completely separa-
ble, it is clear that increases in both precursors
lead  to  increases in the maximum daily oxi-
dant level.
  Further observation of the data base shows
that  80 /xg/m3  (0.04 ppm) NOX occurs simul-
taneously with calculated nonmethane  HC
concentrations  ranging from 200   to 930
Mg/m3  (0.3  to   1.4  ppm   C).  Similarly,
observation  of  the 200 Mg/m^ (0.3  ppm  C)
nonmethane HC level shows NOX in  the range
of 80 to 320 Mg/m3 (0.04 to 0.16 ppm).
   It is also  clear  from the hydrocarbon
 envelopes shown  in  Figure 4-7  and the HC
 values in Table 4-1  that each HC level is as-
 sociated with a wide range of NOX values. For
 example, at a 1.0  mg/m3  (1.5 ppm) HC level
in Figure  4-7, the associated NOX concentra-
tions range from 80 to 340 Mg/m3 (0.04 to
0.18 ppm). Both  of the atmospheric  rela-
tionships  (Figures  4-7  and 4-8)  suggest  a
complex dependence of oxidant level on the
ratio  of  HC  to  NOX  in  addition  to  a
dependence on the absolute concentrations of
the two.  The two figures also suggest the
existence of an optimum HC-to-NOx ratio with
regard to maximum attainable oxidant.
  The results of a series of laboratory photo-
oxidation  experiments,5  designed  to deter-
mine precursor  effects  on   peak  oxidant
values, are shown in Figure 4-9.  The oxidant
values  have  been   converted  to ozone by
correcting  for  the  effect  of NC>2 on  the
oxidant  instrument.1 Comparison of observed
ambient oxidant isopleths in Figure 4-8 with
the experimentally  determined  isopleths in
Figure 4-9 indicates that the two observations
are, qualitatively, quite similar.
   This  similarity  between laboratory results
and  the  data from  ambient  atmospheres
(Figure  4-7)  lends  validity to the  measure-
ments obtained with current  air monitoring
instruments. Furthermore, it suggests that the
measured  total   hydrocarbon  and  the
calculated  nonmethane  hydrocarbons  are
indeed  proportional  to the reactive  HC
fraction taking part in atmospheric oxidant
formation.
          Table 4-1. CHANGES IN MAXIMUM DAILY 1-HOUR AVERAGE OXIDANT
              CONCENTRATION AS A FUNCTION OF 6- to 9 -a.m. AVERAGE HC
                          AND NOX CONCENTRATIONS9'3
No.
of
days
84
167
138
83
6- to 9- a.m. precursor concentrations
Total HC, ppm C NOX, ppm
Median
1.7
2.2
2.7
3.2
Range
1.5-1.9
2.0-2.4
2.5-2.9
3.0-3.4
Median
0.05
0.08
0.10
0.11
Range
0.01-0.15
0.02-0.16
0.02-0.17
0.05-0.24
% of days oxidant 1 -hour daily
maximum" equals or exceeds:
0.06 ppm
42
64
67
68
0.10 ppm
12
17
19
35
0.20 ppm
1
2
3
10
 Combined data from Washington B.C., June through September 1965, 1966, 1968; Philadelphia, June
  through September 1965 through 1968; and Denver, June through September 1965 through 1968.

  ^Oxidant data were corrected for effect of NO2-*
4-12

-------
           a
           a.
           UJ
          a.
          O
          a:
          a.
          Figure 4-9.  Oxidant isopleths  from  laboratory experiments showing effect
                     of varying  initial  precursor  hydrocarbon (propylene)and NO
                     concentrations on maximum ozone concentrations.5
   In  lieu of adequate atmospheric sampling
data,  laboratory simulations must be used as
broad guidelines in discussing the interactions
of the HC-NOX system that produce ambient
OX   concentrations.  Qualitatively,  the
precursor-oxidant  relationship  shown  in
Figure 4-9 is a typical laboratory finding with
certain noteworthy features. At  practically
every HC-NOX concentration, reductions in
hydrocarbons  without NOV reduction  result
                         A.
in a reduction in 63, whereas NOX reduction
without HC reduction does not always lead to
a reduction in O^.
   Sources that emit different proportions of
HC and NOX likewise  make different contri-
butions  to  ambient  levels  of oxidants.
Although the contributing factors are not yet
identified, it  appears that HC-NOX ratios do
vary in ambient atmospheres. Differing HC to
NOX  ratios  could arise  from the range of
operating  conditions in  automobile  engines
that  result  from  variations in  intake-air
temperature and humidity.

4.   Limitations and Interpretation of
     Observational Approach
   The  ideal  approach  to  defining  the
HC-NOX-OX interrelationships would take the
form of a comprehensive model, which would
take into account the interplay of emission
                                                                                   4-13

-------
rates, chemical reactions, and dispersion vari-
ables in space and time.
  The absence of a fully validated model of
this  type required the development of the
observational model discussed above.
  In  order  to prevent misinterpretation of
this observational approach, it is necessary to
consider  in detail the various factors involved.

a.   Assumptions
  One of the major assumptions is that the
parameters  affecting  ambient  oxidant con-
centrations  are the same as the ones deter-
mined in laboratory systems. This does not
mean that all  known  parameters have been
delineated.  In  terms of certain  major items,
however, the laboratory  results  and ambient
observations are in agreement; in the atmos-
phere as  in the laboratory, the major oxidant
formed is ozone.  Similarly, the conversion of
NO to NO2, the disappearance of certain re-
active hydrocarbons, the appearance of alde-
hydes and peroxyacyl nitrates as intermediate
products, and the dependence of the reactions
on sunlight  energy, are apparent in both. In
addition  to  the gross similarities, the rates of
reaction  and  relative product yields  are
consistent.
  Another  major assumption is that early-
morning precursor levels measured at  a
specific  location  are  proportional  to  the
maximum  oxidant  values  that  will  be
measured at this  same location  2 to 4 hours
later.  This  assumption  implies both that
meteorological  conditions  are homogeneous
and  that the precursor emissions are similar
over large metropolitan areas. Although this
assumption  may  be open  to  question, it  is
necessary at this time because the ambient-air
monitoring  data  presently available are gen-
erally restricted to a single station in any one
metropolitan area. It  has  been justified  by
data from Los Angeles  County, which has one
of  the  few  comprehensive  air-monitoring
systems that have been in operation for many
years.  It was  possible  to show  that oxidant
concentrations  measured simultaneously  at
several locations in a  900-square-mile area
were similar and related.1 °
4-14
   Although  evidence  supporting the  simi-
larity of area-wide emissions is less  substan-
tial,  the similarity of ambient concentrations
of HC, NOX, and oxidant for most days of the
year over the entire Los Angeles Basin sug-
gests that the activities leading to emissions
are reasonably consistent within a metropoli-
tan area. The similarity of the data also tends
to confirm the fact that meteorological condi-
tions are generally homogeneous throughout
the area.
  On weekends during the summer season,
however, a shift in observed emission patterns
and  resulting  maximum oxidant  concentra-
tions could  be  related  to  the recreational
activities of the populace.10  These alterations
could  only   be   demonstrated  for  stations
separated  by more  than 20 miles, however.
Otherwise, the assumption  of the similarity
between area-wide concentrations and subse-
quent  maximum  oxidant values appeared to
be valid within a considerable distance of any
one specific location.
  The  HC-NOX-OX relationship  in  Figures
4-7 and 4-8 was obtained by comparing 6- to
9-a.m. precursor values to subsequent oxidant
levels.  The fact that such developed  relation-
ships show similar concentration and concen-
tration-ratio effects  in  both field and labora-
tory  observations  (Figure  4-9) is  strong
support of both  of the major assumptions
made in the observational approach.  If the 6-
to 9-a.m. precursor values were not a measure
of subsequent maximum oxidant, it would be
impossible  to obtain atmospheric confirma-
tion of complex chemical interactions such as
shown in Figures 4-7 and 4-8.


b.   Weekday - Weekend Effect on Model
     Predictions
  Although, as noted, the data in Figures 4-3,
4-4,  and  4-5  are not  adequate  for  a  more
exact delineation of  the HC envelopes than is
found in Figure  4-7,  the  effect of  some
HC-NOX ratios on ambient oxidant levels can
be observed by combining the data from the
three cities and focusing attention on the HC
range containing  the most data. The oxidant

-------
levels for the 2.0- to  2.4-ppm C HC range is
shown in Figure 4-10 as a function  of NOX
concentration.  Although  the HC  range  is
narrow, the NOX  values have a rather wide
range. The  HC  shows a  10 percent variation,
but  the associated  NOX values  vary  60
percent.
  Part of the wide variation in NOX could be
attributed to a weekday-weekend difference.
During the  week, NOX values tended toward
the upper range, whereas during the weekend,
the  values  dominated the lower range, as is
shown on Figure 4-10. This variation in NOX,
with  HC  held  constant,  suggests  that the
reduction in weekend emissions is not of the
same magnitude for NOX as for HC.
  Further analysis of the data in Figures 4-3,
4-4, and 4-5 shows a 12 percent reduction in
median total hydrocarbons and a 43 percent
reduction in median NOX on weekends, when
the traffic patterns were similarly reduced.
   It is not  clearly understood  why the per-
centage NOX reduction was much greater than
the percentage HC reduction on weekends.
Part of the explanation is  undoubtedly as-
sociated with the non-specificity of the total
HC measurement, including, as it does, a large
volume of background methane, which does
not vary in proportion to motor-vehicle emis-
sions. The  calculated nonmethane HC drop,
however, more closely  approximates the NOX
drop.2
   The significant downward shift in weekend
emissions,  and  subsequent  lower oxidant
levels, in the  Los Angeles area suggest that
similar effects might occur in other metropoli-
tan areas.  To investigate  the potential  effect
of such shifts in pollution  patterns on the
            0.20
            0.15
          a.
          a
          X
          o
            0.10
            0.05
                                                                • WEEKDAYS
                                                                • WEEKENDS
  f  _   _      Wv
                                   0.05                  0.10
                                           NOX/ ppm
                             0.15
         Figure 4-10. Maximum daily 1-hour-average oxidants as a function of 6-to 9-
                     a.m. averages of total nitrogen oxides, for 2.0-to-2.4-ppm C
                     total  hydrocarbon  in  Washington,  D.C.,  Philadelphia,  and
                     Denver, June through September, 1965 through 1968.
                                                                                    4-15

-------
relationships charted in Figures 4-2 to 4-8, the
data points in Figures 4-3, 4-4, and 4-5 were
reexamined  on a  weekday  versus  weekend
basis. In all  three  cases,  the  early-morning-
weekend-NOx values  are  significantly lower
than weekday  values.  Approximately  75
percent  of the data below 110 jug/m^ (0.06
ppm) NOX are weekend values, and 75 per-
cent  of  the  data  above  110  jug/m^ (0.06
ppm) NOX  are weekday  values.  With  the
model, therefore, the 80 Mg/m-* (0.04 ppm)
early-morning-NOx  value associated with the
maximum daily oxidant of 200 /ug/m  (0.1
ppm) is  predominantly a weekend  observa-
tion. This raises the question of whether the
weekend effect is due to an actual reduction
in emissions  or to a  shift in  the  time of
emissions, so  that the 6- to 9-a.m. period does
not include peak NOX values on the weekend.
An hourly check of the 6-a.m. to  12-noon-
NOX values on weekends revealed no consist-
ent shift  in the time of peak NOX. It must be
recalled   that  after  9  a.m. the NOX can be
incorporated  into   photochemical  products,
and so the ambient NOX values measured dur-
ing sunlight hours  are  no longer an accurate
indicator  of  peak  emissions.  In this same
study, the time of day for peak OX was exam-
ined for  a weekend shift, but  no consistent
trend could be found.
   The weekend effect  is apparently due to
emission reduction. Figure  6-3 (chapter 6)
shows NO emission in Chicago during 1962 to
1964. Comparison of the weekend and week-
day values in this figure indicates a substantial
reduction of NO emissions on weekends.
   Examination  of  carbon  monoxide (CO)
values provided additional information on the
subject.  This relatively inert substance may be
used as   an indicator  of motor-vehicle emis-
sions. The diurnal variation  in  CO on week-
days, Saturdays, and Sundays, in Chicago for
the 1962 to  1964 period is shown  in Figure
4-11. The  most prominent  feature of  this
graph is the substantial decrease in CO values
on weekends, particularly during the morning
hours, which  suggests that the lower weekend
emissions, and  subsequent lower OX values,
are  predominantly the  result  of  weekend
4-16
reductions in vehicular traffic and other tech-
nology-related activities. Here again, the 6- to
9-a.m.  period still  appears  to be  the best
choice  for relating precursor emissions to sub-
sequent maximum oxidant values. An exten-
sion of this time period to include later hours
would  only associate lower average precursor
values  with maximum  daily oxidant concen-
trations.
   The correlation between lower emissions
and  lower oxidant  values provides  further
support for  the  observational  model. The
weekday-weekend difference is a direct meas-
ure of  the improvement  in air  quality to be
expected from a control method that reduces
early-morning  weekday  emissions  to  the
weekend level.
   Although the foregoing analysis  suggests
the validity of relating precursor and oxidant
values  at a single point, it does not preclude
the need  for an  adequate  air  monitoring
network. The single-station approach is  one
born of necessity and  should be replaced by
better monitoring networks to support a more
comprehensive emission-reaction model.

c.    Effects of Reductions in Photo-
     chemical Precursors

   As noted earlier, reduction in NOX without
simultaneous reduction in HC does not always
lead to reduction in oxidants. Such laboratory
findings  indicate that if the  objective  is
control of oxidant then control of HC is the
most appropriate step to take. Unfortunately,
oxidant control is not the only consideration.
NO, itself, must be controlled or even reduced
from   present  ambient  levels,  because  its
oxidation product,  NO2  is currently at or
near levels that  are  associated with adverse
health  effects in  over 50 percent of United
States  cities having a population greater than
50,000 (see   chapter  10).  The laboratory
findings also raise the possibility  that  NOX
control by itself may, under  certain circum-
stances, lead to increases in oxidant levels.
Such   increases,  however, would  only  be
characteristic  of  partial  NOX  control since
complete  NOX  control, like  complete  HC
control, would prevent formation of oxidant.

-------
          o.
          a.
          O
          U
             2400
0600
    1200
TIME OF DAY
1800
2400
         Figure 4-11. Diurnal variation of mean carbon monoxide levels on weekdays,
                     Saturdays, and Sundays in Chicago, 1962 through 1964.11
  An additional consideration is the effect of
HC or NOX reductions  on other products of
the  photochemical  complex,  such as  eye
irritants and visibility-limiting aerosols, which
are  also  related  to  oxidant concentrations.
Although  these  products  vary  with  oxidant
levels, the  HC-NOX  mixture having  the
greatest potential for oxidant production will
not  necessarily  correspond  to  maximum
potential for eye irritants.5'1 2 The effect can
be visualized in  greatly  simplified manner by
examining known and expected reactions in a
HC-NOX system  containing an excess of either
HC or NO. Excess HC may act as a scavenger
for oxidant and  reduce oxidant levels with an
accompanying increase in the eye irritant and
other HC oxidation products. Excess NO acts
                  as a scavenger  for oxidant and thus reduces
                  oxidant  levels,  increases  NO2  levels,  and
                  possibly  decreases eye-irritant products.  Un-
                  fortunately,  much  less  is  known  about
                  ambient  levels and variations in eye irritants
                  and areosols than about the ambient levels of
                  oxidants and NO2-  For  this reason, only
                  generalized relationships  of oxidant  levels to
                  eye irritation and visibility reduction can be
                  defined.  It  is  assumed  that  changes  in all
                  atmospheric pollutant concentrations  in the
                  years ahead  will   tend to  modify  these
                  observed  relationships.  Such  changes  will
                  most  likely  be small, and though demon-
                  strable in the laboratory, may not be observ-
                  able   in  the  atmosphere,  since  all of  the
                  products are interdependent. Large reductions
                                                                                    4-17

-------
of HC, NOX, or both, however, should reduce
all of the pollutants and the  resulting mani-
festations.
  Fortunately,  A P C O  has  a considerable
body  of  ambient and laboratory information
concerning the  atmospheric relationships of
oxidant  and  NC>2 that are most important
from  a  health-related  standpoint.  What  is
needed in the atmospheric data  is a deter-
mination of the most logical precursor-reduc-
tion method that will guarantee reductions in
oxidant  and  NC>2 to levels consistent with
known public health requirements. Only one
reduction program promises to accomplish
this goal, that is, a vigorous HC-reduction pro-
gram  coupled with NOX reductions designed
to reduce NC>2 levels below the health-related
criteria level. The necessity  for reducing NOX
below this suggested point cannot be ascer-
tained at this time. Nonetheless, the use of
less stringent NOX reductions (coupled with
vigorous  HC  reductions), the possibility of
causing a slight increase in oxidant values in
minimized and, at  the same time, the scaveng-
ing effect of NO for oxidant is maximized.

C.  ADDITIONAL REACTIONS AND
    PRODUCTS OF NOX
  In the previous sections, the major concern
was the NO2 photolytic cycle and its relation-
ship to the major atmospheric oxidant, ozone.
In addition to this major role the oxides also
engage in reactions leading to  formation of
several nitrogen-containing  compounds. Un-
fortunately, less is known about the fate of
NOX  than is known about the fate of the
hydrocarbons.  In laboratory simulations,
various  investigators  have  been able to  ac-
count  for  over  90 percent  of the carbon
atoms" but only  about  50 percent of the
nitrogen.
  One  of the more  important  classes of
nitrogen  compounds resulting  from atmos-
pheric HC-NOX interactions is known as the
peroxy  nitrates   (peroxyacyl  nitrates  and
peroxybenzoyl  nitrates).  Although  these
oxidants  are minor from  a concentration
standpoint, peroxy nitrates are powerful eye
irritants and phytotoxicants, and  hence im-
portant contributors  to photochemical air
pollution.  Peroxyacetyl nitrate (PAN), which
has the highest  atmospheric  concentrations,
has been  extensively  investigated. Pathways
for formation of this class of compounds have
been proposed, and several reaction systems
have been used to produce these compounds
in the laboratory; however, the specific mech-
anism  involved  in  the  atmosphere  has not
been established. Several other members of
this class  of peroxy nitrates  that  have  been
prepared and  studied  in  the laboratory in-
clude peroxypropionyl nitrate, peroxybutyryl
nitrate, peroxyisobutryl nitrate, and peroxy-
benzoyl nitrate,13 but  these have not  been
found  in  ambient atmospheres. Nor has the
formation of the initial member of the peroxy
nitrate  series,  peroxyformyl  nitrate,   been
accomplished in  the laboratory or observed in
ambient atmospheres.
   This class of peroxy nitrates causes eye ir-
ritation and vegetation damage.  Recent in-
vestigations14  suggest  peroxyacetyl nitrate,
and probably  the other  homologues,  are
important to the conversion of NO to NO2 in
ambient atmospheres.
   Additional experimental evidence suggests
that some undetermined fraction of the  NOX
ends up as nitrous and nitric acid, and trace
amounts of N2O. Furthermore, nitrates  have
been identified as an important fraction of
photo chemically formed aerosols.

D.  FUTURE RESEARCH
   This examination of the role  of NOX in
photochemical air pollution points out several
areas  requiring  further  investigation.   The
limitations  of the observational  hypothesis
clearly  indicate the  need for:  (1) a  more
comprehensive and definitive   simulation
model for NOX, HC, and oxidants, (2)  more
precise and definitive air-monitoring instru-
ments, and (3) an intensified effort to identi-
fy the mechanisms of reactions and the fate
of NOX.
   The observational model, for example, can
only  relate  average  concentration of  pre-
4-18

-------
cursors  during  an arbitrarily  chosen  time
period (6-  to 9-a.m.) to the upper-limit  of
maximum observed oxidant levels that occur
later in the  day. The model cannot,  in  its
present form, relate rates of precursor emis-
sions to rates of oxidant formation, nor does
it define the relative contribution  of various
sources, either near or distant,  to  any given
oxidant maximum.
  Coupled  with  the need for broader instru-
mentation is the need for better air monitor-
ing  networks. In many cities the description
of air quality and changes  in air quality are
still dependent upon  the data from a single
station. Such a single station cannot adequate-
ly  describe the variations in air quality that
affect metropolitan areas which occupy sev-
eral hundred square miles. The air-monitoring
stations and  networks that do  exist provide
only  a  limited  description of air quality.
There is a dearth of monitoring information
concerning  specialized community exposures.
For example, the air quality regime to which
the  populace is exposed while traveling on, or
living  adjacent to, heavily trafficked road-
ways, cannot be adequately  described.
  Besides additional measurement of specific
emissions and their  products,  the develop-
ment  of a  more refined, detailed  model  re-
quires more elaborate delineation of the mete-
orological parameters. At the present time,
certain key factors, such as  mixing height and
sunlight intensity, are  not measured on a rou-
tine basis in most polluted areas.
  Better information  is required  to  relate
emission rates to air quality, as  a function of
source and time.
  Although  knowledge of the  NC>2  photo-
lytic mechanism and the ways  in which this
cycle  interacts with reactive HC has a broad
scope, the  fate  of up to 50 percent  of the
atmospheric  NOX  is still not accounted for.
Details regarding these unidentified products
are  required in  order  to clearly assess their
impact on health and welfare.

E.   SUMMARY
  The relationships  between  early-morning
ambient HC levels and the maximum oxidant
level  observed  later in the  day have  been
examined through  an atmospheric  observa-
tional approach,1'2 which  does not attempt
to describe the complex relationship of oxi-
dant  to  ambient  NOX and  HC levels.  This
document is  concerned with a study of the
complete  HC-NOX-OX  precursor  -  product
relationship,  and primarily with the changes
in ambient oxidant and NOX levels resulting
from  absolute and relative concentration vari-
ations of the  HC-NOX complex. The results of
this examination are listed below.

   1.  The observational approach shows that
the same reaction parameters that govern the
HC-NOX-OX system in the laboratory operate
in ambient atmospheres.

   2.  The ambient levels of 6- to 9-a.m. pre-
cursors, HC and NOX, are a reliable indicator
of the  maximum attainable 1-hour-average-
oxidant  concentration  that  occurs  2  to 4
hours later, between 10 a.m. and 2 p.m.

   3.  The atmospheric conditions that lead to
maximum oxidant potential, i.e.,  low wind-
speeds, high  temperature, intense sunlight,
and  surface  inversions,  occur  on approxi-
mately 1 percent of the days.

   4.  Independent studies  of the effect  of
variations in  either HC  or  NOX level are not
possible  in  ambient  atmospheres.  Atmos-
pheric observation shows, however, that when
the   calculated  6-  to  9 - a.m. nonmethane
hydrocarbons are in the approximate range of
200 to 930 Mg/m3 (0.  0 to  1.4 ppm C) and
the 6- to  9-a.m. NOX value is approximately
80 Mg/m3 (0.04  ppm), the maximum  daily
oxidant will  reach 200  Mg/m3 (0.10 ppm) on
only 1 percent of the days.

   Further observation of the data base shows
that 80 Mg/m3 (0.04 ppm)  NOX occurs simul-
taneously with  calculated nonmethane HC
concentrations  ranging  from  200  to  930
Mg/m3 (0.3 to 1.4 ppm C).  Similarly, observa-
tion of the 200 jug/m3 (0.3 ppm C) nonmeth-
ane HC level shows NOX in the range of 80
to 320 Mg/m3 (0.04 to 0.16 ppm).
                                                                                    4-19

-------
   5.  Laboratory   studies  do  permit  the
independent evaluation of the effects of vary-
ing either  HC or NOX. Using such studies as
guidelines,  control  agencies can  consider  a
combinational HC  and  NOX   reduction
program   applicable  to  polluted  ambient
atmospheres.  The  data  base  suggests  that
reductions  in HC should  be the  primary step
for control of  oxidants. Coupled  with HC
control, NOX must be controlled at  a level
that will hold ambient NC>2 values below the
level of adverse health effects.

   6.  NOX  have also been implicated in the
formation  of peroxy nitrates  and aerosols.
The precise mechanism of formation has yet
to  be  elucidated,  but,  to  the extent that
oxidant levels are a measure of the intensity
of the effects of the photochemical complex,
the control of NOX and  HC to the suggested
levels  can  be  expected   to  control  all un-
desirable aspects of NOX and its products.

   These results are largely based on examina-
tion of current  ambient  atmospheric  data as
defined by current monitoring techniques. It
can be expected that these techniques will be
modified  to  provide  improved  information.
Furthermore,  the  application  of  emission-
control measures can  be  expected to change
the relative, as well as the absolute, concen-
trations of atmospheric contaminants. All of
these factors dictate the need for  a continual
reexamination and updating of the observa-
tions in this chapter.
H.   REFERENCES
  1. Air Quality Criteria for Photochemical Oxidants.
    National Air Pollution Control Administration.
    Washington, D. C. Publication No. AP-63. March
    1970.
  2. Air Quality Criteria for Hydrocarbons. National
    Air Pollution  Control Administration. Washing-
    ton, D. C. Publication No. AP-64. March 1970.
 3. Altshuller, A.P., et al. Chemical Aspects of the
   Photooxidation of the Propylene-Nitrogen Oxide
   System. Environ. Sci.  Technol.  1: 899-914,
   November 1967.
 4. Schuck, E. A. and G. J. Doyle. A Study of Ir-
   radiated Auto Exhaust. Stanford Research Insti-
   tute.  Report No.  9, Technical  Report  II.
   February 1958.
 5. Romanovsky, J.  C., R. M. Ingels,  and R. J.
   Gordon. Estimation  of Smog  Effects  in  the
   Hydrocarbon-Nitric Oxide System. J. Air Pollut.
   Contr. Ass. 17:454-459, July 1967.
 6. Korth,  M.  W., A. H. Rose,  and R. C. Stahman.
   Effects of HC/NOX on Irradiated Auto Exhausts.
   Part I. J. Air Pollut. Contr. Ass. 74:168-175, May
   1964.
 7. Dimitriades,  Basil. On  the  Function of Hydro-
   carbons and  Nitrogen Oxides in Photochemical
   Smog Formation. U. S. Bureau of Mines Report
   of Investigations, RI  7 and 33,  September 1970.
 8. Schuck, E.  A., et  al. Relationship of Hydro-
   carbons to Oxidants  in Ambient Atmospheres. J.
   Air Pollut. Contr. Ass. 20:297-302, May 1970.
 9. 1965 Through 1968  Data Tabulations and Sum-
   maries (Denver,   Philadelphia,  Washington).
   National Air Pollution Control Administration,
   Arlington, Virginia. Data on file in APCO Divi-
   sion  of Air Quality  Emissions Data. Durham,
   N.C.
10. Schuck, E. A.,  J. N. Pitts, and J. K. S. Wan.
   Relationships Between  Certain Meteorological
   Factors and  Photochemical Smog. Air Water
   Pollut. 70:689-711, 1966.
11. The  Automobile and Air Pollution, A Program
   for Progress, Part II. U. S.  Department of Com-
    merce. Washington, D. C. December 1967.
12. Hamming, W. J. and  J. E. Dickinson. Control of
   Photochemical Smog by  Alteration of Initial
   Reactant  Ratios. J. Air   Pollut.  Contr.  Ass.
   76:317-323,  June 1966.
13. Stephens, E. R. The Formation, Reactions, and
   Properties of Peroxyacyl   Nitrates  (PANS) in
   Photochemical Air  Pollution.  In: Advances in
   Environmental Sciences and Technology, Vol. I.,
   Pitts, J.N. and R.L. Metcalf (eds.).  New York,
   Wiley - Interscience, 1969, p. 119-146.

14. Schuck, E. A. and E, R.  Stephens, Oxides of
   Nitrogen,  In:  Advances in Environmental
   Sciences and Technology, Vol. I., Pitts, J. N. and
   R.   L.  Metcalf (eds.).  New York,  Wiley-
   Interscience, 1969. p. 73-118.
4-20

-------
                                   CHAPTER  5.

  METHODS  FOR MEASUREMENT OF NITROGEN OXIDES
A.   INTRODUCTION
  The  most accurate,  reliable,  and useful
methods for measurement of the atmospheric
nitric oxide (NO),  nitrogen dioxide (NO2),
and nitrogen oxides (NOX) employ standard
techniques to collect the sample in an absorb-
ing solution. The analytical  determination is
made colorimetrically or spectrophotometric-
ally. Analytical  methods  based  on physical
properties such as absorption or emission of
radiation, oxidation, or reduction and meth-
ods based on gas-chromatographic separation
and detection are not yet in common use but
will probably  receive more  attention in the
future.  Measurement techniques for  other
oxides of nitrogen will not be discussed in this
chapter.
B.   NITROGEN DIOXIDE
1.   Manual Methods
a.   Griess-Saltzman Method
  The colorimetric Griess-Saltzman  method1
is deemed  the most suitable manual method
generally applicable to  the  measurement of
NC>2  in the atmosphere.*  This method is
based on the reaction of NO2 with  sulfanilic
acid to  form a diazonium salt, which couples
with N-(l-naphthyl)-ethylenediamine  dihy-
dro chloride to form a deeply  colored  azo
dye. Air is sampled through  a fritted bubbler
into a solution of the Griess-Saltzman reagent
for periods of 30  minutes or less, and the
color  is allowed to develop for an additional
15 minutes after completion of sampling. The
color  produced,  which is proportional to the
"This method has been adopted as a tentative method by the
 Intersociety Committee on Manual of Methods for Ambient
 Air Sampling and Analysis.2
amount of NO2 sampled, is measured at 550
nanometers (nm). The method can be used to
determine  concentrations  of  NO2 in  the air
from  40 to 1500 Mg/™3 (0-02 to 0.75 ppm).
  Ordinarily, interferences are not a problem,
although high ratios of sulfur dioxide to nitro-
gen dioxide (about 30:1) can cause bleaching
and give misleadingly low  values.1  Interfer-
ence from  other oxides of nitrogen  or ozone
are  negligible at concentrations found in pol-
luted  air. Peroxyacetylnitrate  (PAN) can give
a response  of up  to 35 percent of an equiva-
lent  molar concentration  of NO2, but  in
ordinary ambient air PAN concentrations are
too low to  cause significant error.1 The color
formed  is  reasonably stable,  but  samples
should be read within 1 hour of completion
of sampling.
  In  the original  method  Saltzman^ used
solutions of  sodium nitrite as  calibration
standards. He reported that, under laboratory
conditions,  0.72 mole of nitrite produces  the
same color  as 1 mole of NO2 gas and incorpo-
rated  this  factor  into  his  calculations.  In
recent years  this  stoichiometric factor  has
been the source of considerable  controversy.
Values ranging from 0.5 to  1.00 have been
reported,4'10  and one recent work11 recon-
firmed  Saltzman's  original  findings.  The
method  can  be  standardized by  using an
accurately known concentration  of  NO2  gas
and thus eliminate  the stoichiometric factor
from the calculations.
b.  Jacobs-Hochheiser Method
  The  Griess-Saltzman  method  for NO2
measurement   cannot be  used  successfully
when  the  delay  between sample  collection
and color measurement is more  than  4 to 6
hours or when  sampling  periods  of longer
                                           5-1

-------
than  1 hour are required.  In such situations
the Jacobs-Hochheiser method12 is preferred.
With  this method, NC>2 can be  measured in a
network where a central laboratory is used for
analyzing field samples. With a later modifica-
tion12'14  of the procedure, NO2 sampling
periods can be as long  as  24 hours, and the
delay of analysis can be at least 2 weeks, if
necessary.

  In  the  Jacobs-Hochheiser procedure, pol-
luted  air is passed through aqueous sodium
hydroxide,  in  which nitrite ion is  formed
from  the NO2- The resulting solution, which
contains sodium nitrate in addition to sodium
nitrite, is treated with hydrogen perioxide to
remove  possible sulfur  dioxide  interference,
and acidified. The nitrous acid produced is
measured  by the Griess-Saltzman diazotizing-
coupling procedure, except that sulfanilamide
is used in place of sulfanilic acid. In the devel-
opment of the Jacobs-Hochheiser procedure,
the ratio  of nitrite  to  nitrate14 was  deter-
mined and an average stoichiometric factor of
0.63 was selected for converting nitrite ion to
NC>2 concentrations.

  Because the  sampler  adopted  by  the
National Air Surveillance Network (NASN) of
AP CO employs a 24-hour sampling period,
it  precludes the use  of the Griess-Saltzman
method and all data obtained by this network
since  its inception have been derived by the
Jacobs-Hochheiser method. Because of dis-
crepancies  concerning the NO2-nitrite stoichi-
ometric factor, all NASN  computations are
now based on the assumption that one mole-
cule of NO2 produces one molecule of nitrite
and no nitrate.

   The absorption efficiency for NO2 varies
with  the  sampling  system used.  Detailed
studies  have shown  that,  under  laboratory
conditions, the NASN sampling system has an
analytical efficiency of approximately 35 per-
cent, a correction that is  applied in NASN
calculations  of NO2  concentrations.  Other
investigators13-15  have found  different effi-
ciencies.
2.   Continuous Method
  Most continuous  NO2 analyzers  use the
Griess-Saltzman method or a slight modifica-
tion of it. The sample air is drawn through an
absorber at an  accurately determined flow
rate countercurrent  to  a  controlled flow of
absorbing reagent. The absorber must be care-
fully  designed  and  properly sized  because
NO2  is somewhat difficult to absorb. Suffi-
cient time  is allowed for full color develop-
ment, and then the colored solution is passed
to a colorimeter or spectrophotometer where
the intensity is measured continuously at 550
nm.
3.   Miscellaneous Methods
  Correlation spectrometry,16'17 gas chroma-
tography,18 long-path infrared spectroscopy,
and  electrochemical  oxidation and reduc-
tion19 have been applied  to NO and  NO2
analysis, but have not found wide application
in atmospheric sampling  and  measurement.
Correlation  spectrometry,  a  promising new
technique for remote sensing, requires further
field trials before it can be  adopted for  air
pollution  measurement.20  Gas chromatogra-
phy is potentially a useful and versatile meth-
od for measuring  NO2-2 Even though it is in-
termittent  by  nature  and,  in the  strictest
sense, does not  lend itself  to  continuous
monitoring, gas chromatography  can none-
theless  analyze sequential  samples  rapidly
enough to approximate a continuous measure-
ment. Long-path  infrared spectroscopy uses
large-volume cells, which make rapid response
difficult to achieve in a continuous-monitor-
ing  instrument. Long-path infrared also has
the  disadvantage  that water  vapor interferes
with the  measurements.   Electrochemical
transducers, which operate on the principle of
oxidation or reduction of pollutant absorbed
on a sensing electrode, show excellent poten-
tial  as simple, compact means of analyzing for
NO2 and  NOX  in sources such  as automobile
exhaust.  At present,  however, these  sensors
lack" adequate  sensitivity  for ambient  air
analysis.
5-2

-------
4.   Calibration
  Any measurement method should be cali-
brated to assure accuracy. The preferred pro-
cedure for calibrating a method for measuring
NC>2 is to use the entire system, including the
sampling line and its components, to analyze
an atmosphere  containing accurately  known
amounts of  NC>2  in  the  concentration range
to be studied. These calibration  gases can be
generated by simple-gas-dilution with check-
analysis,21  by  precision-gas-dilution without
check-analysis,9  by  a  gravimetric  pro-
cedure,22 by an electrolysis system,23 or by
use of a permeation tube.24  The permeation-
tube procedure is by  far the most accurate
and simplest of these methods. In another less
complex  procedure  for calibration25  ac-
curately  known  concentrations of sodium
nitrite solutions  are  used to  calibrate  the
analytical portion of the method. Since the
sampling efficiency of the method being cali-
brated is  not taken into account, the pro-
cedure is considered inferior to  the permea-
tion-tube method.
  Permeation tubes  deserve special mention
because  of  their  accuracy,  simplicity, and
growing use. A permeation  tube is a device
used  to  generate artifical atmospheres  with
known  NC>2 concentrations.11  Constructed
of Teflon, it contains  a  liquified gas such as
NC>2, which permeates through the walls at a
slow, but constant rate, as long as the temper-
ature of the device  is  held  constant. The
permeation rate of the gas can be accurately
determined.26  One  suitable  system using a
permeation  tube  is  shown  schematically in
Figure 5-1. The temperature of the bath con-
taining  the  permeation device must be held
constant, preferably within 0.1° C, and the air
or nitrogen passed across the permeation tube
must be dry. This technique has been used for
both field and  laboratory calibrations.  NC>2
permeation  tubes are moisture-sensitive,  and
care must  be  taken  to protect them from
atmospheric moisture during use and storage.
  After  any method  has  been  calibrated
under  closely  controlled laboratory condi-
tions, preferably  using the  permeation-tube
technique, the problem of interferences in the
ambient  air at the sampling site  should  be
considered. One method for dealing with the
interference problem is to use  local ambient
air for which NO2 background  is determined
concurrently, as dilution air. Such a procedure
will be satisfactory if all of the precautions in
the use of permeation tubes, particularly that
of allowing only  dry air or nitrogen to pass
over the tube, are observed.
C.   NITRIC OXIDE
1.   Oxidation to N02
  Because  no  appropriate analytical  tech-
niques for  direct measurement of NO exist,
NO is generally determined by oxidizing it to
N02, measuring the NO2 and then converting
the values to NO concentrations. The oxida-
tion  step is as yet poorly understood. Some
workers find it unreliable and report conver-
sion  efficiencies ranging from 40 to 110 per-
cent.25'27   The oxidizers used are aqueous28
or  solid29 potassium  permanganate, and
dichromate or chromium trioxide in various
formulations27' 30'31  all of which have un-
certain lifetimes at peak efficiency.  In addi-
tion,  the dichromate and chromium troxide
preparations are extremely sensitive to high
humidity. Aqueous permanganate seems  to
have the fewest limitations; hence, it  is the
best current choice.
  It is generally believed that NO  concentra-
tions  are underestimated  as a  result of the
poor  conversion  efficiency, a  factor that
should be kept  in  mind when studying NO
data.  Steps  can  be  taken to minimize the
problem, but oxidation of NO  to  NO2 must
always be  considered  a potential  source  of
error.
a.   Series Mode
  The measurement of NO by  the oxidation
method can be  done  in either  a series or a
parallel mode. In  the  series  mode (Figure
5-2A), which some workers32'33 prefer, the
sample air passes through an NO2 analyzer for
removal  and  measurement  of NO2,  then
through an oxidizer where NO is converted to
NO2,  and  finally  through a  second  NO2
                                                                                      5-3

-------
                                                     FLO* METER
                                                         OR
                                                      CRITICAL
                                                  E3  ORIFICE
                                                          THERMOSTATISTICALLY
                                                              CONTROLLED
                                                          EQUILIBRATION
                                                               COIL
                                           \
                           PERMEATION TUBE   BUBBLER
                          Figure 5-1.  Permeation tube system.
analyzer. The response from the second NC>2
analyzer is a measure of the NO in the sample
air.
b.   Parallel Mode
  The parallel  mode  (Figure 5-2B),  requires
two  separate analyses. In one, sample air is
analyzed for NC>2; and in the other, sample
air is analyzed for total NOX. The NO concen-
tration is the  difference between the  total
NOX and the value for NO2- In order to carry
out the analysis for total NOX, the sample air
is passed  through an  oxidizing scrubber to
convert NO into NO2-  In  addition to the

5-4
usual  problem of uncertain efficiency in NO
to NO2 conversion, the problem of possible
capture of NO2 by the scrubber adds to any
basic unreliability in the procedure.
2.  Chemiluminescent Methods
  Recently, a  device that  operates on the
principle  of  chemiluminescence  has  been
developed.34 A detector measures the lumi-
nescence produced during the gas-phase reac-
tion between   NO  and  ozone. During the
operation,  ambient  air  flows  continuously
through  a calibrated orifice into  a vessel,
where it is mixed with ozone and gives rise to

-------
          A.  SERIES MODE
SAMPLE
AIR
N02
ANALYSIS
AND
REMOVAL


                                                                      N02
                                                                   ANALYSIS
                      NO,CONCENTRATION
                 NO CONCENTRATION
         B.  PARALLEL MODE
                      N02 CONCENTRATION
         SAMPLE
          AIR
         SAMPLE
          AIR
                       NO CON-
                      CENTRATION
                                          NOX CONCENTRATION
                 Figure 5-2.  Oxidation schemes for analyzing NO and NOX.
measurable luminescense. This  approach rep-
resents  a  promising development  in  direct
NO-measurement methododology.
3.   Calibration
  Artifical atmospheres  containing  known
concentrations of NO are not easily prepared.
The extreme ease with which NO at high con-
centrations is oxidized to NO2 makes a dilu-
tion  system  impractical  and  the  very low
boiling point (-151.8° C) of NO makes prep-
aration and calibration of permeation devices
extremely difficult.  Since most NO analyzers
are, in reality, NO2  analyzers preceded by an
oxidizer, NO2 can  be  used for  calibration.
Removal of the oxidizer before calibration is
recommended to avoid  the problem of NO2
capture.
D.   TOTAL NITROGEN OXIDES
1.   Oxidation of NO to NO2
  At some point in the analysis all of the
methods  in general use for  measuring  total
NOX involve  the  oxidation  of NO to NO2,
with the  attendant problems  of this  step.
Either  the series  or the parallel method in
Figure  5-2 can be adapted to give the  NOX
concentration, the former by summing NO
and NO2 concentrations, the latter by on-line
reading.

2.   Chemiluminescent Method
  Total NOX have been measured by a chemi-
luminescent reaction involving oxygen atoms
and  NOX3 5 in a gas-phase reaction similar to
the one used  for NO  measurement.  In this
                                                                                    5-5

-------
system NOX is mixed with oxygen atoms in a
chamber where the oxygen atoms  rapidly and
quantitatively  convert NC>2 to NO in the gas
phase, and excess oxygen  atoms react with
NO to produce chemiluminescence, which is
then measured by appropriate means. Though
the method holds much promise, it has not
yet been widely tested in the field.

E.   SUMMARY
  Nitrogen dioxide is  the only atmospheric
oxide of nitrogen that can be measured direct-
ly with currently used  techniques. All of the
procedures for measuring  atmospheric NO
and NOX rely on some type of converter to
oxidize NO to NO2, which is then measured
and  arithmetically  converted  to  respective
values.
  Any  method used for measuring NO2 in
the ambient air should be calibrated before
use  with sample  atmospheres  containing
known  amounts of  NO2- Permeation tubes
are recommended for  generating such test
atmospheres.
  For sampling periods of 30 minutes or less,
the most suitable method for measuring NO2
is  the manual colorimetric Griess-Saltzman
method1 in which the color depends on con-
version of NO2 to nitrite ion. The degree of
conversion of NO2 gas into nitrite  ion remains
a point of controversy  iri the Griess-Salt/man
and other methods. Nonetheless, this method
is relatively free of interferences and may be
employed to  determine  concentrations  of
NO2 ranging from 40 to 1500 jug/m3 (0.02 to
0.75 ppm).
  For periods from  30 minutes to 3 hours,
sampling  can be done either by taking  the
appropriate  number of discrete 30-minute
samples  with the  manual  Griess-Saltzman
method  or by using the continuous  Griess-
Saltzman method for the entire period.
  For sampling periods of 3 to  24 hours or
for situations where analysis must be delayed
more than 4  hours  after sampling,  the most
suitable currently available method for measur-
ing NO2 is the Jacobs-Hochheiser method.12
Because  the  collection efficiency  of the
Jacobs-Hochheiser method is low and vari-
able, the method must be carefully calibrated
before use.
  The  Griess-Saltzman and  the  Jacobs-
Hochheiser methods are  not interchangeable
and can yield different results; therefore, the
appropriate  one  must  be carefully  chosen
according to the  purpose of the sampling to
be done.
  For continuous  NO2  measurements,  the
most  suitable  method currently available is
based on the  Griess-Saltzman  procedure.
When used in  conjunction with  an oxidizing
scrubber for conversion of NO to NO2, the
continuous Griess-Saltzman method can also
be used, in either the series or parallel mode,
for measuring NO in ambient air. Researchers
disagree as to which of the two modes is more
satisfactory, but the major problem with NO
measurement lies in completely oxidizing NO
to NO2-
  Total nitrogen oxides can be determined by
summing  NO  and NO2 concentrations from
separate samples, or by oxidizing NO to NO2
and then measuring NO2-
  New  approaches  for  measuring  NO  and
NOX based on chemiluminescence are promis-
ing,  but  have  yet  to be adequately field-
tested.
   Research is  still  needed  to  develop  and
thoroughly  evaluate  sensitive, reliable,  and
practical methods for determining NO, NO2,
and NOX.

F.   REFERENCES
 1. Standardization Advisory Committee, National
   Air Pollution Control Administration. Tentative
   Method for Determination of Nitrogen  Dioxide
   in the  Atmosphere  (Griess-Saltzman Reaction)
   (in press).
 2. Intersociety  Committee  on Manual of Methods
   for Ambient Air Sampling and Analysis, Subcom-
   mittee 3. Tentative Method of Analysis for Nitro-
   gen Dioxide  Content of the Atmosphere (Griess-
   Saltzman Reaction). Health Lab. Sci. 6: 106-113,
   April 1969.
 3. Saltzman, B. E. Colorimetric Microditermination
   of Nitrogen  Dioxide in  the Atmosphere. Anal.
   Chem. 26: 1949-1955, December 1954.
5-6

-------
 4. Stratmann, H. and M. Buck. Messung von Stick-
    stoffdioxid in der Atmosphere. Air Water Pollut.
    10: 313-326, May 1966.
 5. Buck, M. and H. Stratmann. The Joint and Sep-
    arate Determination of Nitrogen Monoxide and
    Nitrogen Dioxide in the Atmosphere. Staub 27:
    11-15, June 1967.
 6. Huygen, I.  C. Reaction of Nitrogen Dioxide with
    Griess Type Reagents. Anal. Chem. 42:407- 409,
    March 1970.
 7. Gill, W. E.  Determination of NO and NO2 in Air.
    Am. Ind. Hygiene Ass. J. 21(1): 87-96, February
    1960.
 8. Kooiker, R. H., L. M. Schuman, and Y. K. Chan.
    Nitrogen Dioxide Poisoning. Arch. Environment
    Health 7(1): 13-32, July 1963.
 9. Saltzman, B. E. and A. F. Wartburg, Jr. Precision
    Flow Dilution System for Standard Low Concen-
    trations of Nitrogen  Dioxide,  Anal. Chem. 37:
    1261-1264, September 1965.
 10. Shaw,  J.  T.  The  Measurement of  Nitrogen
    Dioxide in Air. Atmos. Environment 1(2): 81-85,
    March 1967.
 11. Scaringelli, F.  P., E.  Rosenberg,  and  K.  A.
    Rehme. Comparison  of Permeation Devices and
    Nitrite Ion as Standards in  the Colorimetric De-
    termination  of  Nitrogen  Dioxide.  Environ.
    Science  Technol. 1970 (in press).
 12. Standardization  Advisory Committee, National
    Air Pollution Control Administration. Tentative
    Method for Determination  of Nitrogen Dioxide
    in the  Atmosphere (24-Hour Sampling Method)
    (in press).
 13. Jacobs,  M. B., and S.  Hochheiser. Continuous
    Sampling  and  Ultramicrodetermination  of
    Nitrogen  Dioxide in  Air.  Anal.  Chem. 30:
    426-428, 1958.
 14. Morgan, G. B., C. Golden, and E. C. Tabor. New
    and Improved  Procedures for Gas Sampling and
    Analysis in the National Air Sampling Network.
    J.  Air  Pollut. Contr. Ass., 17:  300-304,  May
    1967.
15. Meadows, F. L. and W. W.  Stalker. The Evalua-
    tion of  Efficiency  and  Variability of Sampling
    the Atmosphere for Nitrogen Dioxide. Am. Ind.
    Hygiene  Ass.  J. 27:  559-566,  November-
    December 1966.
16. Barringer,  A.  R. and B. C. Newburg. Remote
    Sensing  Correlation Spectrometry for Pollution
    Measurement.  (9th Conference on Methods in
    Air Pollution, Pasadena,  California,  February,
    1968).
17. Kay, R.  B.  Absorption Spectra  Apparatus Using
    Optical Correlation for the  Detection of Trace
    Amounts of SO2. J.  Appl. Optics. 6: 776-778,
    1967.
 18. Morrison, M. E. and W. H. Corcoran. Optimum
    Conditions and Variability  in Use  of Pulsed
    Voltage in Gas-Chromatographic Determination
    of  Part-Per-Million  Quantities of  Nitrogen
    Dioxide. Anal. Chem. 39: 255-258,1967.
 19. Shaw,  M.  Electrochemical Tranducers. (Paper
    presented at the Symposium on Advances in In-
    strumentation for Air Pollution Control, Cincin-
    nati, Ohio, 1969.)
 20. Remote Monitoring Instrumentation for Source
    Emissions and  Urban  Atmospheres.  Clean Air
    Quarterly  (State of California), 11: 9-12,
    December 1967.
 21. Chrisman, K. F. and K. E. Foster. Calibration of
    Automatic Analyzers in a Continuous Air Moni-
    toring Program. (56th Annual Meeting of the Air
    Pollution Control Association, Detroit, 1963.)
 22. Thomas, M. D.  and R. E. Amtower. Gas Dilution
    Apparatus for Preparing Reproducible Dynamic
    Gas Mixtures in Any Desired Concentration and
    Complexity.  J.  Air  PoHut.  Contr.  Ass,,  16:
    618-623, November 1966.
 23. Hersch, P. and  R.  Deuringer. Trace Addition of
    Nitric  Oxide and  Nitrogen Dioxide to Air by
    Electrolysis. (56th Annual Meeting, Air Pollution
    Control  Association,  Detroit, 1963.)  Paper
    63-36.
 24. O'Keeffe,  A.  E. and  G.  C. Ortman.  Primary
    Standards for Trace Gas Analysis. Anal. Chem.
    38: 760-763,  1966.
 25. Interbranch  Chemical  Advisory  Committee.
    Selected Methods  for the Measurement of Air
    Pollutants. Public   Health  Service,  Cincinnati.
    Publication No. 999-AP-ll, 1965.
 26. Scaringelli, F. P., et al. Preparation of Known
    Concentrations of Gases and Vapors with Perme-
    ation Devices Calibrated Gravimetrically. Anal.
    Chem. 42: 871-876, July 1970.
27. Repley, D. L., J. M. Clingenpeel, and R. W. Hum.
    Continuous Determination of Nitrogen Oxides in
    Air and  Exhaust Gases. Air  Water Pollut. 8:
    455-463, 1964.
28. Thomas, M. D., et  al. Automatic Apparatus for
    Determination of  Nitric Oxide and  Nitrogen
    Dioxide in the Atmosphere. Anal. Chem. 28:
    1810-1816, December  1956.
29. Calhoun, J. D. and  C.  R. Brooks. A Solid Oxi-
    dant for Oxides of  Nitrogen Analyzer. (7th Con-
    ference  on  Methods in Air  Pollution, Los
    Angeles, California, January 1965.)
30. Wilson,  D., and S.  L.  Kopczynski. Laboratory
    Experiences  in  Analysis of Nitric  Oxide  with
    "Dichromate" Paper.  J. Air Pollut. Contr. Ass.
    18: 160-161, March 1968.
                                                                                                 5-7

-------
31. Jones,  E.  E., L. B. Pierce, and  P.  K. Mueller.
    Evaluation of a  Solid Oxidant System. (7th Con-
    ference on Methods in Air Pollution Studies, Los
    Angeles, January 25, 1965.)
32. Mueller, P. K. and  Y.  Tokiwa. Series vs. Parallel
    Continuous Analysis for NO, NC>2 and NOX. (8th
    Conference on  Methods  in  Air  Pollution  and
    Industrial  Hygiene  Studies, Oakland, California.
    February 1967.)
33. Mueller, P. K.,  et al. Series vs. Parallel  Continu
    ous Analysis  for NO,  NO2 and NOX H. Labor-
    atory Data. (9th Conference  on Methods in Air
    Pollution and  Industrial Hygiene Studies, Pasa-
    dena, California, 1968.)
34. Fontijn, A., A. J. Sabadell, and R. J.  Ronco.
    Homogeneous  Chemilumenescent Measurement
    of Nitric Oxide  with Ozone. Anal.  Chem., 42:
    575-579, May 1970.
35. Snyder, A.  D. and  G.  W. Wooten. Feasibility
    Study for the Development of a  Multifunctional
    Emission Detector for NO, CO,  and SO2- Mon-
    santo Research Corporation. Dayton, Ohio. Final
    Report  under National Air  Pollution Control
    Administration Contract No. CPA-22-69-8. 1969.
5-8

-------
                                  CHAPTER 6.
        ATMOSPHERIC  LEVELS  OF NITROGEN OXIDES
A.   INTRODUCTION
    The ambient concentration of oxides of
nitrogen (NOX) varies greatly with time and
place. Values reported by the  National Air
Pollution Control Administration Continuous
Air Monitoring Program (CAMP), the State of
California's Statewide Cooperative Air Moni-
toring Network  (SCAN),  and other special
studies since 1961,  appear  to have several
characteristic  concentration  patterns.  The
data, which include both day-mean values for
different time periods and frequency distribu-
tions of individual concentrations, have been
evaluated  in  terms  of  both  temporal and
meteorological variables, since these factors,
as well as the measurement variables must be
considered in  a complete   description  of
atmospheric levels of NOX.
B.   TEMPORAL VARIATIONS IN NOX
     CONCENTRATIONS
1.   Diurnal Patterns
   On a normal  day in a  city, ambient NOX
levels follow a regular pattern  with the sun
and traffic. Before daylight, nitric oxide (NO)
and nitrogen dioxide (NO2) remain relatively
stable  at  concentrations  somewhat  higher
than the  daily minimum. As human activity,
especially  automotive traffic, increases in the
hours just after dawn (6 to 8 a.m.), the con-
centration of the primary  contaminant, NO,
increases. Then, as ultraviolet energy from the
sun  becomes  available, NC>2  concentration
generally increases until nearly all of the pri-
mary NO is converted to secondary NO2- As
the NO concentration approaches very low
levels (<0.1 ppm),  photochemical  oxidants
begin to accumulate  and reach a peak about
midday. Concurrent  with  NO2 and oxidant
formation,  both NO and NO2 become incor-
porated into products such as the peroxyacyl
nitrates and other nitrated compounds.
  As solar intensity  decreases, atmospheric
stability increases,  and  winds  decrease. In
urban  areas the increased automobile traffic
in late  afternoon and evening (5 to  8 p.m.)
also  increases concentrations of primary  NO,
but  solar energy is generally unavailable to
convert NO to NO2 or to cause  formation of
oxidants. The absence of sunlight does not
completely  halt NO2 formation,  however.
The  principal oxidant present,  ozone (03),
will  continue to react  rapidly  with NO to
form NO2, until the 03 supply  is exhausted,
and thus the evening NOX concentrations may
continue to  rise.  This  condition  may  be
ascribed to meteorological factors  and,  on
cold evenings,  to  increased emissions from
stationary sources.
  Table 3-1 (chapter 3) shows that transpor-
tation  sources contribute 39 percent of the
national NOX emissions.  This table does not
account for the spatial distribution of NOX
sources. In  the  center of major  cities, where
most of the air  monitoring  stations  are
located, the contribution of mobile sources to
NOX levels  may be  considerably higher than
the   reported  39 percent.  For  this  reason,
traffic  volume,  which is highly  variable  and
easily  documented,  becomes an important
tool  in identifying and relating NOX pollution
to a recognized emission - automotive exhaust.
  The  diurnal variations of NO,  NO2, carbon
monoxide (CO), and 03 in Los Angeles, on
July 19, 1965, are shown in Figure 6-1. This
concentration profile graphically depicts the
NO peak and  the associated lag  in NO2 peak
as the  NO is  converted. It also illustrates the
                                          6-1

-------
              0.50
              0.40 —
            CL
            CO
           O
            CN
           O
           z
           d
           z
              0.30 —
0.20 —
              0.10 —
                2400
         0300    0600
0900    1200    1500
   TIME OF DAY
1800    2100    2400
            Figure 6-1.  Average daily 1-hour concentrations of selected pollutants in
                       Los Angeles, California, July 19, 1965.8
subsequent increase in 03 concentration. The
data in  Figure  6-1 do not show  the general-
ized increase in evening NOX levels described
in the previous paragraph. On this particular
day,  evening  dispersion  factors  apparently
prevented  any  substantial buildup  of the
NOX.
   Other  measurements  display  a  similar
characteristic profile. Figure 6-2  shows diur-
nal variations in NO2 concentration for St.
Louis, Philadelphia, and Bayonne, N.J., taken
from data gathered  by personnel of  the Air
Pollution Control Office of EPA.
   The hours at  which the peak concentra-
tions of NOX  usually occur either coincide
with, or take place shortly after, the hours of
                                 peak  automotive traffic. The diurnal pattern,
                                 therefore,  shows  little day-to-day  variation
                                 except  for weekends  and  holidays,  when
                                 traffic differs  from weekday  patterns. The
                                 diurnal variations of NO on weekdays,  Satur-
                                 days, and Sundays are shown for the Chicago
                                 CAMP station in Figure 6-3.l The Sunday 8-
                                 a.m.  peak concentrations of NO are  about
                                 one-third of  the  weekday  peak concentra-
                                 tions. This  is but one example of  weekend
                                 changes. On some weekends, some  locations
                                 have  peak NO concentrations equal  to  week-
                                 day peak values, but occurring 1 to 3  hours
                                 later. Furthermore,  weekend NO concentra-
                                 tions in  certain  recreational  areas  often
                                 exceed weekday values.
6-2

-------
               0.10
               0.08
               0.06
             0.
             0.
             CN
            O

            Z'0.04
               0.02
               0.00
                          V/
                            1       I       I       I       I
                                   A  BAYONNE, NOV. 1968


                            '         \  PHILADELPHIA,   f
                                        \     AUG. 196C    /
                                ST. LOUIS, JUNE 19tt
                         I
                                                                I
                 2400    0300    0600   0900   1200   1500
                                         TIME OF DAY
                                                       1800   2100   2400
           Figure 6-2. Diurnal variation  in  monthly  mean  1-hour-average NC>2
                      concentrations from three urban stations.8
   0.20
   0.15
 Q.
 Q.

 O
 •z.
0.10
   0.05
                          SUNDAY
     2400     0600     1200     1800
             LOCAL STANDARD TIME
                                   2400
F igure 6-3. Weekday  and  weekend  1 -hour
           average NO  levels  in  Chicago,
           1962 through 1964.1

2.   Seasonal Patterns
  \tyhereas NO  displays  a marked seasonal
variation,  NC>2  does  not.  For NO, as with
most  primary  contaminants,  higher  mean
values are observed during late fall and winter
months, when there is less overall atmospheric
mixing and generally less  ultraviolet energy
available for forming secondary products.
  Figure  6-4 shows the  seasonal patterns in
NO, by presenting the mean values by month
of the year, during the time period indicated,
for the  cities of Chicago,  Denver, Bayonne,
and Los Angeles. The distinctly higher levels
during winter are evident.
   The pattern  for NO2 is less distinct, and
shows much less variation from month to
month. Even though a greater amount of NO
is  converted  to  NO2 during  the  summer
months, the actual concentration of NO2 may
be  greatest during  other seasons of  the year
when the NO2  disappearance rate is slower.
The seasonal variation of monthly mean NO2
values for four  cities is shown in Figure 6-5.
   The changes  in NOX  and N©2  average-
1-hour maximum  daily  concentrations  as a
function  of the  month of the year are shown
in Figure  6-6 for the Los  Angeles Basin during
1962. Total NOX varies  by a factor of three
and is highest during the winter and lowest in
                                                                                      6-3

-------
              0.24
             0.20
              0.16
           o
           z
              0.08
              0.04
              0.0
                      I     I     I     T
:	\   CHICAGO, 1966
                     FEB.
    APR.
JUN       AUG
 MONTH
                                  OCT
                                                                        DEC
            Figure 6-4. Monthly mean NO concentrations at four urban sites.8
summer. The high  wintertime NOX  values
correspond in part to increased NO emissions
from power and heating sources. The ratio of
NO2  to NOX is at a maximum during the
summer because of the high conversion rate
of  NO  to  NO2-  Variations  in  dispersion
factors further influence NO2 and NOX con-
centrations.
   Study of California data has indicated that
the same seasonal patterns  are observed when
the  concentrations  are expressed  as:  (1)
monthly peak, the highest peak concentration
observed   during  a month;  (2)   monthly
average,  the average of all  hourly averages
during a month; and (3) the monthly average
of maximum hourly averages for each day of
the month.2
3.   Annual Trends
   The ability to determine possible year-to-
year trends in air quality  depends  upon the
                    extent to which cyclical,  seasonal, and ran-
                    dom  variations  are  considered. A study by
                    Brier3 indicates that no adequate long-term
                    data base is presently available for computing
                    cyclical  variations. Seasonal factors, on  the
                    other hand, are distinct and can be recorded.
                    When data for several years are available, an
                    average  "seasonal index"  can be  calculated
                    and applied to  monthly  means  to  obtain
                    monthly values that are corrected for seasonal
                    variations.
                      The remaining and most troublesome diffi-
                    culty  in  determining  trends is  random varia-
                    tion.  Evidence of this is  shown  in several
                    figures reproduced from a report by Ingels, et
                    al.4 on atmospheric trends of NO2- Figure 6-7a
                    shows a least-squares linear-trend line fitted to
                    raw monthly data from downtown Los An-
                    geles, and Figure  6-7b shows  a  subjectively
                    fitted trend curve applied to  monthly data
6-4

-------
             0.10
             0.08
           E
           £0.06
           CN
           o
           z

             0.04
             0.02
             0.0
I     I     I     I

    LOS ANGELES, 1967
                               CHICAGO, 19tt v  ,.	«%

                                          *              ***
                              DENVER, 19K
                    FEB
         APR
JUN
 MONTH
                                                  AUG
                                       OCT
                              DEC
           Figure 6-5. Monthly mean N02 concentrations at four urban sites.8
                    J      A
                     MONTH
Figure 6-6.  Average and standard deviation of
           daily  maximum oxides of nitro-
           gen and nitrogen dioxide concen-
           trations for seven locations in Los
           Angeles Basin, 1962.9
                                              corrected for seasonal variation. Even though
                                              NOX emissions in  Los Angeles County in-
                                              creased slightly during the periods considered,
                                              no easy interpretation  of these data is possi-
                                              ble. At  this time, there is no clear way to
                                              determine the percent of this increase that
                                              was due to increased NOX emissions and the
                                              percent  due to  long-term  changes in  the
                                              meteorological  factors  that  affect ambient
                                              concentrations.
                                                Nitrogen dioxide concentrations in  several
                                              New Jersey  cities are  shown  for the  period
                                              1966 through  1968 in Figure 6-8.5  No signifi-
                                              cant trends are apparent.
                         4.   Model for Relating Temporal
                              Maxima
                           It is possible to convert the maximum con-
                         centration for one averaging time to that for
                         another time. Larsen6 has worked out formu-
                         las in which he computes the expected maxi-
                         mum concentration (C) for a selected averaging
                                                                                     6-5

-------
                 1.00
                 0.80 -
                 0.60 -
              x
             O
             z   0.40 -
                 0.20 A-'
                 0.0
                     III II I III Illlll Ml I II I II III II I Mil III MM I I I III  Mill I II III I Ml
            Figure 6-7a. Monthly means of daily maximum NOX concentrations at
                        Los Angeles Civic Center, 1957 through 1961.4
             o
             z
                 0.80
                 0.60
                 0.40
                 0.20
                 0.0
                      II II I II I Illlll I |l I III I I III! lill
                                  STATION 58 "	
              Timlin n iiiiiniinTniinTi]
              	 STATION 1
                                  NEW INSTRUMENT
                     i n 1111111111111 ii 11111 inn i mi 11 n inn miii 11 in 111 mi 11 ii
                        •1957-
-1958-
                                              -1959-
                                              YEAR
-1960—*4—1961-
            Figure 6-7b. Trend curves fitted to monthly means of daily maximum
                        NOX concentrations, corrected for seasonal variation, Los
                        Angeles Civic Center,  stations 1 and 58, 1957 through
                        1961.4
6-6

-------
                0.100

                0.090

                0.080

                0.070

              E 0.060
              Q.
              t 0.0501

                0.040J

                0.030

                0.020

                0.010
CN
o
        TTT1 I  I I  I I  I T M I ITT I Mill M I I I I  | |  M I  I I  I I
           NEWARK
       - BAYONNE
                                                      CAMDEN
                      111 I  I I  1 1 II I  M  I I I  I 1  M I  I I  I I I  I I  I I  I III I  I I
                     1965|«     :  1966     »|«      1967     .U      1968 ——
                                              YEAR
             Figure 6-8. Average monthly N02 concentrations in four New Jersey
                       communities, t965 through 1968.5
time by using the geometric mean concentra-
tion (Mg) for the area in  question with its
geometric standard deviation  (<7g), and the
standard normal deviate (Z) for no more than
one occurrence in 1.67 N trials. His formula
is:

              C = Mg(ag)z

This formula and the associated mathematical
model are based on the assumption that aero-
metric  data  are lognormally distributed. The
model  generally gives good estimates of
expected maximum concentrations for averag-
ing times of 1 hour to 1 day, even if aero-
metric   data  depart  substantially from log-
normal. The values of observed-hourly-average
concentrations  at the  0.1 and 30  percentile
levels are used to calculate maximum concen-
trations for other averaging times. In addition,
the model provides predictions of the 1 -year
concentration.  The  fit for averaging  times
from  1  day to about  6 months depends on
how well actual aerometric  data  at a site
fit the(lognormal assumption. When the aero-
metric   data fit  lognormality well, then the
model1 fits well for all averaging times from  1
                                 hour to 1 year. Experience has shown that the
                                 model tends to overestimate expected concen-
                                 trations for  averaging times shorter than  1
                                 hour.

                                 C.   EFFECT OF METEOROLOGICAL
                                      FACTORS
                                      At any given location, the concentration
                                 of NOX is dependent on many factors. Among
                                 them are the emission rate; the diffusion rate,
                                 which is meteorologically determined; and the
                                 reaction rate, which is both meteorologically
                                 and chemically  determined.  All of  these
                                 factors have an important effect on the distri-
                                 bution of pollution  over a city.  As with any
                                 primary pollutant, concentrations of NO (and
                                 therefore, subsequent concentrations of NO2)
                                 on  the  upwind side  of an area of sources are
                                 different  from  concentrations  downwind.
                                 Spatial  variations in concentration make the
                                 location of air-monitoring stations important,
                                 if representative  air quality data are  to  be
                                 collected.
                                   As discussed earlier, the meteorology of an
                                 area also influences diurnal and seasonal varia-
                                 tions; for example,  daily minimum  concen-
                                 trations of NOX usually occur during daylight
                                                                                     6-7

-------
hours, at times of maximum  ventilation and
mixing.  The amount  of ultraviolet energy
from the sun also governs the conversion rate
of NO to NO2-
   From  a  photochemical standpoint, the
highest concentrations  of NO2 should  occur
during late  spring and summer months. In
most regions of the United States, radiation
inversions during this period  are often quite
pronounced in the morning hours, but usually
break by midday. On the other hand, during
the fall  and winter, when low sunlight  inten-
sity  limits  maximum  photochemical  proc-
esses, the atmosphere may remain  stable for
longer periods when the general circulation is
weak. Such counteractive effects account for
the relatively constant  level of NO2 through-
out the year. An additional factor in moderat-
ing NO2 levels is the incorporation of NO2
into photochemical  products. This has the
effect of  lowering daily  average NO2 values
during periods of intense photochemical acti-
vity.
   The effect of  a stagnation period on the
peak concentrations of NO and NO2 is shown
in Figures 6-9 and 6-10. The NO rise occurs
when minimal dispersion of pollutants, due to
stagnating meteorological conditions, coincide
with maximum traffic  conditions.  The sharp
midday  drop in NO  coincides with the NO2
peak; the sharp  NO2  drop coincides  with a
total oxidant peak (Figure 6-11).  The rapid
afternoon rise in both NO and NO2 is, again,
due  to  stagnating conditions coupled with
increases in vehicular traffic.

D.   OBSERVED URBAN NOX
     CONCENTRATIONS
   The  majority of available  continuous air
quality data for oxides of nitrogen is less than
15  years old.  In  1956, the  Los Angeles
County   Air  Pollution  Control  District
(LACAPCD)  began monitoring NO and NO2
continuously. In 1961, the State of California
Department  of  Public  Health  organized  a
Statewide Cooperative Air Monitoring Net-
work  (SCAN)  of 13 stations.  Six  of these
were equipped  and operated  by the Depart-
   °-45M  I I  I I I  LI  I I I  I I  I I I  I I  I I I  I L
E
Q.
a
o"
Z
   0.40
   0.35
   0.30 -i
   0.25 -
    0.20 -
    0.15 -
    0-10
    0.05
    0.00
      2400
      1200
TIME OF  DAY (EST)
2400
 Figure 6-9.  Diurnal variation of NO levels
             during 1963 stagnation in Wash-
             ington, D.C.. October 15  to  19,
             1963.1
 ment;  the  remaining  seven  stations were
 selected from  the  existing  Los  Angeles
 County Network.  As  of 1968 there were 20
 SCAN stations in operation in areas outside
 the Los Angeles County Air Pollution Control
 District.
    The Public  Health Service opened the first
 station of its Continuous Air Monitoring Pro-
 gram (CAMP) in Cincinnati in October 1961.
 Subsequently, additional stations equipped to
 monitor  NO  and NO2  have  been added
 throughout the country.
    Tables 6-1, 6-2, and 6-3 show NO, NO2,
 and total NOY data from the CAMP network.
6-8

-------
   o.20|-r
 E
 a
 a
 x
O
Z
   0.15
0.10
   0.05
                   STAGNATION
        "i  I I I  I I  I I I  I I  I I I  I I  I I I  I I  I I"
2400     0600     1200
           TIME OF DAY
                            1800
                                       2400
Figure 6-10.  Diurnal variation of N02 levels
             during stagnation in Washington,
             D.C.,  October 15  through 19,
             1963.1

Similar  data  for 1963  through  1967 from
selected  cities  in the California  SCAN net-
work and  the  LACAPCD  network  are  in
Tables 6-4 through 6-9. The maximum con-
centrations for each year  and  the frequency
distribution for all years combined are shown
for various averaging times.
   Peak  values (5-minute averages) of  NO
above  1 ppm are  widespread (Table 6-1).
Nitrogen dioxide concentrations  have rarely
been measured at this level. In most major
urban  areas,  peak  NO2  concentrations  are
under 0.5 ppm (Table 6-2).
   Care  should be taken  in analyzing these
yearly  data.   For instance,  the  frequency
distribution for any averaging time is for up
to 7 years of data, thereby masking seasonal
and diurnal effects. Figure 6-12 shows a 1964
frequency distribution of NOX concentrations
for a 3-hour averaging time at  the CAMP sta-
tion  in  Los Angeles.  The measurements are
differentiated by season. As  discussed earlier,
the concentrations of NOX are lowest in the
summer  (June through August),  when  the
photochemical  reactions are the most acute.
In the winter (December  through February)
                                           2400       0600       1200       1300
                                                        TIME OF DAY (EST)
                                                                                      2400
Figure 6-11. Diurnal  variation of  oxidant
             levels during stagnation in Wash-
             ington,  D.C.,   October   15
             through 19, 1963.1
during the period of stagnating weather and
reduced  photochemical activity, the  concen-
trations  are the highest.  Similar CAMP  data
for other cities show the same pattern.
   If yearly NOX values are used as a basis for
control,  the  degree  of NOX  reduction
necessary  to  keep oxidant  concentrations
below a specified value would be much greater
than  is  justified by the established NOX -
oxidant relationship (chapter 4). When exam-
ining ambient NOX values relative  to  oxidant
values, it is, therefore, necessary to investigate
the NOX concentrations  during the summer
months,  when  the oxidant levels  are poten-
tially highest. As  in the case  of  the hydro-
carbon-oxidant relationship,7 the time period
for measuring NOX that  seems most appro-
priate is  6  to 9 a.m. This restriction of exami-
nation to  the  summer months'  data only
applies when the relationship of NOX to oxi-
dant  is being considered. For ambient NO2,
the data from the  entire year  should be  ex-
amined.
   The seasonal  variation in the 6- to 9-a.m.
NOX  values is a result of several factors. It is a
function  of the  more intense radiation inver-
sions, which are characteristic of early morn-
ing hours during the wintertime. It  is also a
function of the increased combustion of fuels
during the  winter season. In Los Angeles, for
example, this increased use of fuels produces
                                                                                      6-9

-------
 oo
 SO
O
*


I
<
£
8 ~
g2
53 3;

IB
«< o
U ft
H 33
•< H
«2s
52^

5>t
3 U A
If

Is
05
a 4
81
yP
so
o

3




GO
0
2
o
c
e
o
^

« i:
dSS



l«l
59
o
X
fc_i
O
(4-1
g
3
_G
s




«1
w £
& s
C 5
o

0

o

o
m
o
~"
— '
o
0
0
a vg
«. 5s*
3
OO

l^

VO

*O

5

m


2 
J C
»• »— 1
3 '-<3
csesm^vo oO'-i'-io4 o o — < — > — > PO^'-- ^-Hr-H(S)CS — 1 — < {N  ^o\^J"O\ r<^(sc~-ro Tj-m-Hvo
cnco(NO r^OTf i^lOiO
inv^'* i^u-itj- •^-roooooo\o t~ t^ r— r- oo oo \ovot~-r~t— o r-t~~r-ooo\o
\Or-io-H^Ht~. ^DcNoorroo oo—'r^.-iir-'*
so *o t tM — i o ^
vo vo ^ (N — i O TfmoOfnoo •*m^-^oO
o'ddddd — i — ' d d d d ddo'do'd
'fr'*tnr^
^D vo m ^ivoo oo'o^r-HO'^j- c-iTi-OTj-cht--
o\o\-^-f^—i'- vo^omn—io rofNvo^^o
dddddd dddddo —.^-.dddd
cnO.0Sjc x0
U'di_i-ic3st-iC-9bii_lcasi-ii-.l9i-iiqai:ai-ie';ai-ii-in)ak.
£P5.c;,r3T3 E X.G E^JS-O S Xg CJ543T3 S X^ ExiJ3-O S X
.H  — i oo — — . -H "lO'-ioo— i — '-i e iy> -^ oo — i — • r^ w 10 T-I oo — < — i -H
£ G 
-------
o o
o o
o o
0 0
0 0
ro ro
0 0
0 0
in in
0 0
0 0
CN CM
0 O
p- NO
ro ro
d d
m ro
ON ON
d d
O ON
NO - ro
0 0
O 00
0 -t
— i O
ro —
in ro
0 0
CN in

-------





1

ro

CN

Jfl
JO
3 u
s |
o
o
0
o
3
o
o
o
r-
o
o
CO
d
00
d
00
d
£
03
o

d
§S£
22
o o

a
o o o
o o o
o o o
o o o
o o o
o o o
O 0 0
o o o
o o o
000
odd
r- ro
d d
r^i
d
£££
«S •-! —1
O O O
ro -H o
O O O
CN <-< O
000
OO "*3" OS
o o o






-H 00 — 1
o
o
o
o
ro TJ-
0 0
o o
s
o
s
o



^H ^^
00 C5
iiO ^h
o o
o o
s s
o o
o o
0 0
0 S
0 O






1 *
0 0
0 0
0 0
0 0
ss
0 0
\O vO
0 0
0 0
2 ~
0 0
o d
d d
\O u~>
d d
ro ro








C~> 'vD
— . d
c— rs
o o


}i5
0 0
0 0
^^ c^
o o
0 0
0 0
\o p-
0 0
0 0
« 2
o o
d d
d

O — i








(N r»
O O
t^ 0
0 O


.S -t
o
o
s
o
Sl/->
0
0 0
in
O
o
0
o



ON O








O O
O O
O ^£>
0 0


I.
6-12

-------
CN
O
O
CO
o
o
co
o
0
s
o
r-
o
0
d
00
!— 1
d
in
CN
d
ON
CM
0
5
o
o\
CM
O
CN
O
r-
co
0
CO
O
CM
0
Philadelphia
5 min
CM CM
0 0
O 0
CO CO
0 0
0 0
S3
o o
o o
0 0
£•«. \Q
0 0
0 0
— i O
o o
00 CO
d d
co
CN
d
ON Tf
r- oo
O CN
CM —
0 O
CO CO
CN rt
0 0
co m
CN ~H
0 0
0 0
0 0
\o ^t
CN ^H
O O
CN ON
CO ~H
0 0
CM '3-
CM — i
0 0
1- t-l
43 43
T-l 00
CN CO
O 0
O 0
CO CO
O 0
O 0
S3
o o
o S
O 0
O 0
O 0
s
d


\o o
00 ON
ON m
0 0
0 O
O m
i— i O
0 0
O in
—i O
0 0
c- •*
O 0
0 0
O in
0 0
^J" **O
1-H ^^
0 0
O u->
— i o
0 0
•3 s




s
o







o
o
s
o
s
o
s
o
s
o
3
o
s
o
3
o
* •"
&
0 0
o o
CN (N
0 0
0 0
co co
o o
0 0
CO CO
o o
o o
0 0
0 0
00 00
o o
d d
CO CO
!— 1 «— 1
d d
— i O
CM CM
d o
00 00
O 00
CN ^H
0 0
f- \£>
o o
— ' O
CM CN
0 0
00 CO
CN r-,
0 0
— < CN
CO CM
O O




£3
•— i
O
0
CN
O
0
CO
o
o
CO
o
o
in
0
o
C--
o
d
d

00
8
o
s
o
2;
o
8
o
r— 1
O




00
•-i CN
0 0
0 0
CN CN
0 O
0 0
CO CO
o o
0 0
co co
o o
o o
S3
0 0
f-
o
d


ON CN
00 ON
in 3f
O O
0 0
NO CO
o o
o o
23
0 0
ss
0 O
ON) ^f
1— 1 ^J
0 O




•o £




co
o
0







o
o
CN
o
0
CO
o
o
CO
o
o
CO
o
o
3
0




lyr
San Francisco
o o
O 0
CO CO
o o
o o
S3
o o
§"O
0
0 0
OO 00
o o
o o
00 00
^H l— 1
o d
—i ON
CO CM
d d
co oo
T}- CO
d d
0 0








CN — i
in ^J"
O O
oo co
CO CO
o o
o c-
in CN
0 O
£3
§ 43
in ~
^H CN
O O
0 0
CO CO
0 O
O O
SS
o o
o o
0 0
§0
0 0
^O CM
d d
CO
CM
d

in in








— < oo
CO —H
o o
co co
CN <-H
0 0
\o o
0 O
^ >,
43 T3
oo 1-1
CM
O
o
3
o
in
0
O
0
0
s
o



CO
oo








OO
o
o
o
o
o
o
8
CN CM CM CM CO
O O O O O
0 O 0 0 0
CO CO CO CO CO
0 0 O 0 0
o o o o o
co co co co Tj- Tt co
o o o o o o o
o o o o o o o
SS3SS
o o o o o
>sO ^«O ^>O ^O W)
o o o o o
o o o o o
O O ON 00
— — i O O
d d d d
odd
CM ~H
CM CM
d d
O ON ON ^f ^O m O
O P- r~ oo oo ON O
O TJ- co oo in in
(N CN •— i O O O
O O O O O O
CM CN -H O O O
o o o o o o
ON oo ON r~- •* co
i— H i — i o O O O
o o o o o o
m co o r~ ^- co
CN CN — H O O O
o o o o o o
NO ^- CO ^- O ^1" Tl"
O CM CM ~H -H O O
O O O O O O O
m ^j- CN -H ON ^- CO
O CM CN — i O O O
O O O O O O O
co t- o o r~ •* co
O CO CO •— • O O O
O O O O 0 O O
o
^Si.g^^^o^
•£,.5 gjSJs-a 6 >>
"1""°°



















o
1
E
a
N
•4-1
$
O)
O
c«
Determined by continue
6-13

-------
5
2
 IU
I


i3


O
X
cfl
§
S
£
o
8
8
<*-H
O

+3 -
Q J




S
M
"e
=>
.§





_ t
T= !
S •
S
o
ON
o

o
v>
0
f>
O
' "*
—
d
o
d
q"
a
S

t^
vo
vo

10

S

m

CM

2P
3)
T)
i i
v£>
O
O
8
o
2
o
vo
O
04
O
O
d
O4
VO
d
r-
d
r ^
o o o
5§ 5 $
o o o
0 S °,
-i 0 0
oo *~* ro
^O to ^^
o o o
ON -3- (S
O O O
^ ^ -I"
0
0
2
o
n TJ-
o o
m
0
£
O



ON O
00 O
IT) CS
0 O
2 2
O 0
ON vo
O O
£ 2
O O
S 2
0 0
2 2
o o
cs 2;
o o
1*
o
o
§
o
o
o
0
0
ro
0
d
d
0

ooooo
CO CO O< O4 O4
ooooo
dddd
t-~ P- — <
d d d
00 ON
CN O
333SS|








^ ^ ?S £ S ^
— < — i O O O O
2 o P 5; 8 2
_ — i o O O O


'3 i2 H ^ <-f H
tt J5 J5 *o C X
IO —< 00 ~H —1 ^H
6-14

-------
CO
o
O
m
O
O
O
o
o
0
r-
o
CN
d
3
00
o
r-
CN
00
T-H
00
— <
s
CN
2
— <
O
in
— '
ON
--i
00
—
Philadelphia
5 min
CO CO
0 O
0 O
m in
0 O
0 O
r- c-~
o o
o o
0 0
0 O
r- vo
o o
r— 1 VO
•<*• co
d o
o o
O OO
— d
ON
vq
^ in
rt 00
VO ON
— 0
CN ON
vo t-
-i 0
CN rt
O -H
CN -H
r- o
o o
CN •*
*-> 0
ON VO
vo O
-H ,—
£ 0
~ -<
t-( VH
— 1 00
Svo
0
0 O
\Q t —
O 0
o o
00 OO 00
o o o
000
2 §
0 0
in CN
0 0
ON
CN
d


VD *-H CO
C^ OO 00
J in O
o o o
CN in O
Tf r-. — ,
O O O
OO 00 O
in — i ~*
d d d
OO O 00
CO -* O
odd
OO O 00
CN — i O
o o o
O CN
o o
VO CO 00
in — H O
O O O
a" ° <-<
t— * i— i »— t
CN CN
0 0
0 0
CO CO
0 O
O 0
o o
0 0
0 0
0 0
CN CN
0 0
•* CO
CN CN
d d
ON r-
CO CO
d d
«-i m
vo m
d d
co co
CN VO
m Tf
O 0
CN oo
0 0
CN VO
o o
ON •*
o o
CN (N
O ON
~ O




M C
115
CN CO
0 0
0 0
ss
o o
vo vo
o o
0 0
r- r-
0 0
o o
r-* O
o o
oo m
*— 1 *— 1
d d
r-
CN
d

00 CN
r- oo
-H >n
CN — >
o o
OO CN
CN CN
0 0
ON m
CO CO
O O
CN — i
O O
vo in
CN CN
o o




t- a
oo i—*
s
o
0
0
vo vo
0 0
0 0
C--
o
o
£
0



in o
00 O
gvo
0 0
00 VO
O 0
O 0
SB
d d
r~- m
o o
d d
oo r-
0 0
0 0




S x




























San Francisco
co
O
O
VO
0
O
O
0
in
O
vo
rs|
0
CN
d
CN
r-
d
CO_
in
VO








ON
O
ON
vo
— «
oo
in
O
C
in
CO CO
O 0
O O
SB
O 0
O -H
0 0
in m
O O
vo m
CN CN
0 0
in •*
d d
00 ON
vo 't
d o
(N
ON
d
in ON
vo vo








VO (N
f— m
0 0
in vo
co in
^H 0
ON —
**}" CO
O O
~* oo
vo r--
o o
o o
g2
0 0
CN CO Tt
000
in •<*
O O
CN VO
CN — H
0 0
0
CO
d


oo in c~-
vo c~- vo








-i O Ti-
ro CN —i
O O O
VO VO Tf
CO -H -H
O O O
CO O
CN — <
0 O
w1 ° >-
CN CN CO CO •*
O O O O O
o o o o o
co Th TJ- in m
00000
o o o o o
m in m vo vo r~
O O O O O O
000000
r- r- r- oo oo
o o o o o
o o o o o
Tf ^f CO CN O
o o o o o
oo oo -H m
CO CO CO CN
dodo
0 -H Ti-
r- f^ in
odd
§r-
ON
*-i O
^ ~H m oo VD O
c~- t-~ r-» r- oo o
O oo ON OO Tj- ON
O f~ TJ- CO — < O
— i O O O O O
t-^ O vo in -< oo
•
-------
   ~
 Jj
 J=
0
o
. f*^
"3
o
o
•a
a>
"3 °
g. w
D
o
1 2
u
o
o
o -1
O „
^v o
o
o'


g VO
1—
i s

ea
S

CO

Place, site No.,
averaging
time
O O O <-H
OOOO
dddd
§•-1 •-! (N
000
dddd
ooooo
ooooo
OOOO
dddd
OOOO
d d d d
(N OO CN
odd
O oo
d d
d
\O **~* 00 I*** ^f
\D ^* ^^ ^^ f^
ooooo
°s^ss
ooooo
o ON r- g
OOOO
C? O"*i ON ^^
CO f^ ^^ fj
OOOO
ON oo — < vn CN
ooooo
7
o o o — <
OOOO
dddd
gO *-i CN
000
dddd
— i o -H -H
do d d do
tS C4 OO
odd
10 m — i o o as »o co — < o CN — i o o o
ooooo ooooo ooooo
10 ^^ *~^ ^J ^O CO C^J *^ ^D ^^ ^D ^3 ^D ^D
dddd doddd odddd
^(N^OO ^Om^^O
d d d d d d d o' d d
SS2SS S3^So
ooooo ooooo
oo^"CNOO ON^oroO
ooooo oooo
t- T
VO ^ g,
fj" 0
M^OO-Hrt^-,^^00^-^— • -g « 00 — < — i —
£ O PH
6-16

-------
ooo
ooo
o' o' o'
ooo
odd

ooo
odd
ooo
d o' o'
O O ON
rt _. O
odd
r~- en oo
CN CN -H
odd
d d
d
oo m — i
in en es
OOO
•<*• m o
r-~ •sf CN
ooo





00
CN
•*
T3
2 >- i- w1
E J3 J2 T3
•3 ~ °° -
s
o
o'
CN
O
d

o o
d d
o
o'
S
d



oo •^t
O O
O O
c~
o
o





1 mo
lyr
Riverside- 126
§o o
o o
o' o' d
ooo
odd

ooo
odd
sss
d d o'
i—H ^^ O*S
^H »™f ^^
odd
ON — i 00
odd
f- 00
<*• en
d d
•*
d
in CN CN
OOO
en NO o
Tj- CN CM
OOO
>n en CN
OOO
o ON oo
— 0 0

T-l 00 — I
o
d
0
d

en TJ-
0 0
d d
in
O
d
d



oo •*
0 O
o o
oo •*
o o
o o
o ^-
-, 6
O 0
-
o

o
eN
-*->
^ I V*
o
888
odd
ooo
odd

ooo
o' d o'
en en Ti-
ooo
d d d
888
odd
C— O\ CN
en exi eN
odd
0 x}-
r~ 
0 0
-H 00
88
d d
0 0
d d

ooo
d d d
en en
0 0
d d
in -3-
0 0
d d
ON
O
d


•—i in en
«-i O O
ooo
8£$
0 O
O ^f
O 0
o o
0 O
ooo
ooo
«" o
8888
d d d d
oooo
d o d d

00000
o d d o d
CN en rj- in
OOOO
d d d d
o o o ON
_ « ^ o
d d d d
oo NO o
en CN CN
odd
>
c — 00 — — —
6-17

-------
    -₯
i
VO
H
o
Os
O
•o r-»
•8
1>
m
o
*. 0
O
«
3 O
cr en
>
j.. -H OO —C -H 1-H
^
oooo
oooo
dddd
oooo
d o' d d
ooooo
ooooo
oooo
dddd
t — t — VO ^O
O O p O
dddd
CN <-< — i
odd
d d
S
d
VO (N -H O O
ooooo
CN oo ^ r-- en
i/l « — i O O
ooooo





7 S
*^^tlfc
^ ^H OO *— 1 ' — ' ' — I
8^n — i CM OOOO OOOO
ooo oooo oooo
dddd dddd dddd
»— 1 (NJ fNJ fSJ Q Q O ""^ O O ^* ^^
OOOO OOOO OOOO
dddd o'ddd dddd
ooooo ooooo ooooo
ooooo ooooo ooooo
c^ oo oo oo *—* CM cs en CM CM en en
oooo oooo oooo
dddd dddd dddd
^,1212^H OOOO OOOO
dddd dddd dddd
^- Os l/"} CM CM 00 ^~ cn Os
en CM CN cn CM '-•' CM CM *~*
odd odd odd
in rf in Tf rj- en
do do do
OO OO OO
vo vo vo
d o' d
voencMOO en^OOO
OOOOO OOOOO
SB 2 £322 SBSS-S
ooo oooo ooooo
cn CM ^H i— i c^Tj-cM' — ^o TfencM^^o
o'ddd ddddd ddddd
VO ^* CM *— i cn '•H t~t O in CM »~H O O
oooo oooo ooooo
in oo vo r^ oo oo ^ ^i-
r- •* CM •— i en~H"O
dddd dddd
0 2 S
^ x 1 >> ^
J+j

-------
%
3
3
£
£
££
p -

II
58
« H
Z w
O ^o
P2^


lei
§ w
S 3
Z O1
S «
U K
w fa
08


I-

P
o

i
S
o>
o
0
r-

o
o ^
E
•s -
^. &
O
O
o-


ta ^O
UH
S *°
3
'S
^ s
ro

Place, site No.,
averaging
time
oooo
d d d o'
oooo
d d d d
CO CO CO ^" ^"
ooooo
ooooo
oooo
dddd
oooo
dddd
in f"> O
odd
tS — '
o d
d
^««^
ooooo
t""- l/^ C^ Vl ^"
£Nj f-H ^-i ^^ ^3
o' d d d d
(N — i — i O
dddd
CM en — < •*
d d d d
°, £ z; o o
ooooo
i£ °i
7 4
p >
"°
vo (N O T3
7 2 1 1
1-1 ^ O ^
(i. O eu ca
                                               6-19

-------
O


o
S3
2.

3^

^s
& ^
ca  5

l|

PE
a vo
M O
5^ i—i ^-v

Be S
U ** n,
5 U &


If"


ss
  Q

  Z
  tu
Z
|


o
u
o
o\
0
1 o
0
3
1 %
o>
O
1 2
3>
O
o r_1
E
o — <
^ o
o
d
r-
VO
i- \O

S -n
1
cs
s ^

en
MS
d
Z °°
S H g
•a 2 .E
oooo
dddd
(N rs CN en
oooo
dddd
OOOOO
OOOOO
S^r ^ -
C-JC^m^- CNtNCNen — i T-H — i rN -H — i ,— i »-i ,— ,
pppp oooo oooo oooo o
dddd dddd dddd dddd o
en^-^^f'^l- mmmen^ (NfNCNfNo) r>irNcnen^t ,—
OOOOO OOOOO OOOOO OOOOO o
OOOOO OOOOO OOOOO OOOOO o
inmmm •* Tf ^- •»!• CMCSCNCN ^-TI-TJ--^ en
popp oooo oooo POOO o
dddd dddd dddd dddd o
pppp pppp pppp o o p o o
dddd dddd dddd do' do o
ooin^" rfc-l' — i \Oin*x£ encNO f\j
odd odd odd odd o
do do do do d
^ O * — i m \s*.
Tl* en • — i r^ rn
ddddd
eneN— -^ooo rs — • i •— i O o §
ddddd ddddd dddd dddo'd d
••sfencNOO enrN^^OO ' — ^OOO i/^
ddddd ddddd ddo'd o
mcN — OOenrN—  C >n g o
 6-20

-------
O O -H
OOO
odd
OOO
odd
O4 O4 OI Ol
OOOO
oooo
ro co co
000
odd
vo in ^f
OOO
odd
d d
00
d

t~- oo ro fN
I-H O O O
oooo
ON O) VO (O
•-i — < O O
OOOO
O4 O4 Tf Ol
Ol -* O O
OOOO
in o-\ vo 04
— i O O O
OOOO
OO O4 VO CO
— -H 0 0
o' d d o
rt* O
J3 T3 C >>
00 _ rt ,_
§OOO
p p o
dddd
d d d d
o o o — •
("™i f~} f™1* ^"^
oooo
oooo
dddd
r)- Tt- ro O4
OOOO
dddd
o o> r-
-H O O
odd
»— « »-H
d d
d


12 " o
ooo
f- OS m 01
i — i O O O
oooo
O ~-< o\ in
Ol —• O O
OOOO

O
7
5? - - & §
a -c^^ s
c ^ °° ~ ~
8
d
0
d
0 g
o d
ro
0
d
o
d
CO
d
in
OI
d
in
ro
d
-H
0
r--
ro
0
O
O
0
0

00
O
o
00
!i .** IH
" c ^
OOO
odd
—i O4 O)
000
d d d
01 01 o) to
oooo
oooo
ro ro ro
OOO
odd
in in >n
OOO
odd
tN O
d d
00
>— 1
d

CO 00 ^ O)
•— i O O O
oooo
in ^ vo to
01 ^H o o
OOOO





-C T3 G >t
00 rt ^ _
oooo
dddd
ro to to '*
OOOO
d d d d
ssss
oooo
oooo
dddd
§o\ oo r~-
000
dddd
OO VO CO
i — i . — i * — i
odd
OI O4
0 0
ro
d
04 -H -H O
oooo
r- o vo vo
OI O4 -H O
oooo
ro vo CN f-
0 0 0 O
co co " O
oooo
o o\ ro r-
ro ~H — i o
d o' d d
VD
CO
j-j — 1 00 — 1 — H


g
0





s
o
s
o


0
o

CO
1
2
a
"1
i-H "*"*
1
O O
d d
o o
d d
O4 O)
0 0
0 0
co ro
O O
d d
0 0
d d
§£
d d
d d
00
d


2§
0 0
— i CO
O O
22
o o

--I 00
o o
d d
o o
d o'
OI CO CO
ooo
ooo
co ro
o o
d d
SCO
0
d d
r--
o
d




VO CO
0 0
oo TJ- co
000
gss
ooo

N
^» I-H
•O 5 >> a
>
— I OO — I —I ^H














•d
0
b
CB
ss-Saltzm
aDetermined by continuous Grie
6-21

-------




e>
z
o

2 \O
£*
3B
II
z <*>
33^
JUl ^^ CU
z jy
zjt
8 g
w Q
Q H
w P
o
t_i
z
VC
a>


T3
-O
§
CO
V-i
T3
"J
o-
sntration
S
o
o
6
•s



53

.8
S



o
«.£
•a ;
•* <
CJ £
O 0
O
O
O

O
ro
O
-
d
0
d
r-

VO

m
VO
^

r^

«
5 73
3
•^ (N r~l CN
oooo
dddd
CO ^^ ^" V^
oooo
dddd
i/^ i/^ I/O ^O ^O
ooooo
ooooo
t— C-- O 00
p p p p
dodo'
dddd
odd
O 00
d d
(N
t-~
d
\o o r~ •* oo
t— 1^1 CS — i O
ooooo
in en  Tfm 00*^3 i— <•— 1
do do do do
O\ O ro ro
10 r^ p c^
d d ~H d
^- O i/o f^l p^ vO OO ^H OS ^i" ^O Tl" I-H i/o Os OO i/o i— i *O ^"
^O^CN^HO ^OrocNOO O^-O^"1 — 'O C^' — 1< — 'OO
ddddd ddddd — * d d d o' o'dddd
raoo^og §-£££ os^^^g osr^ogg
OOOOO OOOOO OOOOO OOOOO
odd ddddd do' odd
££2go- ^pt°2-
OOOOO -HOOOO
^* CN ^^ o os 10 rs o O os i/o r*"i c^
oooo ooooo oooo
OS
S r- 1
± s s 1
^ >^1 1 "
U tN 1 (J
_S _S •« £ S,°_S-S-3 S Si^-S^-S E S,u_S^-S E S,
cd t-i ^ O
CQ fc O CL,
6-22

-------
CN
O
d
CO
o
d
m
O
O
8
d
m
d
CO
co
d
o>
in
d
o
ON
d
0
ON
O
O
00
0






Richmond-428
Ihr
CN CN CN
O O O
odd
co -
CO CN — O
OOOO
in CN NO CN
CN — O O
OOOO
in CN r— •*
CN — < O O
OOOO
in
t"- H
1
o
.S
•T3

| ££•§ E
- 00 - -
A
OOOO
dddd
_ — CN CO
OOOO
dddd
oo CN co co »n in
O OOOOO
d d d d d d
m NO r- oo
oooq
dddd
m m rf co
d d d d
in co o
Tl- CO CN
odd
in —
r-- m
d d
in
oo Tf TJ oo ^t m
O oo •* CN — o
O OOOOO
CN CN NO ON 00
Tf in co — O
— OOOO
oo — ON co in
ON m CN — O
OOOOO
00 ON O CN •*
— NO CO — O
— OOOO
CN NO NO Tt — NO
O ON in co CN o
O OOOOO
O
I1 >. o
kH .^ W fcH Cd ,-1 t-l
>, (5 •£ •& *° « >-
— j., — OO — — —
&














-n
.§
•*-»

-------
    VO
1
•*-»
 §
 u

vl
 01



•o

0 0
0 0
ON ON
0 0

\o
1-H
o








2
0
en 10 CN
— < o o
o o o
oo o en
-H -, 0
o o o








b o
•S s >,

0 0
O 0
0 0
O 0


ss
0 0

VO vo
o o
O 0
CN CN
O 0

ON •*
CN CN
0 0
CN O
vo m
0 0
en
^D
0
en O
vo en
0 0
vo en
0 0











oo
o
1
o
BO

c — i oo
— i CN
o o
o o
CO 'sj"
0 0
0 0


*d~ to vo
000
000

VO C —
o o
o o
2g
o o

ON
1— <
0






— i en 10
CN — ' O
0 O O
Tj- Tf VO
fN — 0
o o o











VC
t-


b o 2
•3 e >,,=
c
c/
en en ^f
000
o o o
10 >o vo
O 0 0
000


00 00 00
o o o
o o o

CN CN en
•-H — H — 1
o o o
CN^O
o o o

\Q oo ro
^" f"l CO
o o o
O vri
r** \o
O 0
ON
oo
o
00 •* CN
000
00^
O 0 0
CN ON *
LO CN CN
000
ON P- en
oo c — ^*
o o o
ON >o ON
oo vo en
o o o


5


ij ^

M '•"' OO *~H
!
o o
o o
r^ CN
0 0
o o


2S S
00 0

<0
— i O
0 O
r-
0 0

J^.
CO
o
VO
VO
o
^
c
o
10 ON
o o
g 3
0 0
r^ vo
' — i C~-
0 0
-i C-
CN) T^-
0 0
^
CN
o

10
2
a
•s

£ >> ca -^
J
o
o
CN
0
0


0
o

.0
o
o
-i
o

J —
CN
0
_
"^1"
0





z
0

^~
o
m
fNj
0









X
00
CN CN
o o
o o
CN en
o o
o o


o o o
o o o

•0 vo
0 0
0 0
-ON
0 0

(N
CN
o








CN — <
0 0

CN ' — i O
o o o
en vo
-, O
0 0





*fr
)
3
"** O r~\
•S S >, «
1
S 0
o o
O 0
o o


0 0
0 0

CN en
0 0
0 0
0 0
0 0

O 00

0 0
VO — i
I— 1 1— t
0 0
o
CN
0
CN -H
o o
~ §
0 0















Js J3
— i oo
0 O
o o
CN CN
0 0
o o


000
000

m en
0 0
0 0
S 0
0 0

10

0






o o o
o o o
vo en
o o
o o














>> 0
•3 S >,

6-24

-------
--i  >-i  (N CN
q  q  p p

d  o°  d d
                           —< *-i  cs  M
                           p q  p  p
                           d d  d  d
p  q  p p
dodo
                            r') ^fr  in
q  p  p p  p

d  d  d d  d
                          r^ ro  rf  *t 3
                          p p  p  p p

                          d d  d  d d
^J-  in  m m
0  0  0 0

d  d  d d
                          m  m  m  •*!•
                          q  q  q  q
                          d  d  d  d
o  —i  o ON
i-H  T—H  ^^ ^J

d  d  d d
                          ON 00  C--  VO
                          q q  q  q

                          o' o'  d  d
CO  (N  tN

odd
                          C- ro  1-1

                          odd
(N  ro

"1  ^
d  d
                          ^o  oo
                          (N  --H

                          d  d
C-;

d
Ol  •—(  C4 >n  ^
•*  CN  ^ q  q

d  d  d d  d
o   (N  '—i  O

d  d  d d  d      dddd
roNOTfCNm     m oo  ro  NO ^f

d  o d d  d     d d  d  d d
OO  ON OO ON  ^O
NO  f*"> '—I O  O

d  d d d  o'     d d  d  d d
                      oo
                      (-S)
    d  o d  d
                      NO in  ^H
                      >  o
        cd  is
       T3  K
               >, 3
                                      2
                                      S

                                                   o

                                                  1
                                              O

                                               f/3


                                               O
                                               O
                                               o
                                               
-------
  l
 ** ^
 z -

 ?>
 z u
 5 z

 Pg?
  a:

  Q
 u w
 w s
 Q e

 §1
 £3
 u

 I
0
o\
o
13 "^
o
V
§ 0
O
T3
"rt o
1 ^
C
O
'•S o
J= -1
1U
§
o T—
S
^ -H
^v O
O
d
t-

^

6 S
§ SO
("& *^"

CO

Place, site No.,
averaging
time
oooo
dddd
oooo
dddd
-H —i — i (N C4
OOOOO
OOOOO
opoo
dddd
opoo
dddd
^ 0 0
odd
o' d
d
o r- c4 ro /•> '•o oo O oo o «— i m TJ- ro ir> vo r~- as m •** i/-> i/-> \o
OOOO—i o -H " — ' -< OOOOO OOOOO
OOOOO OOOOO OOOOO OOOOO
o <"~t CN Tt° r"~ oo oo oo os o "^ *~^ c — t — t — r*-
dddd o'ddd dddd dddd
OVO^-H l/O— o uu £r* o ^^ ^*^ o ^^ ^* o
CH M M TO ^ VH ^^ ^WlTO^S— (f^(— it— (Cn^l— lC^H*^^5'~ '
nxJ^'O C X'QXJ^'^ C X^J^^Td G >,^^^^ G >>
3 >» O O
PQ ffi hJ J
6-26

-------
•-« CN
0 0
d d
CO Tfr
0 0
0 0
vo o
0 0
0 0
CN CN
d d
r— in
CN CN
o d
co 3
m •3-
o d
r-- vo
d o
in
p
in ~H
O vo
-< 0
O Tj-
oo in
0 O
in •<*
r- 3-
0 0
m co
CN OO
— i O
co in
ON vo
d o
CN
•s
a
60
C JB b
hJ -H 00
z ;
co co
0 0
d d
•>tf- m
0 0
d d
r- oo •-<
0 0 -H
000
CO •> 0
O O O
odd
-H CN CO
o o o
odd
CN CO ^
o o o
o o o
in m vo
O O O
odd
CO -H ^H
d o' o'
co •* ON
ro CN -^
d d d
ON co
Tf CO
d d
CN
vo
O
CO CN O
in co CN
0 O 0
O 00 C~
m CN -H
O O O
CN CN VO
rj" CO '—|
O O O
CN r- co
VO CO CN
O O O
ON ON in
VO CO CN
O O O
1 ~ °° ~"
s
d
CO
o
d
Sin
0
0 0
VD
0
d
o
d



O in
rt o
0 0
O in
—i O
0 0
ON •*
0 0
0 0
co in
— i o
0 0
£ S
O O
in
t^
1
o &
CS t G
0 0
d d
CN CN
0 0
d d
ss
0 0
00 00
0 0
o d
o d
co in
CO CN
0 0
«n CN
TT ro
d d
O
d
vo co
0 0
in ^H
Tf CO
0 0
C"*"11 ON
0 0




l-l 1-1
*-< 00
o o
d d
0 0
d d
in vo oo
o o o
o o o
OO f-
o o
o d
d o'
o
CN
d


^f m oo
CN — i O
O O O
ON CN VO
rt — i o
o o o
00 O
o o




& $ a
^3 C >•»
1—1 1— 1 1— I
0 0
o d
CN CO
0 0
d d
f^ f"~^
0 0
CN Tj
i— ( »— i
d d
o d
O VO
00 VO
d d
ON vo
CN O
in
0 0
OO Tt"
-1 rt
in VD
r~ o
— --I
CN 00
O vo
— 0




VO
r-
g >- M
a " °°
CN •*
0 0
d d
•^- vo
0 0
d d
OO ON CN
O O ^-i
o o o
d d
CN CN
o d
in
in
O


T)- •— < m
ON co -H
O O O
O O CN
f- CO -H
O O O
in vo O
m CN <— i
O O O




^ O
•S 6 x
o o o
odd
^H « CN
o o o
odd
-H 
-------

T3
0>
13
o
x
o
13
cr
CO
e
centratic
o
CJ
1
'o



S3
*
e
|
I



d
D .
O
cS
O
ON
o
o

o
CO
0
-
d
o
d
r--

XO

m

^

ro

averaging
time
O 0
d d
O 0
d d
o o
o o
0 0
O 0
ON oo
0 0
o d
d d
ro O
o d
04
d
ro 04
0 0
— 0
0 0
03
o o
S£
0 0
So!
0 0
s_
O4 CO
0 0
d o
0 0
d o
£"} C) f~ ^
o o o
0 0
d d
00 NO
0 0
d d
d


— < o o
o o o
— < o o
o o o
\O 'O t
—.00
o o o
2 0 0
000
~8S
o o o
^ c b*
'OCX
o o
d d
o o
d d
O 0
o o
00 00
^3 ^3
d d
»— 1 r-H
d d
NO CO
d d
00 NO
ro ro
o* d
d
i/-> •*
O 0
T)- en
O 0
5:^;
0 0
O m
O 0
ro ro
0 0
O
m Tf
0 0
d d
Svo
0
d d
O 0 0
000
OO 00
0 0
d d
d d
d


— i m o\
ro —i O
000
CN — < O
O O O
cnog
000
^g%
o o o
Tl- i — i \O
04 —i O
o o o
^ C *"•
^ c >>
ro m
0 0
d d
0 0
d d
\o t-~
0 0
0 0
00 O\
0 0
d d
d d
— H
d d
d


O4 — i
0 O
ON O 00
O O O
°,~§
0 O O
ro O oo
O O O
04 - 0
O O O
•1 i s;
O4
0
d
s
d
o
o
00
o
d
d
O)
d
O4
rf
d
0
d
in
0
00
o
o
t-~
0
0
o
IS
o
7
V3
—
"a3
ro ro •*
0 O 0
odd
SS8
odd
o o o
0 0 O
00 OO 00
o o o
odd
ro 04 O
odd
O4 — i
o d
ro
ro
0

ro 01 — <
000
ro 04 •— <
O O O
ss =
000
ro 04 0
O O O
ors-
o o o
00 — 1 — i
0
d
s
d
o o
0 0
r-
o
d
O4
d
d
o
t
0
0
d
t-~ oo
0 0
oo oo
o t
0 0
o >/->
0 0
§ ?
d d
s s
o o
o
1
fSl
"«S
*^ VI *~*
O4 ro in
O O O
odd
o o o
odd
^O NO 'O C —
OOOO
OOOO
oo oo r^
o o o
odd
O4 — < O
odd
d d
ro
d

ON oo ro r~-
OOOO
m ON O NO
OOOO
ro 01 — < o
OOOO
Os O^ 0s! I""-
04 ^H O O
OOOO
ro (N ^! 0
OOOO
(-1 C3 ^ ^M
_C T3 C >>
00 -H « — <
6-28

-------
CM CM
0 0
d d
en •*
o o
d d
in m
0 0
0 0
o o
d d
en CM
d d
•3- -*
CM CM
d d
in ON
en CM
d o'
0
m
d
so en
^J- en
O O
CN 00
Tf CM
0 0


ON en
in "tf1
0 0
r- oo
m en
0 0
CM
f-
i
(O
*" 43 43
^ ~ 00
z
en
O
d
o
o'
in
O
O
oo
O
d
d
ON
d


CM
o
r-
o


CM
O
00
CM
0
-a
3
d
0
d
SO s£>
O O
O O
oo
O
d
ON
O
o'



en oo
d d
ON so
0 0
0 O


og
o o
2§
o o
1 mo
lyr
Pasadena— 64
CM CM en
O O O
odd
d d d
m in in
O O O
o o o
t- r— r--
o o o
odd
*-i O ON
odd
O oo en
CM — > — i
odd
oo en
CN CM
d d
d
C*~ C""" so
en CM --<
O O O
en t — so
en CM I-H
O o O
<* r- m
o o o
so CM in
en CM I-H
O 0 0
en en m
en CM i— i
O O O
51?
3
d
in
O
d
o o
o o
o
d
00
o
d



ON SO
0 0
0 0
£-— SO
O O
O O
r** ^o
0 0
O 0
OO \O
0 0
O 0
oo in
0 0
o o
in
1
| >, §
~ -• i
en en •*
o O o
odd
^f in in
o p o
odd
o o o
000
OO OO OO
o o o
odd
CM — O
odd
o p- •si-
odd
r- CM
CN CN
d d
d
r- CM so
en cs T-H
o o o
ON CM so
CM CM •— i
o o o
ON ^- so
CM CM — (
O O O




*-H 1-1 ffj
— i 00 — i
in
O
d
d
0 0
0 0
t—
o
d
00
o
d



o r-
~ o
o o
t"- so
O O
o o
00
o
o




so
o A
a &!
1-H 1— 1 2
C^ ro ro
o o o
odd
sss
d d d
in so so
0 0 O
0 0 O
t~- 00 f-
000
odd
en en CM
odd
oo r- ^
CM CM CM
odd
-H 0
in in
d o'
in
so
0
r~- •*!• CM
so *n Tf
O O O
O oo O
•^- CM CM
O O O
en so ON
tj- CS — <
0 0 O




>•
VH VH TO
43 XI 13
— i 00 *-i
s
d
in
O
d
so so
o o
o o
r-
0
d
o
d



SO 00
0 0
O so
— i O
0 0
-H SO
-H O
o o




"H ^1
&
•— i CM CM Tt
oooo
dddd
CO CO ^J ^O
oooo
d d o' d
ooooo
ooooo
so so so SO
oooo
dddd
^^ ^^ C^ C"*1
-H -H O O
o' d o' o'
CM rt ^H
odd
en oo
en CM
d d
CM
d
^f tj- en ON so
•* en CM O O
OOOOO
in m in so m
CO CM i-i O O
OOOOO
en in f- so in
ooooo




i? ^
CM CM en ^f
OOOO
dddd
oooo
d d d d
ooooo
ooooo
SO so so f-
o o o o
dddd
rt -H O 00
d d d d
m o so
CM CM .-H
odd
t- 00
en CN
d d
ON
d
in en CM O O
OOOOO
r~- ON ON oo so
•3- CM — i O O
ooooo
£ $ CM ~ 0
ooooo
O m oo C~- in
Tt" CM -H O O
OOOOO
« oo oo ON in
•3- CM — i O O
OOOOO
7

CO
Determined by continuous Grie
6-29

-------
8-
w \0
v-i o\
a -

d S
w
O
O


O
c«
z2
i—i O\
f)  ~*

15
H Z
•< W


Z W I
w otf -S
Wfe

is
u %
1/5  ,
w w


§1
z z
w rs

o!
B|

z^


2"
s
o
Os
O
•0 ^
o>
o
% °
0
1 °
a- f
ID
C2
^O
2 o
e
H
o
o
1
'o _,
^^ o
o
d
c-

^

"s ss
3 \0
03
^

en

Place, site No.,
averaging
time
ooop
dddd
en •sfr "^ ^O
OOOO
dddd
o o o o o
o o o o o
oo oo oo t^
o p o p
d d d o
fS i-i O 00
-^ —i -^ p
dddd
odd
m oo
d d
10
d
ir> en eN O O
O O O O O
oo en ~ oo g
ddddd
m csl ~ o o
o o o o o
oo ON ^" ON C*^
0s! *™t *-*^ (^ ^^
o o o o o
go^gg
o o o o o
o
3 -H 00 —1 ^H — 1
rf Tt w^ oo
p o o o
dddd
r- oo o\ — i
dddd
^csen^vo
00000
O^ ON ON *— <
OOOO
ON ^O *A> ^H
CO CO CO CO
dddd
—i O -*
odd
iy-i r-
d d
2
oo OA *™^ ^O ON
^^ OO C*"* CO '~~'
-^ o o o o
g £ £ $ S
-^ o o o o
•— i O "-O en —^
-H O 0 0 O
00 IO Tj- CN •— i
O O O O O
es f— m en —<
— i O O O O
T
•g ^H 00 « -H ^H
t"^ ON O CN ^" ^d~ V) oo Tl* w^ 10 oo
p o -^ — i oooo oooo
dddd ddo'd dddd
»— H fNj co 1^*1 ^o r"^~ oo ^D £**• r** oo ^^
T— H i—1 ^H f— H f^ ^^ ^^ i— < ^^ ^^ ^^ 1—1
ddo'd dddd dddd

ooooo ooooo ooooo
io*o^or^ OOOOONO Tfi/">i/iio
dddd dddd dddd
^t^t'ten enenenes eNe>|e>i-^
dddd dodo' dddd
\o— n^t en^oo oooen
odd odd odd
>/1 OO ON f- 00 C~~
do do do
rs rs rj-
« fN OO
-; — ' d
OooenO ^t'enO'^l'C^ ^tesu^enen
- — ^r^^O1^" enoo^oen*-H oo^o^fcN' — •
— ' O O O — < O O O O OOOOO
g^en^oo) °=g§:°^ ooSme^^
— OOOO OOOOO OOOOO
c.go^oc, -«^;^- SSSSS
«ddo'd — < d d d d ddddd
SSSPi?3 SS3£8:S ££££2
•-^ o o
6-30

-------
Sin
0
d d
OO 00
O 0
d d
CN ro
O 0
ON ON
d d
r- m
ro ro
d d
r- ON
NO in
d d
ro oo
ON F-
d d
ON
•> 0 c
•Sis;.]
i-H *— 1 »— 1 C
C
a
ro ^f m
O O O
odd
NO r— oo
o o o
odd
00 ON O
0 0 -H
o o o
CN CN ro
odd
CN ON 00
CN -H •-!
odd
m f- oo
•>§
~H ^H g
£
Sm NO
O O
odd
P- 00 ON
o o o
odd
- ~CN
O O O
NO in NO
odd
NO ro ro
CN CN CN
odd
ro NO oo
•* ro CN
odd
f- ro
m ^1-
d d
NO
d
t~- T)- ro
NO ^ ro
0 O 0
ON ro ON
in ^ CN
000
ON m c^
in ro CN
O O O




V. H &
X X T3
^H 00 »-1
ON
O
o'
d
CN m
O O
in
d
o
CN
d



ro m
O O
ON CN
0 0
in
O




NO
o ^
ro m NO
000
d d o'
P- oo ON
O O O
odd
— i CN ro
O O O
O CN CN
CN CN CN
odd
NO CN O
"fr rj- •*
o o o
§ro ON
00 t-;
^ d d
c- r-
m CN
ON
00
O NO ro
— H \O -H
CN -H -H
in -H ro
ON CN oo
— , —. o
-H oo r--
CN f- NO
— < o o




X X -S
— i 00 -*
00
o
d
d
m oo
o o
CN
d
r-
ro
O



CN ro
0 0
O 00
0 0
ro —i
O O




O
o.l
SfcH C^
>> *&
~~!
o q
d d
0 0
d d
0 0
0 0
»— 1 r— 1
d d
d d
ON OO
in rf
d d
ro m
00 NO
o d
o
q
O m
— 0
NO O
00 NO
o o
rt ON
-H 0




X X
— < 00
S 0
d d
0 0
d d
8 g 2
o o o
d d
d d
ON
ro
d


r- CN ro
0 0 O
O ON CN
^" r-H i—*
o o o
CN r-c O
000




>, ^
•§ G x1^
|
o o o o
d d d d
o o q q
dddd
o o 2 « 2
ooooo
dddd
-H oo r- ro
ro CN CN CN
dddd
m O NO
NO m rl-
odd
in ^H
ON r-
d d
00
ro ^H ro oo ro
-H 0 0 O 0
oo NO ro CN ro
•— i p- in CN ~H
— ' O O O O
in oo ro r-- ^J-
^H oo in CN '~H
»— i CO ^3 f^ f*^
in oo NO ro CN
O m TJ- CN — i
-H 0 0 O 0
oo ~H in ON ro
—i oo m CN -H
-H 0 0 0 0
X X 13 B >>
-H 00 «-l "— 1 ^H














.
•^*
2
H
c
«^
^
A
Determined by continuous Grie
6-31

-------
           X
          O
          z
              1.0

              0.8

              0.6


              0.4
              0.2
              0.1
             0.08
             0.06
              0.04
              0.02
                0.01
0.1   0.5 1       5  10            50
   % OF TIME CONCENTRATION IS EXCEEDED
                                                                      90
            Figure 6-12. Frequency distribution of 3-hour-average concentrations
                       of NOX at Los Angeles CAMP Station, December 1, 1963,
                       to December 1, 1964.
a 17 percent increase in NOX emissions during     E.
the winter  months.  NOX values  are  also
affected  by ambient  temperatures  and by
photochemical reactions. Comparison of the
NO and NOX values by season clearly demon-
strates the influence  of the photochemical
reactions that convert  NO  to NO2- The NO
median  value in the summer season (Figure
6-12) represents  50  percent  of the NOX
median, whereas in the winter season the NO
median  represents 80 percent of the NOX
median. This comparison of seasonal NO and
NOX  values demonstrates the increased con-
version of NO to NO2  during the summer due
to the presence of increased sunlight.
                           EFFECTS OF MEASUREMENT
                           SYSTEMS ON DATA
                         In  addition to the continuous monitoring
                       of pollutants at  CAMP  stations,  the  Air
                       Pollution  Control Office  of EPA  obtains
                       data from its Gas-Sampling Network in the
                       National Air  Surveillance Networks (NASN).
                       This  latter  network,  a  cooperative effort
                       with  local  health and  air  pollution agen-
                       cies,  determines  24-hour  concentrations  of
                       certain pollutants on a biweekly schedule at
                       approximately 150 sites.  The yearly  average
                       24-hour NO2 concentrations are shown  in
                       Table 6-10 for the NASN network  for the
                       years 1967, 1968, and 1969.
6-32

-------
Table 6-10. AVERAGE 24-HOUR NO2 CONCENTRATION3 AT NATIONAL AIR SAMPLING
                  NETWORK SITES, 1967 THROUGH 1969
                               (ppm)
State
Alabama


Alaska
Arizona


Arkansas

California












Colorado

Connecticut



Delaware


District of Columbia


Florida



Georgia


Illinois



City
Birmingham
Mobile
Montgomery
Fairbanks
Grand Canyon Park
Phoenix
Tucson
El Dorado
Little Rock
Anaheim
Berkeley
Fresno
Glendale
Humboldt County
Long Beach
Oakland
Sacramento
San Bernardino
San Diego
San Francisco
San Jose
Santa Ana
Denver
Denver CAMP
Bridgeport
Hartford
New Haven
Waterbury
Kent County
Newark
Wilmington
Washington
Washington CAMP
Washington CAMP
Jacksonville
Miami
St. Petersburg
Tampa
Atlanta
Columbus
Savannah
Chicago
Chicago CAMP
Peoria
Rockford
Station
number
3
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
2
1
1
—
1
1
1
1
1
3
2
2
2
2
2
1
1
1
1
2
—
1
1967



0.046



















0.070

0.109
0.094


0.045
0.096


0.132








0.144


1968



0.052

0.1 17b

0.057

0.159




0.2 16b


0.131
0.113
0.110
0.123


0.092b
0.112
0.071
0.107

0.047b
0.075
0.104


0.101

0.072

0.080
0.118



0.155


1969
0.093b
0.025
0.016
0.045
0.0 12b
0.089
0.028b
0.047
0.012
0.148
0.027b
0.042
0.08 lb
0.012
0.1 82b
0.053b
0.0 19b
0.106
0.106
0.095
0.116
0.059
0.032
0.076b
0.1 06b
0.077
0.072
0.037
0.042
0.054b
0.07 lb
0.040
0.069

0.038
0.061
0.020
0.079b
0.096
0.025b
0.031
0.054
0.160
0.050b
0.036
                                                                    6-33

-------
      Table 6-10 (continued). AVERAGE 24-HOUR NO2 CONCENTRATION3 AT NATIONAL AIR
                   SAMPLING NETWORK SITES, 1967 THROUGH 1969
                                    (ppm)
State
Indiana








Iowa

Kansas

Kentucky


Louisiana

Maine
Maryland
Massachusetts


Michigan




Minnesota
Missouri


Montana
Nebraska
New Jersey





New Mexico
New York



City
East Chicago
Evansville
Gary
Hammond
Indianapolis
Monroe
New Albany
New Albany
South Bend
Des Moines
Dubuque
Topeka
Wichita
Covington
Lexington
Louisville
Carville
New Orleans
Acadia National Park
Baltimore
Boston
Springfield
Worcester
Detroit
Flint
Grand Rapids
Lansing
Saginaw
Minneapolis
Kansas City
St. Louis
St. Louis CAMP
Glacier National Park
Omaha
Burlington County
Camden
Glassboro
Jersey City
Newark
Patterson
Albuquerque
Albany
Buffalo
Buffalo
New York City
Station
number






1
2
2
1
1
1
1
1
1
1
1
2
1
1
1
2
1






1
1
2
1
1
2
1
1
1
1
1
1
1
1
3
1
1967
0.081
0.044


0.051
0.028b
0.086b


0.051
0.030


0.088






0.063b


0.100




0.071
0.076
0.094
0.116

0.076
0.097
0.106
0.071
0.131
0.116




0.082b
0.179
1968
0.081
0.032


0.099
0.032

0.08 lb

0.070
0.091

0.057
0.110
0.077
0.115
0.053
0.080b
0.024b
0.095b
0.055b
0.1 15b
0.090b
0.130
0.093
0.093


0.076
0.08 lb
0.1 16b
0.115

0.072
0.091
0.121
0.059
0.066b
0.137
0.129
0.060
0.084b
0.079

0.1 48b
1969
0.086
0.031
0.045
0.054
0.079
0.034

0.072b
0.031
0.033
0.072
0.023
0.064
0.096
0.062b
0.096
0.043
0.061
0.020
0.099
0.040b
0.086
0.087
0.119
0.085
0.090
0.038
0.031
0.076
0.045
0.135
0.108
0.0 12b
0.075
0.062
0.128
0.020
0.065
0.092
0.099
0.048
0.071
0.029

0.1 42b
6-34

-------
Table 6-10 (continued). AVERAGE 24-HOUR NO2 CONCENTRATION3 AT NATIONAL AIR
              SAMPLING NETWORK SITES, 1967 THROUGH 1969
                               (ppm)
State



North Carolina


Ohio








Oklahoma

Oregon
Pennsylvania











Puerto Rico



Rhode Island
South Dakota
Tennessee


Texas




City
Rochester
Syracuse
Utica
Durham
Greensboro
Greensboro
Akron
Canton
Cincinnati
Cincinnati CAMP
Cleveland
Columbus
Dayton
Toledo
Youngstown
Oklahoma City
Tulsa
Portland
Allentown
Clearfield County
Indiana County
Johnstown
Lancaster
Philadelphia
Philadelphia
Pittsburgh
Reading
Warminster
West Chester
York
Bayamon
Bayamon
Guayanilla
Guayanilla
Providence
Custer
Chattanooga
Memphis
Nashville
Austin
Beaumont
Corpus Christi
Dallas
El Paso
Station
number
1
1
1
1
1
2
1
1
1
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
2
2
1
1
2
1
1
1
1
1
2
1
1
2
1
1967









0.087
0.072

0.032

0.088
0.074
0.039
0.027b






0.125
0.079

0.075
0.049


0.040b
0.029

0.051

0.076

0.084




0.043
1968
0.096b
0.093b
0.062
0.102

0.099
0.107
0.103
0.102
0.096
0.119
0.105
0.043
0.093
0.096
0.090
0.056
0.074
0.077



0.101b
0.088b

0.132
0.112
0.066
0.066
0.096
0.047

0.03 lb

0.100
0.025
0.089b
0.090b
0.101



0.100
0.071
1969
0.086
0.074
0.070
0.076b
0.076

0.038
0.092
0.099
0.091
0.099
0.087
0.057
0.096b
0.083b
0.048b
0.033
0.056b
0.090
0.033
0.040
0.080

0.024

0.113
0.081
0.054
0.042
0.076
0.041


0.034
0.087
0.015b
0.042
0.078
0.068b
0.030
0.042b
0.026b
0.074

                                                                     6-35

-------
       Table 6-10 (continued). AVERAGE 24-HOUR N02 CONCENTRATION3 AT NATIONAL AIR
                      SAMPLING NETWORK SITES, 1967 THROUGH 1969
                                          (ppm)
State








Utah
Virginia




Washington

West Virginia
Wisconsin
Wyoming
City
El Paso
Fort Worth
Houston
Lubbock
Pasadena
Pasadena
San Antonio
Tom Green County
Salt Lake City
Norfolk
Page
Shenandoah National
Park
Richmond
Seattle
Tacoma
Charleston
Milwaukee
Casper
Station
number
2
1
1
1
1
2
1
1
1
1
1

1
2





1967




0.035



0.048





0.052

0.076
0.093
0.035
1968

0.083
0.1 16b

0.074b

0.068

0.084
0.087
0.035


0.102
0.096

0.109
0.109
0.032
1969
0.048b
0.071
0.105
0.02 lb

0.050
0.065
0.015
0.060
0.077


0.015
0.088
0.095b
0.023
0.092
0.090
0.025
   Determined by integrated Jacobs-Hochheiser method.
   bThe number and distribution of individual values making up this average do not meet the NASN criteria
   for calculating a yearly average. Nevertheless, it is felt that the average is adequate for drawing the general
   relationships reported in this chapter.
  Since there is an NASN station at each of
the CAMP  stations, the  NO2 values  deter-
mined by the CAMP continuous instrument
with the Griess-Saltzman method of analysis
can  be readily  compared  with  the NASN
24-hour-integrated-value based on the Jacobs-
Hochheiser method.
  Table 6-11 presents the ratios of NASN to
CAMP NO2 values for the years 1967, 1968,
and  1969. The table shows that NASN NC>2
values are three times greater, on the average,
than the corresponding CAMP values. There is
no satisfactory explanation for the discrep-
ancy  at this time. Both methods have been
carefully checked under laboratory conditions
and  are  internally  consistent. In  the field,
however, it is possible that other factors, such
as difference in type and age of sampling-line
particulate filters, may be affecting the meas-
urements.
   Limited laboratory testing suggests that the
specific particulate  filter  employed in the
sampling line  and the age of the filter may
exert a  major  influence on measurement
values.  Frequent  in-field  calibration  and
servicing  of the instruments  is, therefore,
essential.
   The results  are further complicated by the
fact  that  the  CAMP continuous  instruments
cannot be expected to give accurate or precise
results for 1-hour NO2 concentrations below
190  Mg/m3  (0.10 ppm).  CAMP  instruments
6-36

-------
        Table 6-11. RATIO OF AVERAGE
          YEARLY NASNaTOCAMPb
            NO2 MEASUREMENTS
City
Chicago
Cincinnati
Denver
Philadelphia
St. Louis
Washington
Median
Minimum
Maximum
N02 ratio: NASN/CAMP
1967
3.2
3.3
2.0
3.2
5.5
3.0
3.2
2.0
5.5
1968
3.0
3.2
2.7
—
4.3
2.9
3.0
2.7
4.3
1969
3.0
—
2.3
-
4.5
1.5
2.7
1.5
4.5
 aNational Air Sampling Network.
  Continuous Air Monitoring Program.
are designed  to measure NC>2 concentrations
from  0 to  1.9 mg/m3 (0 to  1 ppm), but are
calibrated by the  use of NC>2 mixtures in the
range of 0.2  to  1.9 mg/m3 (0.1 to 1  ppm).
Many of the  days during which comparisons
in Table 6-11 were made, had 24-hour  CAMP
values of 20 to  130  jug/m3 (0.01 to 0.07
ppm), however;  and  since  these 24-hour
values are established by a process involving
summation of the hourly values, a large pro-
portion of the included hourly values were in
the 20- to 60- Mg/m3 (0.01  to 0.03 ppm) con-
centration range.  Under these circumstances
large  errors (100 to  200%) would not  be
uncommon. By contrast, the NASN technique
has been calibrated in the 0- to 190- Mg/m3 (0
to 0.1 ppm) range, as well as at higher concen-
trations, hence should be more  trustworthy.
F.   SUMMARY
  Continuous measurement of the oxides of
nitrogen by various monitoring networks has
made it  possible  to  compile  selected-time-
averages of mean concentrations for selected
time  periods.  From  these  values, various
temporal patterns have  been analyzed  and
 related to photochemical and meteorological
 parameters.
   Both NO and NC>2 concentrations display
 distinct diurnal variations dependent on the
 intensity  of  the solar ultraviolet energy and
 the  amount  of atmospheric mixing. These
 concentrations  also  vary with  the  traffic
 patterns in the sampling area.
   Nitric oxide concentration shows a seasonal
 variation,  with higher values  occurring during
 the  late  fall  and  winter  months. Nitrogen
 dioxide,  however, does  not  display  such
 distinct  seasonal  patterns.  An  analysis of
 limited air-quality data for total NOX concen-
 trations has not clearly indicated  any yearly
 trends.
   The effect of meteorological factors on NO
 and  NO2 concentrations  has   been   well
 documented. Periods of stagnation in urban
 areas have resulted in high peak levels of NOX
 and resultant high-oxidant levels.
   Continuous   measurement has  indicated
 that peak  values of NO above 1.23 mg/m3 (1
 ppm) are  common, but NO2 concentrations
 have rarely been measured at this level. Most
 NO 2 concentrations measured in urban areas
 have been  under 0.94 mg/m3  (0.5 ppm).
   Methods for  measuring atmospheric con-
 centrations of  NOX are  still in need of im-
 provement. In one instance, measurements of
 NO2  taken at  the same  site by different
 methods were found  to differ by a factor of
 3, but  there is no satisfactory explanation for
 the discrepancy at this time.
G.   REFERENCES

1.   The  Automobile and Air Pollution, A Program
    for Progress Part  II. U.S. Department of Com-
    merce. Washington, D.C. December 1967.
2.   Nitrogen Oxides and Air Pollution, California Air
    Resources Board. Sacramento. January  1966.
3.   Brier, G.W. Some Statistical Aspects  of  Long-
    Term Fluctuations in Solar and Atmospheric
    Phenomena. Ann.  N.Y. Acad. Sci. 95: 173-187,
    1961.
4.   Ingels, R.M., et al. Trends in Atmospheric Con-
    centrations of Oxides of Nitrogen, 1957-1961.
    Los  Angeles  County Air Pollution Control
    District. August 1962.
                                                                                     6-37

-------
5.   Green, M.H. New Jersey Air Monitoring Systems
    and Air Quality Data, October  1965 through
    December   1968.  N.J.  Department  of  Health,
    Division of Clean Air and  Water. Trenton. Tech-
    nical Bulletin No. A-69-1. July 1969. 216 p.
6.   Larsen, R.I. A New Mathematical Model of Air
    Pollutant  Concentration  Averaging  Tjme  and
    Frequency. J. Air Pollut.  Contr. Ass. 19: 24-30.
    January 1969.
7.   Air Quality Criteria for Hydrocarbons. National
    Air Pollution  Control Administration.  Wash-
ington, D.C. Publication No. AP-64. March 1970.
p. 5-1 to 5-13.
Comprehensive Technical Report on All Atmos-
pheric Contaminants  Associated  with Photo-
chemical   Air  Pollution.  System Development
Corporation.  Santa Monica, California. Report
No. TM-(L)-4411/002/01. June 1970.
Schuck E. A., J.N. Pitts, Jr., J.K.S. Wan. Rela-
tionships  Between  Certain  Meteorological
Factors and  Photochemical Smog. Air and Water
Pollut. 10: 689-711, 1966.
  6-38

-------
                                   CHAPTER 7.
       EFFECTS OF NITROGEN OXIDES  ON MATERIALS
A.   INTRODUCTION
  Field studies and laboratory research have
successfully linked nitrogen oxides (NOX) to
problems regarding effects on textile dyes and
additives, natural and synthetic textile fibers,
and metals. While not as well known as those
caused by sulfur dioxide or ozone, these ef-
fects,  nevertheless,  are important and  re-
present a significant economic burden.
  Nitric oxide (NO) and its principal reaction
product, nitrogen dioxide  (NC^), participate
in a number of complex atmospheric reac-
tions. Eventually, atmospheric processes bring
the  NOX to  the nitric acid  stage where it
rapidly reacts to produce  various particulate
nitrates. This  chapter discusses only the mate-
rial effects that are  attributed  to NOX and
particulate nitrates, but does not preclude the
possibility that other reactants such as nitric
acid (HNC>3)  can cause damage.  One must
not, however, lose sight  of the role of NOX in
the photochemical formation and buildup of
ozone  and other oxidants, as well as in  the
photo-oxidation of sulfur dioxide in the pres-
ence of reactive hydrocarbons to produce sul-
furic acid aerosols. The photochemical reac-
tion products  are  believed to  cause more
damage than NOX directly; thus, control and
reduction of NOX  should  also be valuable in
reducing damage from these products. AP-63,
Air  Quality  Criteria for  Photochemical
Oxidants,  discusses  the  effects  of  photo-
chemical pollutants on materials.
B.   EFFECTS ON TEXTILE DYES
1.   Acetate Rayon Fading
a.   Historical Background
  Just prior to World War I, a German dye
manufacturer investigated some  unusual cases
of dye fading on stored  wool goods.1  Fading
was  most  noticeable on  the edges of  the
goods, and the primary cause was traced to
NOX in the air. Open electric-arc lamps and
incandescent gas mantles were major sources
of these pollutants by  virtue of a process of
high-temperature  fixation.  The investigators
found that all  the susceptible dyes  contained
free  or  substituted  amino  groups,  and they
suggested  that these might  become either
diazotized or nitrosated by the NOX,
  During  and   following  the  war  years,
increased replacement of older forms of light-
ing  with  electric  filament  lamps  led  to a
general decline of the wool-fading problem. In
the  mid-1920's,  however, researchers  de-
veloped and  introduced a new  fiber,  cellu-
lose acetate rayon. Traditional dyes were  of
little  use  on  this fiber, but chemists soon
developed disperse dyes that were effective.
Some  of  these   dyes,  however,  contained
amino-groups  that  the  Germans  had  pre-
viously found susceptible to NOX fading.

  Shortly thereafter a puzzling type of fading
began to occur increasingly on dyed acetate.
Blue, green,   and  violet  shades of fabrics,
either stored  or in use, faded mysteriously.
Because this fading was frequently observed
in rooms heated by gas heaters, it was called
"gas-fume fading" or "gas fading."
  During the  1930's, acetate fading became a
serious problem. Dye and fiber chemists, ap-
parently unaware of the earlier German work,
devoted  considerable  efforts  toward  a
solution. These  efforts culminated in  1937
when Rowe and  Chamberlain2 systematically
investigated  the   fundamental chemistry  of
dye  degradation  and independently  reached
the same  conclusions as the earlier German
team. Since then,  much  research has been
                                          7-1

-------
carried  out in an effort  to understand gas
fading and  to develop ways of preventing
it.3-5
b.   Gas-Fading Characteristics
  Gas fading  is  marked  by a definite  and
characteristic  reddening of the fabric. This
reddening effect took on  special  importance
in the  mid-1950's  when  investigators  dis-
covered that ozone  also fades many of the
same  dyes  found to  be sensitive  to NOX. In
many  cases   the  characteristic  reddening
enables  scientists to discriminate between gas
fading and the bleached,  washed-out fading
that ozone produces.

c.   Test for Sensitivity
  Recognizing the importance of fading by
NOX, the  American  Association of  Textile
Chemists and  Colorists (AATCC)  developed a
fading-test  procedure  that they tentatively
adopted in  1941 and formally approved in
1957.  It is presently designated  as Standard
Test   Method  23-1962,  Colorfastness  to
Oxides of Nitrogen in the Atmosphere.6 This
method  calls  for suspending test specimens
along with  a control sample in a chamber, and
exposing them to combustion gases from a gas
burner  that has  been adjusted  so that the
chamber temperature does not exceed 60° C.
The  resulting  concentration of NOX is about
38 mg/rrP (20 ppm).7  Specimens remain in
the chamber until the control sample shows a
change  in  shade corresponding  to a  fading
standard. When this happens, the specimens
are said to  have had a treatment equivalent to
6 months of actual exposure to average air, as
represented by three separate locations in
southern New Jersey. If the test specimens do
not  fade  appreciably  after  one  exposure
period, the procedure  is  repeated  (using a
fresh control sample) as many times as neces-
sary to make  an  evaluation. Dyed fabrics are
classified and  rated according to the number
of exposures necessary to produce appreciable
changes in shade.
  Despite an earlier research report that pure
NO   slightly fades some  dyes,4  current
opinion1 >7  favors N©2 as  the fading agent.
This is largely  because  (1)  photochemical
interactions convert  NO to  NO2  and (2)
controlled  environment studies  using con-
centrations of NO2 below 188 mg/m3 (<100
ppm) produce color changes similar to those
found in field exposures. Since the test con-
centrations were so far above normal ambient
levels,  the  validity  of  these  controlled-
environment  studies  is  open to questions.
Cellulose acetate fibers are excellent absorbers
of NO2,^ and this property undoubtedly also
plays  an  important  role  in   dye-fading
mechanisms.
  The disperse  dyes  sensitive to gas fading
have an amino anthraquinone structure. The
reactions  that  cause fading  include nitro-
sation,  diazotization, and oxidation.1  Blue
dyes are especially susceptible; some violets
and  reds  also  fade.  Even  blends  of the
susceptible  colors  with  resistant  dyes are
subject  to  gas  fading.  Nevertheless,  these
sensitive  dyes  are  economically  valuable
because they are moderately  priced and have
desirable dyeing  properties.
d.   Attempts at Protection
  To alleviate the fading problem, researchers
have  developed a  number of chemical  in-
hibitors that can be added to the dyed fabric
and  selectively  react  with  the  nitrogen
oxides.7-8  Inhibitors,  however,  are  only
effective temporarily, since they are eventual-
ly  used up   and  the   dye  again  becomes
vulnerable  to fading. Subsequently,  re-
searchers developed dyes that were resistant
to fading;  but  those  dyes had  poor  light-
fastness. In the  1950's,  chemists  synthesized
new dyes that resisted  the fading effects of
both  NOX   and  light.7'9'10   These  dyes
however, are more expensive  and have poor
dyeing  properties, which result in slower
processing  rates and require greater  care
during  application.  Other  means  to cir-
cumvent the  fading  problem include  using
alternative  fibers,  or  fibers with  colored
pigments incorporated in them.
     Notwithstanding, costly gas-fading  inci-
dents still persist.1l~13 Many incidents of fad-
ing have occurred in  warehouses, and caused
7-2

-------
whole truckloads of materials to be returned
to manufacturers. Fading of acetate linings in
men's suits has been a particularly  vexing
problem.  Some fiber manufacturers do not
allow labelling with their brand names unless
the resulting fabrics  meet established color-
fastness standards against both pollutants and
light.13  Even  though dyers are aware of this
fading problem, many  still use the sensitive
dyes for economic  reasons. Nevertheless, use
of fibers vulnerable to fading is decreasing as
both  fabric producers and retailers  become
more sensitive to the  public's  demand for
products of higher quality.

2.   Cotton and Viscose Rayon Fading
     (Cellulosics)
   While NOX fading of certain disperse dyes
on cellulose acetate fabrics is now well docu-
mented, the textile industry has  paid  little
attention to the fading, caused by pollutants,
on cotton and viscose rayon (collectively  re-
ferred to as cellulosics).  Consumers and  re-
tailers  have   registered  color-change  com-
plaints, but textile people generally attribute
this fading to light.
   An early indication of the  cotton problem
occurred in the mid-1950's. Dye chemists  in-
vestigated  a series  of complaints that some
colored cotton fabrics were fading  during the
drying cycle in home gas-fired dryers.14 The
investigators  traced  the fading to  oxides  of
nitrogen formed during the combustion  of
natural  gas used to heat the dryers.  Fading
occurred only while the textile materials were
moist. Subsequent research about  10 years
later confirmed the  gas-dryer fading problem
and found that NOX levels (expressed as NC>2)
in such dryers  ranged from  1.1 to 3.7 mg/m^
(0.6to2ppm).ls
   Within the last decade,  additional evidence
of this type  of fading has  emerged.  When
exposed to a number  of  different field en-
vironments in the absence of light, some dyed
cellulosic fabrics showed fading after 2 to 3
months.16 >l 7  Fading occurred in certain blue
and green shades, representing dyes from four
major classes: direct, sulfur, vat, and reactive
dyes. Laboratory exposure of these shades to
the  standard  gas-fading  test  pro-
cedure  did  not produce fading.  The test
procedure,  was  designed  to evaluate  dyed
acetate  fabrics,  and has no provisions  for
controlling relative humidity, because humid-
ity  is  not  a  critical  factor  in  acetate
fading.  The  field  exposures, however,  in-
dicated  that relative humidity might be an
important  controlling  factor  in cellulosics
fading.  Subsequent  laboratory exposures to
the gas-fading test  procedure were therefore
conducted  under high-humidity  conditions
(probably  greater  than  50  percent).  The
results were dramatic; the same concentra-
tions  of NOX produced  color changes that
generally  agreed   with  the  field  results.
Presently,  the textile-dye industry  is con-
sidering modifying the gas-fading  test  pro-
cedure to  call for  humidity-controlled con-
ditions.1 8-20

3.   Yellowing of Whites
   A  new  problem that  has received little
publicity, but  is of considerable concern to
the textile industry is the yellow discoloration
of  undyed-white  or  pastel-colored  fab-
rics.15-20 These  fabrics may be woven from
any number of common  fibers, but most of
the discoloration has  occurred  on Nylon,
acetate, and permanent-press materials. Dis-
coloration  has  usually occurred  on  items
stored or on display, including dresses, shirts.
curtains, and lingerie.  Returned items have
represented  major  losses to  some textile
companies.
   Since most discoloration occurs on whites,
dyes  were  ruled out  as a  source of the
problem.  Investigators  turned to various ad-
ditives, including optical brighteners; cationic,
antistatic, and soil-release  finishers; softeners;
and resinous processing agents, which are ap-
plied  to fibers and  fabrics to enhance certain
properties. When tested by  standard labora-
tory  procedures,  including  the  gas-fading
procedure, many of these additives yellowed
on exposure  to NOX. Washing  the fabrics
sometimes removed the yellow discoloration,
but this is impractical for  items that yellow in
warehouses or on display.  Today, the problem

                                        7-3

-------
is best solved by selecting resistant additives.
Such  selection  may  result  in  diseconomies,
but the textile industry and retailers recognize
that some action must  be taken to avoid an
increasing number of complaints about the
discoloration of whites.
C.   EFFECTS ON TEXTILE FIBERS
1.   Cellulosic Fibers
  Researchers have  not studied  the direct
effect of NOX on cellulosic fibers, but Morris
et al.21  did  conduct a field study  of cellu-
losics from which they attributed damage to
ambient levels of NOX in Berkeley, California.
They exposed combed cotton yarn samples at
a 45-degree  angle in cabinets facing south.
Poly vinyl fluoride film, rather than glass, was
used to cover the cabinets in order to allow a
greater  amount  of sunlight  to enter.  One
chamber in each cabinet was  set  up as a
control, in  which entering  air  was filtered
through carbon  canisters. Ambient  air was
circulated through the other chamber. (The
investigators  did not  note  the rate  of air
change.) Some samples were exposed directly
to  daylight,  while   others  were  shaded.
Samples  were  exposed  for  three separate
28-day periods (December through February),
as well as for consecutive combinations  of
these three periods, as is shown in Table 7-1.
During  exposure,  both  air  pollution  and
weather  measurements were  made  (Table
7-2).
         Table 7-1.  BREAKING-STRENGTH OF COTTON YARN SAMPLES EXPOSED TO AIR
                       AND SUNLIGHT IN BERKELEY, CALIFORNIA21
Exposure
Number of
days
28
28
28
56
56
84

Period3
I
II
III
I, II
II, III
I, II, III
Breaking-strength,
%loss

Filtered air
11
15
15
20
19
26

Unfiltered air
15
18
16
24
29
32
Pollution effect,
increase in % loss
breaking-strength
4
3
1
4
10
6
          al - December, II - January, III - February.
             Table 7-2. AIR POLLUTANT LEVELS3 AND WEATHER MEASUREMENTS21
Exposure
period
I
II
HI
Total oxidant,
ppm
0.03
0.03
0.03
;ug/m3
60
60
60
Nitric oxide ,
ppm
0.19
0.23
0.07
Mg/ni3
230
280
90
Nitrogen
dioxide,
ppm
0.08
0.08
0.05
jug/m3
150
150
90
Temp,
°F
50
50
50
Total
sunshine,
%days
100
100
72
Rain,
in.
0.5
6
10
    aFor clock hour with highest average value.
7-4

-------
  At least 20 yarn specimens were evaluated
for each exposure  period. The investigators
assessed  deterioration  by measuring  loss of
breaking strength. Table 7-1  shows the mean-
breaking-strength in percent losses,  for un-
shaded  samples, exposed  to  filtered and un-
filtered  ambient air, for the various exposure
periods.  An  analysis  of variance of  these
mean-loss values revealed that  unfiltered air
deteriorates cotton  yarn to a significantly
greater  extent  than filtered  air. The shaded
samples did not develop a corresponding dif-
ference  between filtered  and unfiltered air.
While textile investigators are well aware of
the pronounced deteriorating action of direct
sunlight, the results of this study serve to em-
phasize  the importance of sunlight in stimu-
lating reactions between some  air pollutants
and materials. It is noteworthy that the small-
est difference between breaking-strength losses
for filtered and unfiltered air occurred during
Period III—the period having the most rain
(which  cleans the air), the  lowest levels of
NOX, and the lowest total hours of sunshine.
While it is impossible  to designate  the ag-
gressive  pollutant in this study, the investiga-
tors  proposed  that  NOX, either directly or
indirectly,  were  instrumental  in  causing
damage.  Sulfur dioxide was not measured,
since levels were known to be low.
  The investigators did not mention the fact
that,  although the filter material, activated
carbon,  absorbs NC>2 effectively,  it  absorbs
NO poorly. Levels  of NO in the filtered air
chamber could, conceivably, have contributed
to the decrease in  breaking  strength. If this
were  the  case, cotton  specimens exposed to
clean  air  would  show  a  smaller  loss in
breaking strength, and the relative pollution
effect would increase.

2.   Synthetic Fibers
  Ambient levels of NOX do not appear to
cause noticeable damage to  synthetic fibers.
Several items, however, are worthy of mention
  In  March  1964,  New York  City news-
papers22 reported an episode of runs in Nylon
stockings worn by women in the vicinity of a
demolition  project. Investigators  identified
the guilty  agent as NO2 gas released during
dynamite blasting operations. Local weather
conditions  at  that  time were unfavorable—a
temperature inversion existed along with wind
stagnation  and  high  humidity.   The  in-
vestigators  proposed that the combination of
these  conditions  in  the  presence  of  ab-
normally high levels of released NO2 and dust
had  produced  nitric acid  aerosols that
damaged the Nylon stockings.
   Travnicek,23 noting the New York City
episode,  suggested that "the corrosive effect
of NOX is rather strong, not only for Nylons,
but  for  practically all  other fiber-forming
polymers, because it combines acid  corrosion
and oxidation."
   Spandex  is  a  synthetic,  elastomeric
material. On exposure to  the standard gas-
fading test procedure,  it develops  a yellow
cast.2 ° This is not a dye-fading problem, since
the oxides  of nitrogen react directly with  the
polymeric material.

D.   EFFECTS ON  NICKEL-BRASS
     ALLOYS
1.   Stress-Corrosion
   In 1959 the Pacific Telephone  and Tele-
graph  Company  noticed  considerable
breakage  of  their  nickel-brass  (65 Cu-23
Zn-Ni) wire springs in some relays located in
Los  Angeles  area   central offices.24'26
Failures  often occurred within 2 years after
installation. These  failures were totally  un-
expected, since the wire springs had been used
with  excellent results for  years, throughout
the nation. Investigators found that  breakage
occurred on wires that were under moderate
stress and a positive electrical potential, and it
was concluded that the failure mechanism was
a  form  of  stress-corrosion-cracking.  Bell
Laboratories  subsequently  showed that high-
nitrate concentrations in airborne dust, which
had   accumulated  on  surfaces adjacent  to
cracked  areas, produced  the  failures. The
nitrate content of  dust from Los Angeles is
from 5 to 15 times greater than from most
eastern and midwestern cities. Furthermore,
Los  Angeles dust  consists of light-colored,
very  fine, claylike materials and considerable
                                        7-5

-------
organic matter in a highly oxidized, polar con-
dition. This combination results in dust that is
more sensitive and reactive to moisture  than
the carbonaceous, oily, siliceous, high-sulfate
content of most eastern  dusts.  Nitrates are
also  more  hygroscopic than sulfate salts. Ad-
ditional tests have shown that  failures  take
place only when  surface  nitrate concentra-
tions are  above  2.4 jug/cm^ and when  the
relative humidity, a very important control-
ling  factor, is above  50 percent.  Other  salts
will cause stress corrosion, but only when the
relative humidity is  greater than 75 percent;
for sulfate salts it must exceed 95 percent.
   Sometime  after this stress-corrosion prob-
lem  was first observed in Los Angeles,  scat-
tered failures were reported, not only  in wire
springs, but in other nickel-brass components
elsewhere  in California,  and in  New York
City, and  Philadelphia. Westinghouse25  has
reported failures in Texas and New Jersey.
   Bell Laboratories26 also reported a totally
different type  of  corrosion problem that has
been observed  in  such widely scattered  loca-
tions as Cincinnati,  Cleveland, Detroit,  Los
Angeles, New  York,  and Philadelphia.  The
nickel bases  of palladium-topped contacts of
crossbar switches  corroded forming  bright-
greenish corrosion  products that gradually
crept up over the palladium cap of the contact,
resulting in electrically open circuits. Investiga-
tors concluded that the "creeping green" corro-
sion was promoted by the presence of anions,
principally nitrates, in accumulated dust.

  Nitrates  in  urban  atmospheres have been
identified as  one of the end products of the
photochemical  reactions  between oxides of
nitrogen and hydrocarbons (see chapter 4). In
Cincinnati  and  Philadelphia, the cities  for
which data are available, the 1965  average
airborne nitrate  particulate  concentrations
were  3.4  and  3.0 jug/m3 respectively.  The
corresponding  average  gaseous  NOX levels
were  124 and  158 Mg/m3  (0.066 and 0.084
ppm).
2.   Protection
  The telephone company took several meas-
ures  to correct the stress-corrosion problem.
7-6
Researchers found  that when zinc is left out
of the nickel-brass alloy, stress-corrosion no
longer occurs. To prevent future problems, a
copper-nickel material was specified  for sub-
sequently manufactured wire-spring relays. In
high-nitrate areas,  local central  offices pro-
tected existing nickel-brass relays by installing
high-efficiency filters in outside-air intakes of
ventilating systems, and  by redesigning their
cooling systems  to  keep relative humidity
below 50 percent.
E.   FUTURE RESEARCH NEEDS
  At  present,  the accepted  procedure for
evaluating the fading characteristics  of dyed
fabrics uses NOX generated from a natural-gas
flame. This produces levels of NOX that, while
of low magnitude,  vary considerably. In addi-
tion, since temperature and relative humidity
are  not  controlled,  experimental results do
not  correspond,  absolutely, to practical envi-
ronmental conditions.
  Although dye  chemists use  this procedure
to assess  the vulnerability of dyes and addi-
tives to NOX, they  have  no sound idea of the
dose-response relationships. Reported  com-
plaints and field  studies have been the princi-
pal means for establishing the actual existence
of atmospheric fading problems.
  To  estimate   costs  resulting  from  NOX
damage, investigators must determine reliable
dose-response  relationships  for  vulnerable
materials.  Such research would also delineate
the  influence of  temperature, relative humid-
ity,  sunlight, and other possible  parameters,
including  interaction with other pollutants.

F.   SUMMARY
  Investigators have found that  oxides of
nitrogen   and their  corresponding reaction
products  cause certain textile dyes to fade,
cause some textile  additives to yellow, deteri-
orate cotton fabrics, and accelerate corrosion
of certain metals. The most serious problems
concern textile dyes and additives.
  Many  years  ago,  complaints  by  retailers
lead to the discovery that NOX  fades a number
of sensitive,  disperse dyes used  on cellulose
acetate  fibers.   The sensitive colors  were

-------
mainly shades of blue. To correct this fading
problem,  dye chemists developed  both dye-
additive  inhibitors  that  provide temporary
protection and  fade-resistant dyes. In  both
cases, especially  in the latter, the end result is
a more expensive fabric.
  During the last decade or so,  investigators
found that certain blue  and green shades of
dyed cellulosics  faded under  high-humidity
conditions. They first encountered this prob-
lem when investigating complaints that cotton
fabrics faded during the  drying cycle in home
gas-fired dryers.  They found that oxides of
nitrogen,  which  are  among the  products of
combustion  of  natural  gas, readily reacted
with  some  dyes under  moist  conditions.
NC>2 levels in the dryers ranged from  1.1 to
3.7  mg/m3 (0.6 to  2 ppm). Later, field ex-
posures,  in the  absence of light, showed fad-
ing occurred  in certain dyed cellulosic fabrics.
Previous laboratory   exposures  to NOX had
not  faded these fabrics.  Subsequent  labora-
tory exposure to NOX, under conditions of
high humidity  did  produce fading in  these
fabrics  and  confirmed   the  importance  of
atmospheric moisture in the fading reactions.
  A more recent problem  concerns the yel-
low  discoloration of  undyed-white  and
pastel-colored fabrics. Chemists  have traced
this yellowing to the action of NOX on certain
additives applied to  fabrics to enhance their
marketing properties. These additives include
optical brighteners,  softeners,  antistatic and
soil-release finishes,   and  resinous  processing
agents. To prevent this yellow discoloration,
textile processors must be critical in selecting
NOx-resistant additives,  which  are  generally
more costly.
  Information concerning  the  damaging ef-
fects of NOX on textile fibers is  meager. One
study has been reviewed here in which investi-
gators exposed cotton fabrics to ambient air
containing above-average  levels of NOX. The
exposed yarns showed increased losses in fiber
strength.
  New York City scientists who investigated
an episode of runs in Nylon stockings traced
the problem  to abnormal levels of NC<2 re-
leased during  dynamite  blasting operations.
The combination of unfavorable weather con-
ditions and NC>2 produced acid aerosols that
damaged the Nylon.

   Nitrogen  oxides  react  with  spandex,  a
synthetic elastomeric fiber, producing a yel-
low discoloration.  This is  an inherent  prop-
erty  of the fiber  rather than  a dye-fading
phenomenon.
   High-particulate-nitrate levels have caused
stress-corrosion failures  of nickel-brass  wire
springs on relays  used  by telephone  com-
panys.  Failures take place  when  surface ni-
trate concentrations  exceed 2.4 jug/cm2 and
the  relative  humidity  is above 50  percent.
This  problem  was first  noticed in  the Los
Angeles area, where airborne nitrate levels are
5 to 15 times greater  than in most eastern and
midwestern cities.
   Another  type of  this  corrosion has  been
associated  with annual  average  particulate
nitrate concentrations of 3.0 and  3.4 jug/m.3
with  corresponding  NOX levels of  124 and
158 jug/m3 (0.066 and 0.084 ppm).
  The  apparent lack of dose-response rela-
tionships for the various materials sensitive to
NOX points the way for future research.  Until
these  relationships  can  be developed, eco-
nomic estimates will be, at the very best, gross
projections.

G.   REFERENCES
 1.  Giles, C. H. The Fading of Colouring Matter. J.
    Ap'pl. Chem., 15: 541-550, December 1965.
 2.  Rowe,  F.  M. and  K. A. J. Chamberlain. The
    "Fading" of Dyeings on Cellulose Acetate Rayon
    - The Action of "Burnt Gas Fumes" (Oxides of
    Nitrogen, etc. in the Atmosphere) on Cellulose
    Acetate  Rayon Dyes. J. Soc. Dyers Colour., 53:
    268-278, July 1937.
 3.  Seibert,  C. A. Atmospheric (Gas) Fading of
    Colored  Cellulose Acetate.  Amer. Dyest.  Rep.,
    29: 363-374, July  1940.
 4.  Greenspan, F. P. and P.  E. Spoerri. A Study of
    Gas  Fading  of Acetate Rayon Dyes.  Amer.
    Dyest. Rep.,30: 645-665, November 1941.
 5.  Ray, T. K., et al. A  Comparison of the  Effect on
    Rayon  Fabrics of Various  Gases Under Con-
    trolled  Conditions.  Amer.  Dyest. Rep., 37:
    391-396, June 1948.
                                                                                        7-7

-------
 6.  1969 Technical Manual of the American Associa-
    tion of  Textile  Chemists and  Colorist, Volume
    45.391-396, June 1948.
 7.  Salvin, V. S., W. D. Paist and W. J. Myles. Ad-
    vances in Theoretical and Practical Studies of Gas
    Fading.  Amer. Dyest. Rep., 14:  297-304. May
    1952.
 8.  Mousalli, F.  S. and W. J. Myles.  Gas Fading of
    Acetate   and Triacetate  Prints.  Amer.  Dyest.
    Rep.,54: 1136-1140. December 1965.
 9.  Salvin, V. S. and R. A. Walker. Relation of Dye
    Structure to Properties of Disperse Dyes - Part I
    Anthraquinone Blues. Amer.  Dyest.  Rep.,  48:
    35-37, July 1959.
10.  Salvin, V. S. and  R.  A.  Walker. Correlation Be-
    tween Colorfastness  and Structure  of  Anthra-
    quinone  Blue Disperse  Dyes.  Text.  Res.,  30:
    381-388, May 1960.
11.  Gas  Fume  Fading. Dyer,  Text. Printer, 128:
    89-90, July 1962.
12.  Fume Fading. National Institute of  Drycleaning,
    Silver Spring, Maryland, FF-141, 2  pgs., June
    1966.
13.  Moreley,  D. J.  Upholstery Fabric Fading  by
    Impurities Present in the  Air. Bull. Furniture Ind.
    Res. Assoc.,  2-3, March 1967.
14.  A Study of the Destructive  Action of Home  Gas
    Fired  Dryers on Certain  Dyestuffs. Amer. Dyest.
    Rep., 45: 471, July 1956.
15.  McLendon,  V.  and  F.  Richardson.  Oxides of
    Nitrogen as  a Factor in  Color  Changes  of Used
    and  Laundered Cotton  Articles. Amer. Dyest.
    Rep.,54: 305-311, April 1965.
16.  Salvin, V.  S. Effect of Atmospheric Contami-
    nants on Light-Fastness Testing. Amer. Dyest.
    Rep.,  47: 450-451, June  1958.
17. Salvin, V. S. Relation of Atmospheric Contami-
    nants  and Ozone to Lightfastness. Amer. Dyest.
    Rep.,5J: 12-20, January 1964.
18. Rabe, P., and R.  Dietrich.  A  Comparison  of
    Methods  for Testing the Fastness to Gas Fading
    of Dyes  on Acetate. Amer.  Dyest. Rep., 45:
    737-740, September 1956.
19. Salvin, V.  S.  Testing  Atmospheric Fading  of
    Dyed  Cotton and Rayon Amer. Dyest. Rep., 58:
    28-29, October 1969.
20. Salvin, V.  S. Effect of Atmospheric  Contami-
    nants  on  Fabrics  - Dyed and Undyed. Test. Qual.
    Contr. Pzp.,14: 56-64, 1969.
21. Morris, M.  A., M. A. Young and T. Molvig. The
    Effect of Air Pollutants on Cotton. Text. Res.,
    34: 563-564, June 1964.
22. City Finds Nylon Culprit: Blasting Gas. The New
    York Times, March 11,1964.
23. Travnicek,  Z. Effects of Air  Pollution on Tex-
    tiles,  Especially Synthetic  Fibers. Int.  Clean Air
    Congr. Proc., I: 224-226, London, Eng., October
    1966.
24. Hermance,  H. W. Combatting the Effects  of
    Smog on Wire-Spring  Relays.  Bell Lab. Rec.,
    48-52 .February 1966,

25. McKinney,  N. and H. W. Hermance. Stress Corro-
    sion  Cracking Rates  of a Nickel-Brass Alloy
    Under Applied Potential, Stress Corrosion Test-
    ing. ASTM STP 452: 274-291,  1967.
26. Hermance, H. W., C. A. Russell, E. J. Bauer, T. F.
    Egan and H. V. Wadlow. Relation of Air-Borne
    Nitrate to Telephone Equipment Damage. For
    publication in  "Environmental  Science  and
    Technology",  1970.
7-8

-------
                                  CHAPTER  8.

                          EFFECTS OF NITROGEN
                        OXIDES ON  VEGETATION
A.   INTRODUCTION
  The primary importance of nitrogen oxides
(NOX) as phytotoxicants was first inferred 20
years ago in California  when photochemical
oxidants were shown to have adverse effects
on  vegetation.  The essential role of NOX in
the production of atmospheric oxidants, in-
cluding peroxyacyl  nitrates (PAN)  and re-
search activities in vegetative effects of these
pollutants are thoroughly reviewed in A P C O
Publication  AP-63, Air Quality Criteria for
Photochemical   Oxidants.  Although  the
importance  of NOX in  photochemical re-
actions is well recognized, the separate effects
on  vegetation attributable to the oxides of
nitrogen per se are difficult  to  assess. Never-
theless, plant  scientists  agree  that  indirect
exposure of sensitive plants to NOX,  through
the photochemical  reactions  producing
oxidants, constitutes the most significant and
widespread  mechanism  for  NOx  phyto-
toxicity. These effects, however,  have yet to
be thoroughly examined.
  Evidence  of damage  to  plants  resulting
from direct  exposures of NOX  in the atmos-
phere is usually confined to the proximity of
specific industrial  sources.  For example,
damage from high ambient levels  of nitrogen
dioxide (NC>2)  has been observed near nitric
acid plants. Direct effects due to nitric oxide
(NO)  and   other  NOX  components or
derivitives have not been delineated  in  the
field.
  In view of the foregoing,  this chapter is
limited to a  discussion of the direct effects of
NOX  on vegetation determined  principally
from the results  of  laboratory  studies. The
bulk of discussion is focused  on the effects of
NO2, but very recent, limited evidence of NO
effects are also treated.
   Evidence  of plant response to phytotoxic
levels of NOX can be divided into three major
categories:  (1)  acute  injury,  (2)  chronic
injury,  and  (3)  physiological  effects. Acute
injury  is manifested by  collapse  of cells  with
subsequent  development  of identifiable
necrotic patterns.  Symptoms  usually result
from short exposures (hours) to varying levels
of NO2, and appear within 2 to 48 hours  after
exposure.
   Chronic  injury is caused by  intermittent
exposure, over long periods, to low concentra-
tions  of gas. It  results in chlorotic or other
pigmented patterns in leaf tissue and may be
accompanied by loss of leaves (leaf-drop).
   Growth  alterations, reduced  yields,  and
changes  in  quality  of  plant products are
among  the  physiological  effects frequently
associated with  pollutant  exposure. At
present atmospheric  NOX exposure has not
been associated with any of these effects. In
the laboratory physiological effects are often
measured in terms of more subtle responses
such as reduced photosynthesis or changes in
rates  of transpiration and  enzymatic  pro-
cesses. Although  most research concerning the
physiological effects of NOX has dealt  with
laboratory  exposures  to NO2, recently the
effect of NO exposure has been measured in
terms  of the  reduction  in apparent photo-
synthesis.  Apparent  photosynthesis  is
measured by the amount of  carbon dioxide
(CO2) absorbed by the plant.
   Descriptions of plant  injury  from NO2 are
found  in two recent publications containing
color plates.4 •! 8
                                           8-1

-------
B.   ACUTE INJURY
  Historically,  vegetation in the vicinity of
nitric acid plants has  been injured by  gases
that escape  during plant operations.1 Leaves
develop brown or black spots, especially on
the margins. The levels of exposure at which
these effects become apparent are unknown,
but  probably  vary  widely.  Experimental
studies  have not revealed  any  direct acute
effect of NO exposure on vegetation.
  Benedict  and Breen2 exposed ten selected
annual and perennial weeds to a mixture of
NO and NO2  for 4 hours.  These plants were
chosen as indicators of plant injury  from air
pollutants. Two types of markings developed:
(1) a discoloration associated with collapse of
cells and  necrosis and  (2)  a general, overall
waxy appearance  of the  leaf.  With broad-
leaved plants  the collapsed, irregular-shaped
neerotic  markings were  intercostal,  but
nearer the margins  of  the  leaf. Middle-aged
leaves usually  developed the most markings;
but  with sunflower,  annual  bluegrass,  and
nettleleaf goosefoot, the middle-aged and old
leaves were equally sensitive.  On  mustard
plants, the oldest leaves were the most sensi-
tive.  In  all species, the young  leaves  were
injured  least. Mustard  was the most sensitive
indicator tested, but all plants were somewhat
resistant.
  Heck3  fumigated  cotton, pinto bean, and
endive plants with 1.9 mg/m^ (1.0 ppm) NO2
for 48 hours and observed slight, but definite,
spotting on the leaves. When  concentrations
of 6.6 mg/m^ (3.5  ppm)  were used for 21
hours, mild necrotic spots appeared on cotton
and  bean, but  endive leaves were completely
necrotic.
  Van Haut and Stratmann4 grouped plants
in relation to their sensitivity or resistance to
injury by NO2 and  NO. When 60 species of
plants were fumigated with a  1:1 mixture of
NO  and NO2, they developed gray-green or
brownish  spots on  the leaves, which  later
became necrotic. Leaf injury  reported was
similar to that caused by SO2, except that the
NOX were 2 to 5 times less toxic. Very young
and  mature leaves   were  less sensitive  than
rapidly  growing ones. The authors made no
mention of duration  or  concentration of ex-
posures.
   MacLean et al.5  exposed  14 ornamental
species  and  6 citrus  varieties to high  con-
centrations of NO2 for short periods and clas-
sified them according  to their susceptibility to
injury. The fumigations  caused marginal and
intercostal necrosis, which was often visible
within 1 hour. Collapsed tissues on the upper
leaf  surface  suggested an  initial  injury  to
palisade  cells. The   necrosis  later  spread
throughout  the leaf.  Citrus  responded with
the wilting of young leaves and  leaf-drop.
Older leaves developed  marginal and inter-
costal necrosis. The  varieties, listed  in  the
order of  decreasing  sensitivity,  were Marsh
seedless grapefruit, pineapple orange, Valencia
orange,  Tangelo orange, Hamlin  orange, and
Temple  orange.  Ornamentals  represented  a
wide  range  of susceptibility.  Tissue collapse
and subsequent leaf-drop occurred rapidly in
azaleas.  Croton plants exposed to 282 mg/m^
(150  ppm)  for 4 hours showed slight inter-
costal necrosis, but Carissa given a like treat-
ment showed no observable effects. Although
defoliation and/or necrosis were  complete in
several  instances,  all species survived and
axillary  buds developed within 2 to  6  weeks.


C.   CHRONIC INJURY
   Evidence of chronic NOX injury is limited.
No evidence is  available that would indicate a
chronic  effect of NO on plants. Leaf-drop and
chlorosis were  observed in navel orange trees
after 35 days of fumigation with 940 Mg/m^
(0.5 ppm) NO2 and leaf-drop  after  8 months
exposure to 470 /zg/ni^ (Q.25  ppm).6 Similar
effects  were  found   on tobacco,  tomato,
spinach, and soybean plants exposed  to the
Los Angeles Basin smog; on  the same plants
raised  in  greenhouses  equipped with car-
bon  filters; and in irradiation chambers sup-
plied  with automobile exhaust from motors
equipped  with afterburners.  Since  leaf le-
sions  typical of ozone injury were not ob-
served,  Glater  postulated that low,  constant
levels of NOX were the cause of injury.7
8-2

-------
D.   PHYSIOLOGICAL EFFECTS

1.   Observed Responses
  All of the evidence of physiological effects
is based on controlled experiments. Heck, et
al.8  recently  fumigated  bean  and  tomato
plants with 12.3 mg/m3 (10.0 ppm) NO and
observed an  immediate  60  to  70  percent
reduction in apparent photosynthesis. An im-
mediate  reduction  in photosynthesis  also
occurred  with  concentrations   above   5.6
mg/m3  (>3  ppm).  The rate of CO2  ab-
sorption returned  to normal as  soon as  the
fumigations  were  discontinued.   No  visible
injury developed after exposure.
  Hill and Bennett9 have compared the effect
of NO with the effect of NO2 on the rate of
apparent photosynthesis of alfalfa and oats. A
threshold concentration of about 0.7 mg/m3
(0.6 ppm) NO or  1.1 mg/m3 (0.6 ppm) NO2
was  required  to  reduce  CO2 assimulation.
Combining  the two gases  gave  an additive
physiological  effect.  Nitric oxide produced a
more rapid  reduction in  apparent photo-
synthesis than NO2, and  recovery  was more
rapid when the fumigation was stopped. Con-
centrations   causing  up  to   50  percent
reduction in apparent photosynthesis after 1
to 2 hours did not induce leaf injury.
  Yields of the navel oranges exposed to 470
Mg/m3  (0.25  ppm) for  8 months  were
reduced.6
  Taylor and Eaton10 exposed  pinto bean
plants to 560 Mg/m3 (0.3 ppm) NO2 for 10 to
19 days and reported a decrease in dry weight
and  an  increase in  unit-weight  chlorophyll
content. Similar studies with tomato plants
exposed to concentrations ranging from  280
to 490 Mg/m3 (0.15 to 0.26 ppm) showed a
decrease in dry  weight and leaf area, a darker
green color,  and a strong tendency  for  the
leaves to curl downward.
2.   Biochemical Mechanisms
  Little  work  has  been  done on   the
biochemical mechanism(s) by which oxides of
nitrogen cause direct injury to plants. When
NO2 reacts   with  water,  it  forms a
stoichiometric mixture  of nitrous and nitric
acids. This reaction probably occurs as the gas
reaches  the  wet  surfaces  of  the  spongy
parenchyma in  the leaves of plants. In vitro
studies by Weill and Caldwell11 on the effect
of 1 molar nitrous acid on beta-amylase from
barley showed  the  enzyme  to be slowly in-
activated.  They  concluded  that easily
oxidized groups such as sulfhydryls could be
affected. DiCarlo and  Redfern12 conducted
more  detailed  studies  with  alpha-amylase
obtained from Bacillus subtilis and concluded
that the nitrite  was reacting  with an essential
amino group in proteins.  It reacts with two
oxidation states of catalase, a hemoprotein
from mammalian tissues,  and with a perox-
idase obtained from horseradish.13  To date,
all  enzymatic studies have  used one molar
nitrate. The low concentrations of nitrate that
would occur in plant tissues  after exposure to
a few parts per million of NO2 in air have yet
to be examined.

E.   FACTORS AFFECTING RESPONSE TO
     NITROGEN DIOXIDE
  Several  factors such  as species  of plant,
stage  of plant  development, plant  environ-
ment  (temperature,  light,   humidity,  soil
moisture,  mineral nutrition), and  variable
susceptibility within species (variety or clone)
influence the degree of injury to vegetation
by air pollutants. Depending upon the kind of
plant and its environment, one factor may be
of much greater importance than another.
  Benedict and  Breen2  reported that moist
soil conditions caused several times  as much
injury  to  their ten test species as did dry
conditions.
  Taylor  and  MacLean14  recognized  the
increased sensitivity of plants to NO2 caused
by low light intensity and reported that 5.6
mg/m3 (3.0 ppm) NO2 in darkness caused as
much injury  as 11.3  mg/m3 (6.0 ppm) in
light. Czech and Nothdurft,15 and Van Haut
and  Stratmann4  also  reported  that night
fumigations may cause more  injury than day-
time exposures. Van Haut and Stratmann,4
reporting on the effect of time of day, found
that  NO2  caused more leaf injury during
                                                                                    8-3

-------
certain periods than at others. Rye plants were
most sensitive from noon to 4 p.m. Oats had a
bimodal  sensitivity,  with  more  injury  oc-
curring  from midnight  to 2 a.m. than from
noon to 2 p.m.
  Of special interest is the study by Dunning,
et al., which showed that leaf injury occurred
on  tobacco after a 4-hour  exposure to  a
mixture of 188 Mg/m3 (0.10 ppm) NO2 + 260
/ug/m3 (0.10 ppm) SO2.17 If these pollutants
act  synergistically, perhaps very  low  con-
centrations of each in  combination with the
other  will cause plant injury.
influence  of  time  and concentration and
permits  the  development  of  a  three-
dimensional injury -response surface:
        C = A0 +  Ail  + A2/T
        C = Concentration, ppm
        2 = Constants (partial regression co-
            efficients) specific for pollutant
            plant species, and environmental
            conditions.
         I = Percent injury
        T = Time, hours
F.   DOSE INJURY RELATIONSHIP BE
     TWEEN NITROGEN DIOXIDE AIR
     POLLUTION AND VEGETATION
  Dosage is a measure of the combination of
duration of exposure and pollutant concentra-
tion; therefore, it is an essential element of air
quality  criteria. The general types  of in-
formation needed and problems involved have
been thoroughly discussed in a preceding air
quality criteria document.16
  Presently, it is possible to  cite only cursory
results  concerning  the  effects  of  variable
dosages of  NOX on the  physiology of plants.
Results  available are summarized in Table 8-1.
  The acute effects  of NO2 have not  been
widely  studied.  Threshold  levels  of injury
reported are tabulated in Table 8-2. The thresh-
old  injury  level is  defined  as  the  exposure
necessary to injure 5 percent of the area of
the leaf (upper surface). Preliminary  time-
concentration curves for several sensitivity
groupings  have  been  developed.  Based on
these curves, Table 8-3  suggests  exposure
durations  and  concentrations  necessary  to
produce injury in sensitive, low-sensitive, and
resistant plants at  the  threshold-injury level.
Table 8-4 gives a complete list of plants that
have been studied, and  places them in  three
NO2-susceptibility  classes on the basis of the
projections in Table 8-3.
  Heck, et al.6 present  a modified model for
reporting injury to vegetation from a group of
common air pollutants, including  nitrogen
dioxide. The model  recognizes the separate
Constants developed for  selected test species
are shown in Table  8-5. These species are
classed in susceptibility groupings based on
Table  8-3 projections. The values  of  the
constants are dependent  upon the prevailing
environmental conditions. The  experimental
design  suggests that these  equations can be
used by  control agencies to  predict injury to
vegetation  under  environmental  conditions
when  vegetation would be most  sensitive
(high humidity and temperatures between 75°
and  90°  F). These  equations are preliminary
attempts to place usable tools in the hands of
control agencies. They should not be applied
to time periods less than 30 minutes or over 8
hours,  or for concentration averages below
0.5 or above 25.0 ppm.
G.   NEED FOR FUTURE RESEARCH
  Plants vary  in sensitivity to NOX. Further
testing in the high concentration range  [28
to 47 mg/m3 (15 to 25 ppm)]  for more than 1
hour will yield little more information, except
to show the exact kind of leaf lesions that will
occur on a given  species. This information
would  only  be worthwhile  for  evaluating
injury  caused by  spills,  rocket  fuels,  acci-
dental release of concentrated gas, or elevated
atmospheric levels in the vicinity of poorly
controlled, industrial use or production of
NO2.
8-4

-------
o
Z
o
O
Z
ti.
o
u
s
b
w
O
1
Q
O
Z
o
ft!

g
oo
p
t!
J5
(2




GO
-t--
o
S3


03^
(L) !?
H ^
0) ro
1 ;s
o c ~£o
ft O £

W '-g
c
8
c
r") ^
^ ft
ft










ca
S
6


ON




O
O
CD
1-4
ca
ft
ft
ca
o
CD
ta
Reduced :
synthesis

>


d




>-q
o










Gramineae
Oat (Avena sativa, L.)
i


ON




o
0
-a
"H
i_i
ca
ft
ca
o
CD
ca
Reduced
synthesis

>


d




^o
d








p-5

Leguminosae
Alfalfa (Medicago sativ



oo



g
o
r-
o
ft .52
il
-4-J t
<-> M
-§2
CD O
CD 0,
Immediat
apparent

>

co_




G
o
o
o
ft .S
0 c
'" >>
-§ 1
CD O
CD O,
Immediat
apparent

>

CN
It
co


O
T— t
CO


^
^
;3
R-
"1

o
s
Solanaceae
Tomato (Lycopersicon



ON ON O ^O O

ft
0
+j i~* ca
.S -a 2s 53 "H
I I 11 !! sf
I i !1 11 Is
1 • 1. s>^ s " ^1>
ft ft Cu£ r^ft CDQ^v:
ca ca > 1-1 ;>ii>
o o b § S 2 ^^ J
^ «* c ^ o ^ S c ^
T3 .22 X3 .2 o ^ -S3 1? o "^ ^
II II Is 1 I ill
•05 -o-SoO ^S "h£
(21 ^Is^ S« a i s
ON r^
> > S £ § 6
— fXl i— i
VO <> ^ ^
* ' * 	 | * i*"S
— „ o -r o 9
ON co
0 0

VO
P ^ ^
^«0 ^jO CO »— * CN O
d do >A d in
d ".
o
2s §
<-> « s
h-J c/) ^
0 K
*IC " ^

i_j fc"0 2 o
^ g ^ 2
'^ *G § >E; §
J. s "i § •§
•§ 1 1 S 1
§ si i . i
« §" ll c g5 «£
H 1! * is I
S3 , > M M— 1 *. >J T— 1 0 Cd
l~ s>a 1 f > 1 g
Cca wSi S 3ca oO
OO J^ £ B5Z i» H
i
                                                                                                                                
-------
i

I
Ci
u


o
H


£

J

S

P




|


1/5

Q


O



U
Qi


H

H
r

oo


c
..
2
"o
•51 «
I;
c
(
<
C.


Reference
Effectsa
>
— tj
3 .-&
H
r?
2 e°
D
_>
5 6
") o-
J ft-
txj
.52
'o
a.
00
co r*i co oo  O co co co co co co oo ^o ooOOoo
° — — 	 °— °--- - -o,,-
\D
^ ^ \O 10
CTs ON ^O *^ ^O "^ W) Q\ ^f 00 ON ON >v^ ^^ Os ^*D 1*«O ON iy^ r^ ^t~
i™^ '""^ ^*O C*~* C"*^ ON r~* ^~^ ^D OO '~~< *~^ ^O »•—( I-H vD i-O <~H i/-\ ^^ ^t- f™^
^ "^ *n -^ r*-) vr>
(N ro •— *
§ o ^
8 >, - § ^ 1
£j f>S ^^ «J J>- iT"1 /^ .5^ ^
^^H M"~l >»^ ^ J (^" MH ^^ ti^ ^^
"-> ^^
0 -g S
T^ O S*
^J ^^ * ^~*4 ^ j^ -. ^J ?3 ••
•a j .g g > ^ -J | 2 1 1 .1 1
1 1 |i ^S I'I 1 ^ 1 ii 1
s- go" -^ § a K S a .&«<£,
! Ill ii ll { ! I 1! I
«S All sil 111 -I I.I i^t |
|| !l|||l||f |! I S| JlS 1
OC Q^^fc-cSj^flj^o c^"o OS«j"3Jon ^
^W hJSCNOO^J^OQ So Offi^c/^F-Hf-H aL
 C
 a


 aT
~^-
 UJ
 Ul


T3
 O

 E

 n

fN
 a
                                                                                                                                         b
                                                                                                                                         3
                                                                                                                                         O

                                                                                                                                         >>
8-6

-------
            Table 8-3. PROJECTED N02 EXPOSURES FOR 5 PERCENT INJURY
                        LEVELS ON SELECTED VEGETATION3

Time,
hr
0.5
J.O
2.0
4.0
8.0
Concentrations producing injury
Sensitive ,b
ppm
6-10
4-8
3-7
2-6
2-5
mg/m^
11.3-18.8
7.5-15.0
5.6-13.2
3.8-11.3
3.8-9.4
Low-sensitive,
ppm
9-17
7-14
6-12
5-10
4-9
mg/m-'
16.9-32.0
13.1-26.3
11.3-22.6
9.4-18.8
7.5-16.9
Resistant,
ppm
>16
>13
>11
>9
>8
mg/irr
^30.1
>24.4
>20.7
>16.9
^15.0
         aReference 6.
         bPlant type.
             Table 8-4. RELATIVE PLANT SENSITIVITY TO N02 INJURY
             Species
        Variety
Reference
Sensitive plants
   Apocynaceae
   Periwinkle (Vinca minor, L.)
   Balsaminaceae
   Sultana (Impatiens sultani, Hook.)

   Begoniaceae
   Begonia (Begonia Rex, Putz.)
   Chenopodiaceae
   Spinach (Spinacia oleracea, L.)
   Compositae
   Chrysanthemum (Chrysanthemum, sp.)
   Lettuce (Lactuca saliva, L.)
  Cruciferae
  Broccoli (Brassica oleracea botrytis, L.)
  Radish (Raphanus sativus, L.)
  Radish (Raphanus sativus, L.)
Bright Eyes


White Impatiens


Thousand Wonders White

Bloomsdale Long Standing

Oregon
Ruby
Early Prize Head
Iceberg
Grand Rapids
Great Lakes
Romaine
Burpee Bibb
Black Seeded Simpson
Butter King
Big Boston
Butter Crunch


Calabrese
Cherry Belle
     8


     8
     8

     8
                                                                                     8-7

-------
               Table 8-4 (continued). RELATIVE PLANT SENSITIVITY TO NO2 INJURY
                      Species
         Variety
Reference
          Gramineae
          Bromegrass (Bromus inermis, L.)
          Oats (Avena Sativa, L.)
          Wheat (Triticum vulgare, V.I 1.)
          Solanaceae
          Pepper (Capsicum fnttescens, L.)

        Low-sensitive plants
          Chenopodiaceae
          Spinach (Spinacia oleracea, L.)
          Compositae
          Dahlia (Dahlia variabilis)

          Cruciferae
          Mustard (Brassica arvensis, Rabenh)

          Cucurbitaceae
          Cucumber (Cucumis sativus, L.)
          Musk melon (Cucumis melo, L.)
          Squash (Cucurbita, sp.)


          Gramineae
          Barley (Hordeum vulgare, L.)

          Leguminosae
          Bean (Phaseolus vulgaris, L.)
          Lima Bean (Phaseolus lunatus, L.)
           Soybean (Glycine max, merr.)
           Malvaceae
           Cotton (Gossypium, sp.)
           Cotton (Gossypium hirsutim, L.)

           Solanaceae
           Tobacco (Nicotiana tabacum, L.)
Sac Smooth
Clintland 64
329-80
Pendek
Wells

Cal Chili 505
Noble
Early Hybrid # 7
American
Heinz Pickling
Black Diamond
London

Early White Bush
Golden Summer Crookneck
Pinto
Henderson
BF
Thaxter
Scott
Bonsei
Kanrich
 Acala 4-42


 White Gold
 Bel-B
 Beltsville W3
 Beltsville C
     10
8-8

-------
         Table 8-4 (continued). RELATIVE PLANT SENSITIVITY TO N02 INJURY
              Species
         Variety
Reference
  Tomato (Lycopersicon esculentum, Mill.)
Resistant plants
  Amaranthaceae
  Pigweed (Amaranthus retroflexus)

  Caryophyllaceae
  Chickweed (Stellaria media, cyrill)

  Chenopodiaceae
  Beet (Beta vulgaris, L.)
  Lamb's - quarters (Chenopodium album, L.)
  Nettle-leaf goosefoot (Chenopodium murale)

  Compositae
  Dandelion (Taraxacum officinale, Weber)
  Sunflower (Helianthus annuus, C.)

  Cucurbitacea
  Cucumber (Cucymis salivus, L.)
  Ericaceae
  Azalea (Rhododendron, sp.)

  Euphorbiaceae
  Croton (Codiaeum, Juss)

  Gramineae
  Annual bluegrass (Poa annua, L.)
  Coin (Zea mays, L.)
  Kentucky blue grass (Poa pratensis)
  Orchard grass (Dactylis glomerata, L.)
  Sorghum (sorghum, sp.)
  Malvaceae
  Cheeseweed  (Malva parviflora, L.)
  Solanaceae
  Tobacco (Nicotiana tabacum, L.)
Roma
A
B
C
D
Pearson Improved
Perfected Detroit
Long Marketeer


Alaska
                                  2

                                  2
                                  2
                                  2

                                  2
                                  2
Pioneer and Golden Cross
Potomac
Martin
Beltsville B
W3
Burley 21
                                                                                         8-9

-------
          Table 8-5. NO2 TIME-CONCENTRATION RESPONSE EQUATIONS (0.5 to 7 hours)3
                                    c = AO + A!I + A2/T
Species
Sensitive plants
Oats (Clintland 64)
Radish (Cherry Belle)
Oats (329-80)
Bromeglass (Sac Smooth)
Begonia
Chrysanthemum
Oats (Pendek)
Wheat (wells)
Sultana
Broccoli
Periwinkle
Low-sensitive plants
Cotton (Paymaster)
Cotton (Acala 4-42)
Tobacco (Bel B)
Tobacco (Bel W3)
Tobacco (White Gold)
Resistant plants
Azalea
Corn (Pioneer)
AO

1.45
2.40
1.75
2.49
2.45
3.16
2.79
2.80
3.93
3.07
2.92

3.97
3.68
3.62
3.65
4.03

3.79
2.60
A]

0.13
0.14
0.15
0.16
0.15
0.16
0.14
0.13
0.13
0.20
0.23

0.23
0.22
0.21
0.18
0.30

1.90
2.70
A2

24
1.02
3.24
1.9
2.99
2.14
2.88
2.94
1.73
2.94
3.02

1.94
3.15
3.98
4.40
3.56

3.39
4.10
R2

0.76
0.83
0.56
0.71
0.63
0.72
0.50
0.52
0.67
0.53
0.55

0.58
0.50
0.38
0.31
0.40

0.33
0.37
D

4.5
4.1
5.7
5.2
6.2
6.1
6.4
6.4
6.3
7.0
7.1

7.1
7.9
8.7
9.0
9.1

16.7
20.3
E

2.3
3.2
2.9
3.5
3.5
3.7
3.8
3.8
4.8
4.5
4.5

5.3
5.2
5.2
5.2
6.0

13.8
16.5
D+E

6.8
7.3
8.6
8.7
9.7
9.8
10.2
10.2
11.1
11.5
11.6

12.4
13.1
13.9
14.2
15.1

30.5
36.8
         aSee text for explanation of equation.

          R2  = coefficients of determination; represent percent variation in model.

          D   = concentrations that will cause 5 percent injury in 1 hour.
          E   = concentrations that will cause 5 percent injury in 8 hours.

          D+E= Basis for placing the plant in a specific susceptibility category; exact cut-off sums
                not yet established.
  To determine in detail how lower levels of
NOX affect vegetation will require much time
and  effort.  Representative species must be
grown under a variety of carefully controlled,
reproducible,  environmental  conditions.  The
plants selected as standards would then have
to be fumigated  with specific concentrations
of NC>2 and/or NO for specified times.
  Evaluation  would require  preliminary in
vitro  work  to determine  the effect of NOX
gases  on  enzymes,  growth regulators, plant
pigments, and other metabolic  systems.  The
same  reactions could then be looked  for in
selected fractions of the treated plants and in
the untreated, but similarly grown, controls.
Enzymatic  effects  should  be  pursued
intensively,  so that  the  way in which the
particular pollutant affects plant  metabolism
could be defined. As biochemical effects were
understood, correlations could be made with
reactions known  to occur in other organisms
such as mammals, insects, and birds.
  The tentative findings by  Heck  involving
synergism show that studies should be broad-
ened  to  explore  more fully the interactions
between NOX  and other air pollutants.
8-10

-------
H.   SUMMARY
   Many kinds of plants develop severe acute
leaf injury (lesions)  when exposed  to con-
centrations of NC>2 greater than  47 mg/m3
(25  ppm)  for  a   1-hour  period.   Though
characteristic for each plant,  the  lesions are
difficult  to distinguish from similar injury
caused by SO2-  The fact that very young and
mature leaves are more resistent to NC>2 than
those  that are expanding rapidly may provide
aid in  identifying NC>2 injury.

   Under controlled  growth  conditions, the
injury-threshold  value  for NC>2  (level that
injures 5 percent of the leaf area) for certain
sensitive plants is 7.5 to 15.0  mg/m3 (4 to 8
ppm)  for  1  hour. Increasing the exposure-
duration  decreases  this threshold concentra-
tion;  with  a rough time-concentration rela-
tionship in which  4.3 to 6.6  mg/m3 (2.3 to
3.5 ppm) NO2 administered to sensitive plant
species for 8 to 21  hours, or 1.9 mg/m3 (1
ppm)  NC>2 for  48  hours,  causes leaf injury.
Continuous fumigation with 940 Mg/m3 (0.5
ppm)  NC>2 for 35 days resulted in leaf drop
and chlorosis in  citrus,  but no actual necrotic
lesions developed.

   The effects of exposure to low levels of
NC>2 for extended periods are  less  evident.
Recently completed studies suggest that 470
Mg/m3 (0.25  ppm) or  less of NC>2  supplied
continuously for 8 months increased leaf-drop
and reduced  yield  of navel oranges.  The
degree of injury  from lower, atmospheric con-
centrations of NO2 remains to  be determined.
Very  mild chronic  effects,   resulting from
fumigation of pinto bean plants with 560
Atg/m3 (0.3 ppm) and  of tomato plants with
280 to 490 jug/m3 (0.15 to 0.26 ppm) NO2,
approximate  the possible  effects  of persist-
ently high ambient concentrations.

  The  limited information on the effective-
ness of  NO in reducing  apparent  photo-
synthesis  indicates that it  would reduce the
growth of plants  if concentrations  in  the
range  of 3.8 to 7.5 mg/m3  (2  to  4  ppm)
persisted continuously.
I.    REFERENCES
 1.  Thomas M. D. Gas Damage to Plants. Ann. Rev.
    Plant Physiol.2:293-322, 1951.
 2.  Benedict,  H.  M.  and W. H. Breen. The Use of
    Weeds  as  a  Means of  Evaluating Vegetation
    Damage Caused by Air Pollution. Proc. 3rd Nat.
    Air Pollut. Symp., Pasadena, Calif., April 18-20,
    1955. 177-190
 3.  Heck,  W. W. Plant Injury Induced by Photo-
    chemical  Reaction  Products  of  Propylene-
    Nitrogen Dioxide Mixtures. J. Air Pollut. Contr.
    Ass. 74:255-261, July 1964.
 4.  Van Haut, H. and H.  Stratmann. Experimental
    Investigations of the Effect of Nitrogen Dioxide
    on Plants. (Experimentelle untersuchungen uber
    die  Winkunge   von  Stuckstoff dioxid  auf
    Pflanzen).  Transactions  of  the  Land Inst.  of
    Pollution  Control  and  Soil Conservation of the
    Land of North Rhine-Westphalia, (Essen). No.
    7:50-70, 1967.
 5.  MacLean, D. C.,  et  al.,  Effects  of  Acute
    Hydrogen  Fluoride  and Nitrogen  Dioxide
    Exposures on Citrus and Ornamental Plants of
    Central   Florida.  Environ.   Sci.  Technol.
    2:444-449, June 1968.
 6.  Thompson, C. R., et al., Effects of Continuous
    Exposure  of  Navel  Oranges  to NO2-  Atmos.
    Environ. In Press, 1970.
 7.  Glater, R. A. B. Smog and Plant Structure in Los
    Angeles County,  March  1970. Reports  Group,
    School  of Engineering  and  Applied  Science.
    University of California at Los Angeles, p. 1-39,
    Report No. 70-17.
 8.  Heck, W.  W., O.  C. Taylor, and D. T.  Tingey,
    Response  of Plants to  Acute Doses of Nitrogen
    Dioxide. BioScience, 21 (In Press, 1971).
 9.  Hill, A.  C.  and  J.  H. Bennett. Inhibition  of
    Apparent  Photosynthesis by  Nitrogen  Oxides.
    Atmos. Environ. (In Press, 1970).
10.  Taylor, O. C. and F. M. Eaton. Suppression  of
    Plant  Growth  by  Nitrogen  Dioxide. Plant
    Physiol. 47:132-135, January 1966.
11.  Weill, E.  C. and M. L.  Caldwell. A Study of the
    Essential Groups  of  B-amylase. J. Amer. Chem.
    Soc. 67:212-214, February 1945.
12.  DiCarlo, F. J. and S.  Redfern. a-Amylase from B.
    subtilis  II. Essential Groups.  Arch.  Biochem.
    75:343-350, 1947.
13.  Nicholls.  P.  and  G.  R.  Schonbaum "Catalases"
    In:  Enzymes 8,  Boyer, P.  D.  et  al  (eds.)
    Academic Press. 1963, 180-181.
14.  Taylor,  O. C., and  D. C. MacLean. Nitrogen
    Oxides and the Peroxyacyl Nitrates. In: "Recog-
    nition of  Air Pollution  Safety to Vegetation: A
    Pictorial  Atlas." J. S. Jacobson and A.  C. Hill,
    eds. Air Pollution Control Association, Pittsburg,
    Pennsylvania, pp. E1-E14, 1970.
                                                                                          8-11

-------
15. Czeck, M. and W. Nothdurft. Investigation of the
    Damage  to Field and  Horticultural  Crops by
    Chlorine,  Nitrous and  Sulfur Dioxide  Gases.
    (Untersuchungen  uber  Schadigungen land-
    wirtschaftlicher und gartnerischer Kulturpflanzen
    durch  Chlor-Nitrose-und  Schwefeldioxydgase).
    Landwirtshaftliche  Forschung. Darmstadt,,  4:
    No. 1,  1-36, 1952.
16. Air Quality Criteria for Photochemical Oxidants.
    Nat. Air  Pollut. Contr.  Admin., Government
    Printing Office, Washington, D. C. AP-63, 6-1 to
    6-23, March, 1970.
17.  Dunning, J. A., D. T. Tingey, and R. A. Reinert.
    Nitrogen Dioxides and Sulfur Dioxide Interact to
    Injure Horticultural  and Agronomic Crops. Hort.
    Sci. 5:333, 1970.
18.  Recognition of Air Pollution Injury to Vegeta-
    tion: A Picturial Atlas. Jacobson, J. S. and A. C.
    Hill 1, Ed. Air Pollution Control Association, Publ.
    1970.
8-12

-------
                                   CHAPTER  9.

     TOXICOLOGICAL EFFECTS OF NITROGEN  OXIDES
A.  INTRODUCTION
  Data from  human  and  animal  studies
indicate that both  nitric oxide  (NO) and
nitrogen dioxide (NO 2) have untoward effects
on health. Animal mortality  studies indicate
that NO2 is about four times more toxic than
NO.1  This chapter represents an attempt  to
evaluate health studies and  to  identify the
probable hazardous doses of  both pollutants.
  The  potential  role  of  nitrogen  oxides
(NOX) in atmospheric reactions that result in
photochemical smog is discussed in chapter 4.
Although  considerable   information  is
available regarding the toxicology of some  of
these  photochemical products, they are not
discussed in this chapter. Many are treated in
AP-63, Air Quality Criteria for Photochemical
Oxidants, and AP-64, Air Quality Criteria for
Hydrocarbons.

B.  NITRIC OXIDE
  No  cases of human poisoning due to NO
have been reported in the scientific literature,
although NO is undoubtedly included in oc-
cupational exposures to NOX. When present
in high concentrations,  NO is  rapidly con-
verted to NO2, a property that makes it diffi-
cult to control NO  concentrations in labora-
tory experiments.
  In animals, extremely high concentrations
of  NO produced   central  nervous  system
paralysis and  convulsions.2 Mice exposed  to
3,075  mg/m3  (2,500 ppm) were narcotized in
6 to 7 minutes and died within 12 minutes,
but when the narcotized animals were re-
turned to  fresh air  after 4 to 6 minutes  of
exposure, they recovered rapidly.
  D,ata  from  limited experiments on guinea
pigs indicate no effect on pulmonary function
from  4-hour  exposures to concentrations of
19.7 to 94.0 mg/m3 (16 to 50 ppm) NO.3

1.   Methemoglobin Increase
   Methemoglobin   (MeHb)  does  not  bind
oxygen; hence, it reduces the oxygen-carrying
capacity  of blood  when it  replaces normal
hemoglobin.  Low   concentrations  of MeHb
ranging from  0.01 to 0.5 gram MeHb per 100
milliliters (g/100 cc)  are  usually present in
normal human blood,4 although the range in
normal  subjects has been reported  to vary
between  0 and 8 percent  of  total  hemo-
globin5  (i.e.,  up to 1.2 g MeHb/100 cc, as-
suming a hemoglobin  value of  15 g/100 cc).
The earliest  clinical  evidence  of  MeHb in
human blood  is  cyanosis, which begins when
the concentration reaches 10 to 15 percent of
the total hemoglobin, a  concentration  not
likely to result from exposure to ambient con-
centrations  of  NOX.  Symptoms  such  as
exertional dyspnea (labored breathing),which
reflects hypoxia  (low oxygen supply) are not
likely to appear until blood levels of MeHb
reach 30 to 45 percent of total hemoglobin.4
   High quantities of MeHb and cyanosis were
detected in the blood of animals that died as a
result  of  poisoning  from exposures to 1,500
mg/m3 (1,200 ppm) NO.6'7  Evidence con-
cerning MeHb formation following exposure
to lower  concentrations of NO is conflicting.
   The significance  of low  blood  levels  of
MeHb is  not well understood. In chapter 10,
the use of blood levels of MeHb as an index of
exposure  to ambient NOX is discussed,8  but
few epidemiological investigators actually use
this parameter in their experimental  design.
  In vitro, NO combines with hemoglobin to
form   NO-hemoglobin  (NO-Hb)  and  with
                                          9-1

-------
MeHb  to  form  NO-methemoglobin  (NO-
MeHb), both of which  have biological half-
lives of about 2 days. A preliminary examina-
tion of the blood of rats exposed for 1 and 9
days to 12.3 mg/m3 (10 ppm) NO in air by
electron-spin resonance revealed neither  NO-
Hb nor NO-MeHb.9
2.   Enzyme Inhibition
  A concentration of about 24.6 mg/m3 (20
ppm) NO inhibited hydrogenase activity of
the bacteria Proteus vulgaris, both in vivo and
in vitro.1 ° This activity could be restored by
the addition of ^28204, which reduces NO
to nitrous oxide (N2O), except at concentra-
tions of 12.3 x 103 mg/m3 (10,000 ppm) NO
and  above  where inhibition was irreversible.
  Although ambient levels of NO  are not
inherently toxic, toxicological potential lies in
its relationship to  NO2 which is discussed in
AP-63, Air Quality Criteria for Photochemical
Oxidants.11
C.   NITROGEN DIOXIDE
1.   Effects in Animals
a.   Mortality
  Many  early  studies  produced  negative
evidence  of NO2  toxicity  in animals.  This
may have been due to the presence of large
amounts of NO in the experimental mixtures,
for recent studies have produced more reliable
and  positive evidence  of toxicity. Several
factors affecting the mortality due to NO2 are
discussed in the following section.
(1)  Length of exposure  and concentration
(ct).  Short exposures of rats to high concen-
trations are  more  toxic than  equivalent
exposures to low concentrations for longer
times. This is illustrated in  Table 9-1.12
   Most investigations involving inhalation of
high concentrations of NO2 have implicated
pulmonary  edema  as the  major  cause  of
death.  When 112 animals of various species
were exposed to 56.4 to  1,880 mg/m3 (30 to
1,000 ppm) NO2, 84 animals died,  74 from
pulmonary edema,  5 from asphyxia, and  5
from  pneumonitis  (inflammation   of  lung
tissues) (Table 9-2). Exposure to 56.4 mg/m3
(30  ppm)  NO2  for  3 hours  produced no
apparent  immediate or delayed  harmful
effects in guinea pigs. The harmful effects of
2- to 3-hour exposures of 103.4  mg/m3 (55
ppm) NO2 were questionable in experiments
with rats and  mice. None were found at the
                    Table 9-1. COMPARISON OF LETHAL LEVELS OF ACUTE
                            EXPOSURE OF MALE RATS TO NO2'2
Number
of
experiments
3
2
6
10
10
7
Exposure
time,
min
2
5
15
30
60
240
LCSO,a'b


ppm
l,445d
833d
420 (362-487)
174 (154-197)
168 (153-185)
88 (79-99)

mg/m^
2,715
1,566
790(680-916)
325 (290-370)
315 (290-350)
165 (140-185)
LCtso,b'c

ppm x min
2,890
4,165
6,300 (5,430-7,305)
5,220 (4,620-5,910)
10,080 (9,180-11,100)
21,120(18,960-23,760)
         aLC5Q represents the concentration lethal to 50 percent of the animals.
         "95 percent confidence limits in parenthesis.
         cLCt5Q represents the exposure (concentration x time) lethal to 50 percent of the animals.
         dNo confidence limits.
9-2

-------
    Table 9-2. TIME NECESSARY TO PRODUCE
      DEATH IN ANIMALS3 EXPOSED TO
      HIGH CONCENTRATIONS OF NO2'3
Concentration,

ppm
30
100
150
400
600
800
1,000

mg/rrp
56.4
188.0
282.0
752.0
1,128.0
1,524.0
•1,880.0

Time until death,
min
	
318
90
58
32
29
19


% deaths
0
74
70
92
93
100
100
 aAnimals included the following species: cats, guinea
  pigs, mice, rats, and rabbits. Death occurred in 84 of
  112 animals.
 same concentration  with rabbits, cats, and
 guinea pigs.1 3
   It  is rare  that any  species  can withstand
 exposure  to  NC>2 concentrations of 188  to
 1,880  mg/m3  (100  to  1,000 ppm).  In
 experiments with these concentrations,  death
 occurred  after 10  minutes to 21 hours  of
 exposure,  depending on the animal.12 Four-
 hour exposure to  99.6 mg/m3 (53 ppm) NC>2
 was fatal to rats and  122.2 to 141 mg/m3 (65
 to 75 ppm) NC>2, to dogs.
 (2)  Temperature. An  ambient temperature
 increase of about  11° C (20° F) increases the
 toxicity  of  NC>2  for  rats  by  about  25
 percent.
14
(3)  Presence of other irritants. The survival
time for rabbits was increased from 60 to 140
minutes when nonlethal amounts of sulfur
dioxide (862) were added to the NC>2 dose
that caused 50 percent mortality.14

b.   Respiratory tract effects
(1)  Changes in pulmonary function.  Short-
term   (less  than  4  hours)  exposure  to
NC>2   produced  reversible  changes  in
pulmonary  function.3'15  In general,  both
short-term  and continuous exposures caused
respiratory rates to increase and tidal volumes
(the volume of air inhaled in an average single
breath) to  decrease during exposures to con-
centrations ranging from  1.5 (0.8 ppm) to 94
mg/m3 (50 ppm).  The  degree of response
varied  with the  exposure and  the animal
(Table 9-3 and Figure 9-1).
  Other  pulmonary  functions  have  been
examined,  but show no evidence  of  N(>2
effects. Guinea pigs exposed to 9.4 mg/m3 (5
ppm) NC>2 for either 4 or 7.5 hours a day, 5
days a week,  for periods up  to 5.5  months
had no  changes  in expiratory flow resist-
ance.16 Four rabbits exposed  to 47.0 mg/m3
(25 ppm) NC>2 continuously  for  18  months
had a transient increase in the rate of oxygen
consumption.  The  rate reverted  to  normal
within  48 to 72 hours after exposure. In 16
rabbits exposed to 1.9 and 9.4 mg/m3 (1 and
5 ppm) NC>2 for the same length of time, no
change in oxygen consumption occurred.17
  Beagles exposed to 0.9 to 1.9 mg/m3 (0.5
to 1.0 ppm) NO2 plus 245  Mg/m3 (0.2 ppm)
NO, and to 2.8 to 3.8  mg/m3 (1.5 to 2.0
ppm) NO2 plus 245 jug/m3 (0.2 ppm) NO for
16 hours  a day for 18 months, showed no
changes  in  body  weight,  or  in  carbon
monoxide  (CO)-diffusing  capacity,  com-
pliance, or total expiratory resistance of the
lungs when compared with a control group of
20 breathing filtered air.18
(2)  Chemical  effects.
  (a)  Structural proteins—NO2 can alter the
configuration of the lung tissue structural pro-
teins, collagen and elastin.19 One animal from
each of five sets of four litter-mate female
rabbits served  as  a control; one breathed 1.9
mg/m3 (1 ppm) NO2 for 1 hour;  and the re-
maining two were exposed to 9.4 mg/m3 (5
ppm)  NO2 for 1  hour. All the animals were
sacrificed immediately after exposure, except
one in the last group, which was sacrificed 24
hours after exposure. The lungs were excised
and converted to a lipid-free powder  from
which collagen  and  elastin were isolated by a
combination  of  solvent-extraction  and
enzymatic-hydrolysis  methods.  Differential
ultraviolet  spectrophotometry  indicated that
                                                                                    9-3

-------
   O
   z
   Q

   -1





   W
   2
   P
   I
   Q.

   z
   U

   O
   X
   u
   ON


   
o
C

t E .S
£ 3 E
H .5
E
Gl d>
B E
G 5
S >
eu
3 6
H 1
0
.53 "PJJ
OH *"'
^>
as
0
n «
8,1
w


cT e
•2 1
d "™
(-<
8 co
§ ^

e




g
'e



m in in

l-H
fN °
•Si- 00 O ' £
* * 5 0.3
^ 
C ca «
§ c c


d d d



•
S


i ,
\ +



+i + +

3
O
3
C
'5 O °
§ e s
U (N —

CN G ON in O
rt T"~<


00 r^ ON ^" OO
m c ^ ON 06
CS ^H «

^~N W5

_O O K>
S | | |
~~^ W 00
     £ Q S Z
                                            + , O C
9-4

-------
             z
             o
             u
             o
             o.
             X
             LU
120

100

 80

 60

 40
             u
             OL
             IU
             0.
                                           TIDAL VOLUME
  "-RETURN TO AIR
J	I
                                                          LRETURN TO AIR
                                      2345

                                 N02, ppm | EXPOSURE, hr | SUBJECTS
                                    13
                                    6.5
                                    5.2
                                           10
                                           10
                                           10
180

160

140

120

100

 80
                                                          RETURN TO AIR
                                            TIME, hours
            Figure 9-1.  Effect  of N02 on tidal volume and respiratory rate of
                        guinea pigs.10
the molecular structures of both collagen and
elastin were altered in both exposure groups.
Similarity between the spectra of collagen and
elastin from  the animal sacrificed 24 hours
after  exposure  and  those from  the  control
animal  indicated  that  the  changes  were
reversible under the conditions  of  the  ex-
periment. The authors suggested that meta-
bolic activity of the lung tissue was reduced in
the highly acidic  environment produced by
exposure  to  NC>2  and speculated  that re-
peated  exposure  to  NC>2  with  concurrent
repeated denaturation of  collagen and elastin
may be a factor in the etiology of pulmonary
emphysema.
   Rabbits exposed  on  a  schedule  of 470
Mg/m3 (0.25  ppm) NO2, 4 hours a day for 6
days; and sacrificed  immediately,  developed
                               structural changes  in  the lung  collagen,  as
                               determined  by  electron  microscopy.  These
                               changes were not reversible and were apparent
                               in  animals sacrificed 7 days  after the final
                               exposure.2 °
                                 (b)  Lipid  peroxidation—Short-term low-
                               level exposures  [4  hours at  1.9 mg/m3  (1
                               ppm)]  produced  evident  peroxidation  in
                               extracts of lipids from rat lungs.21 Peroxida-
                               tion was maximum  24  hours  after exposure,
                               and lasted  for at least 24 hours more. In six
                               daily 4-hour exposures to  1.9 mg/m3 (Ippm),
                               the lipid peroxidation increased, possibly as a
                               cumulative  effect.  The  significance  of this
                               peroxidation is not yet established; however,
                               rats  fed a vitamin-E-deficient diet and  ex-
                               posed  to  NC>2  had more peroxidation  in
                               both  surfactant  and tissue lipids than  did
                                                                                       9-5

-------
those  on  vitamin-E-supplemented diets.22
Vitamin E is an antioxidant.
(3)   General  Pathological  Effects.  Patho-
logically, all  animal lungs demonstrate  the
same  general pattern of response to NC>2, ir-
respective of the duration of exposure,  but
the severity  tends  to  increase as the  con-
centration  increases.  An inflammatory re-
action characterized by macrophage infiltra-
tion and epithelial degeneration occurs during
the 24-hour period immediately  following
exposure and can  develop  into  pulmonary
edema at  sufficiently  high  concentrations.
Subsequent  regeneration   follows.  Such
changes were described in rabbits exposed to
188 mg/m3 (100 ppm) and in rats and guinea
pigs exposed  to  141 to 150 mg/m3 (75 to 80
ppm)  NC>2 for 2 hours.2 3 In experiments in
which rats and guinea  pigs were exposed to
28.2  to  37.6 mg/m3  (15 to 20  ppm)  and
rabbits to 47.0 mg/m3 (25 ppm) NC>2 for 2
hours  a  day  for 5 successive days, initial
edema and inflammation were not as severe as
they were in the single-dose experiments; but
chronic  peribronchial  and   perivascular in-
flammation followed later. On the fourth day
after  epithelium was destroyed by a solitary
high dose, lung tissue did repair itself in both
rabbits and guinea pigs. Regeneration became
intense  and  appeared  to be complete  in  2
weeks; after a series  of high intermittent doses
repair  was  slower.  Animals exposed inter-
mittently  during a period  of 21  months
required  up to 3 weeks  after the last exposure
to recover.2 3>24
(4) Cellular Changes.
   (a) Structural  damage—short-term  ex-
posure—In  short-term exposures to low con-
centrations (0.5  ppm for 4 hours or 1 ppm, 1
hour) rats sustained  reversible   lung-tissue
change.25  In  tissues from animals sacrificed
immediately  after  exposure,  the  mast  cells
were  ruptured  and  disoriented, and showed
loss of cytoplasmic granules. This occurred
primarily in the pleura, bronchi, and  sur-
rounding tissues, but most  markedly in the
mediastinum. This response seemed reversible,
since  animals sacrificed 24 to 27 hours  after
exposure appeared  to  have  only a few rup-
9-6
tured mast cells. The investigators considered
that the release of granular material from the
lung mast cells in response to NC>2 inhalation
signified  the  potential onset of an acute in-
flammatory reaction.
  After 4 hours exposure to 32.0 mg/m3 (17
ppm) NC>2,  the  appearance of bits of free
fibrin in both  alveolar  spaces and tissue in-
dicated transient  leakage of plasma from rat
alveolar capillaries. With continued exposure
in the same  experiment, researchers  noted
that focal injury  to type I alveolar epithelial
cells was exaggerated within 48 hours,  as bits
of  amorphous  cellular  debris  appeared  in
ductal alveoli. The facts that fibrin and debris
tended to  disappear  and that  injured cells
healed as  exposure  continued  suggests  the
development  of tolerance.26

  Primary  lesions appeared  in lung alveoli of
squirrel monkeys exposed for 2 hours to 18.8
to 94.0 mg/m3 (10 to 50 ppm) NO2- Progres-
sive  alveolar  expansion occurred with in-
creasing  concentrations of  NO2- At 18.8
mg/m3 (10 ppm) NC>2, many septal  breaks
appeared and the alveoli expanded markedly.
In some areas, large air  vesicles with extreme-
ly thin septal walls were seen. Other  tissues
appeared to  be normal. At  28.2 mg/m3  (15
ppm) NO2 alveolar tissue was expanded with
minimal  wall thinning and patchy interstitial
infiltration  with  lymphocytes.  The  bron-
chioles  were  normal. At  65.8 mg/m3  (35
ppm) NO2 areas of lung were collapsed and
alveolar  septa  became  very  basophilic.  In
other areas,  the  alveoli were expanded and
haa  thin  septal  walls. The bronchi were
moderately inflamed; some showed epithelial
proliferation.15 Exposure to 94.0 mg/m3 (50
ppm)  NO2  resulted  in extreme vesicular
dilatation or total collapse of alveoli, along
with  extensive  edema  and  lymphocyte in-
filtration.  The  bronchi showed  epithelial
surface erosion and absence of cilia.

  (b) Orientation  changes—continuous  ex-
posure—After 3 days of continuous exposure
to  3.8  mg/m3 (2 ppm) NO2,  the  terminal
bronchiolar epithelium  of rats changed from
an  active,  inhomogeneous, lining layer, to  a

-------
uniform layer  of  enlarged cells. This subtle
biochemical effect on epithelial cell metabo-
lism was further documented by the demon-
stration  of abnormal ciliogenesis.  Although
ciliary basal bodies developed normally, they
failed to orient appropriately at the apical sur-
faces of the cells. They either formed no cilia
or  directed  them  intracytoplasmically  into
vacuoles.  Intracytoplasmic,   crystalloid  in-
clusion bodies also developed in time.2 6'2 8
   (c)  Hy perplasia—In  another case,  con-
tinuous exposure  to 32.0 mg/m3 (17 ppm)
NO2 precipitated  a wave of accelerated cel-
lular replication (hyperplasia)  of  terminal-
bronchiolar and type-II  epithelial  cells  and
macrophages in rat lungs. This response was
measured  both by  nuclear autoradiography
with tritiated thymidine incorporation and by
mitotic-figure  counts.29 The activity peaked
at  24  hours and returned to  baseline values
within  a  week,  despite   the  continued
exposure and without apparent vital injury to
the epithelial cells involved.
   In hamsters continuously exposed to higher
NC>2 concentrations  [84.6 to  103.4 mg/m3
(45  to 55  ppm)] for 2.5 months, researchers
noted  transitory hyperplasia  (increased cell
production)  of the  respiratory  bronchiolar
epithelium.  The  lung  volumes  increased
during exposure and  reverted  toward normal
after  a  2-week  recovery.30   In  hamsters
exposed  to 188 mg/m3 (100 ppm) NC>2 for 6
hours, tritiated thymidine uptake indicated an
intense  epithelial proliferation in  the major
bronchi, that   was  maximal  24 hours after
exposure and  normal again within 4 days. In
the periphery,  the response was delayed and
less intense.31
   (d)  Emphysematous  lesions—With
exposures of 3 months or longer, 0.9 to 47.0
mg/m3 (0.5 to 25  ppm) NC>2 caused changes
in animal lung  tissue  similar to those seen in
human  emphysema.  Pre-emphysematous
lesions indicative of the development of early
focal emphysema were  observed in the lungs
of mice exposed to 940 Mg/m3 (0.5 ppm)
N(>2 for 6, 18, and 24 hours a day for periods
from 3 to 12 months.32 The data are given in
Table 9-4. Alveoli, consistently increased in
size  by cell-distension rather than by septal
breakage.  The investigators  also  noted  in-
flammation of the bronchiolar epithelium and
reduction  in distal airway size.
   Continuous exposure of young rats to 18.8
to 47.0 mg/m3 (10 to  25  ppm)  NO2 pro-
duced  voluminous, heavy, dry, air-retaining
lungs in expanded, kyphotic thoraces—a con-
dition  grossly resembling emphysema  (see
Table  9-5).33-36   Such  animals   died  of
respiratory complications after several months
of exposure.  Pathological findings included
narrow, occasionally  fibrotic and  occluded
terminal bronchioles with no cilia; distended,
ruptured peripheral alveoli; and alterations in
the staining  characteristics of both  collagen
and  elastic tissue, particularly in tissue from
the  alveolar  ducts.37  Terminal  bronchiolar,
ductal, and adjacent alveolar epithelium was
hypertrophic (enlarged).
   When such lungs were  allowed to recover,
the hypertrophic  epithelium receded  and cilia
reappeared.   With the  recession,  the  lung
weights  returned  to  normal.  Additional
pulmonary connective tissue usually  appears
in rats as a part  of  aging; however, the
previously exposed, healing rats  developed
excessive  lung weight, compared to  controls
of the same  age. This  suggested an  "anam-
nestic"  connective-tissue  response   in  old
age.38  Abnormally huge collagen fibrils and
overtly thick basement laminae  were revealed
by electron  microscopy  of  the lungs  from
such rats,  including those exposed to  as little
as 3.8 mg/m3 (2 ppm) NO2.39
   Early, emphysema-like lesions could also be
produced  in  larger animals by  exposure to
NO 2- Twelve pure-bred beagle  dogs,  divided
into  two  equal groups, were exposed con-
tinuously  for  6 months to either filtered air,
or filtered air and approximately 47.0 mg/m3
(25  ppm) NO2-  Autopsies on the NC>2-
exposed dogs revealed bullous emphysema in
the  lungs  of one  animal and  increased
firmness with scattered, small bullae  in each
of the remaining five  animals.  A diffuse in-
crease in collagen  was noted in the lungs of all
                                                                                      9-7

-------
ppm) NO2
(*1

I
3
o\
g
>H
5!


ii
X
«l
ss
5 o
« a)
%%

Is
5 <
2>
zg
i—i O
S3
o
z
<
9-4. PATHOLOGICAL
a
exposure, h
        (N
        00
        SO
     •$ 3
       S
           a •§

             - i
            -
          S M 1 1
          ^a^^g-
          '53.«T3"'
          • = •
                 <*-'„•, c o 5

            i
           - §
          H
            1
           ll
                1 S
                H

                H
§ x
ill
                ij.fl
                ! js 1 '§• 1
          c c c o
          I a -S 1

             !
          ft|
          •S
                Mil

                           ii
                         lii
                         CCL,>


            •



                               I
                   8.

                            S S
                                22

9-8

-------
I5
*o
-*-*
£














en
(U
00
c3
JS
o

r^

"Ji^
£
oo
3
"3 £
•s 2
oo o
c ft
£ x
J <"
c

ca
G
o
0
u
ro
g

g
g
a.
G.

nchiolar epithelium; colum-
0
x
C3
en
^
^
CL>
T3
1
>^
bH
•S
o
In
>.
ca
T3


S
o
^+-1
Q>

e

l-H
tS
'S
^

ca
r-
CO

p
*

in
,
i-^ (^
o O-
oo o
it
ttJ ***
fH
O T3
^ m
§ ^-
•g !
tn >,
2 -=!
11
Tl C
^^ n
3|
^3 e*5
8 &
S "°
0, [^
>> IO
1 <

o
_S
^o?>
cu
^

,^
XI
_-, 3
0 O
+3 "^
C V-,
o ^
o ^*
O
1
i








uamated cells in alveolar
I
m

u
connective tissue stroma, some free macrophag'









^^
o
"c
8










o
o
(U
00
1
o
t/i
OS
o>
13
"G.
on
"o
(D
13
T3
C

c/5
spaces. Considerable distension of alveolar duel
sidered emphysematous.























long periods of exposure,
two had focal adenomatosis
oo • -
II
•S3
3 XI
en ^
"2 'ft
"*~J flj
Cy
G-^2
SI-TJ
pi^
&B -+-*
_j_ UH
^ Cd
11
O C
^ "- '
^-H ^3
o <^>
O f i i
-4—1
1 g
V3 O
0) ,/,
X *r^
o -^
T3 g
O 0
*' t4_

^

•^3 'S^
o .«
wa q>
O ^
o ^

O
^-^ *«^
3 o
O Q>
*5 *


00
en
in
CN
"

ids of mucous material in
Iveoli contained bits of
a »
^ ^

05 §
M .5
D VH
| ft
15 X
^0 W


 *rt
X ."-i
n C
Cfl Q
"o a
Grossly, like controls. Microscopically, bronchi
with few remaining cilia and increased size (are


.,_,
g "o

TO C3

CJ

*^3 r^
CO ^^

o "*
.S "

13 g
5-1
13
Z 3

00
m

c^


cytoplasmic blebbing. Rod-
c
CO
•^
.^
o
suggested dormancy, lacking most exfoliative a>
like cytoplasmic inclusion bodies seen.























ial evidence of bronchiolar
3 (2 ppm).
•2-1
|1
O .
1-1 ^
,0 .£
Microscopically, essentially unchanged except i
epithelial changes similar to those obtained wit!


g
o
<4-<
O
o
e

}_i ^_^
£ 2
•3 "S


13 g
|l
3 -s,
Z 3

m
1 '
00
o

9-9

-------
six  animals, but no  such changes  were ob-
served  in  the  controls.40  Further  studies
suggested that adding 1  Mg/m3 ferric oxide
dust to  air containing 48.9 mg/m3  (26 ppm)
NC>2 may protect  dogs against the develop-
ment of pathological changes. This was the
case in a subsequent 6-month experiment.41

(5)   Combined Effects of NO2 With Tobacco
Smoke.  Exposure either for 2 hours  to 28.2
mg/m3  (15  ppm)  NC>2 or for 1  hour to 3
percent  (v/v) tobacco smoke had only a slight
effect on the surface structure of hamster
lungs. When hamsters  were  exposed to the
same concentrations  for the same times, but
sequentially  (NC>2  followed by  tobacco
smoke), the combination  of exposures pro-
duced marked alterations in surface morphol-
ogy. The structure  of mucus-secreting cells
was not typical, and a marked loss of cilia was
noted 2  days after exposure.  These  changes
were not reversible.  Seven days after expo-
sure, the  surface  structure  of  the  main
bronchi and secondary airways appeared to be
even more disrupted; small  patches  of cilia
and deep  holes appeared in  the  bronchial
mucosa. The authors concluded that NC>2 and
tobacco  smoke act  synergistically  on the
bronchial epithelium, producing pathological
changes much greater  than  those caused by
either pollutant alone.42

c.   Systemic Effects

(1)   Tissue Changes. Kidney, liver, and heart
tissues of squirrel monkeys given 2-hour expo-
sures to NO2 were damaged in proportion to
the  dosage. At  28.2 mg/m3 (15 ppm) some
tubular erosion  appeared in the kidneys; liver
cells ballooned and developed clear cytoplasm
with displaced nuclei and congested intersti-
tial  spaces.  At  65.8 mg/m3 (35 ppm) renal
glomerular tufts were swollen and heart tissue
showed  areas of interstitial fibrosis.  At  94
mg/m3 (50  ppm) the hearts developed inter-
stitial edema  and  lymphocytes  infiltrated
kidney and liver tissue. Centrolobular necrosis
also appeared in the liver.l s
(2)  Weight Loss.  Studies on the effects of
long-term  exposure to NC>2  have  produced
conflicting results with respect to weight loss.
No significant reduction in the rate of weight
gain  was noticed in rabbits, guinea pigs, rats,
or hamsters exposed to 1.9, 9.4, or 47 mg/m3
(1, 5, and 25 ppm) NC>2 and  dogs to 1.9 and
9.4 mg/m3 (1 and  5 ppm) NC>2, for 6 hours a
day  for 18  months.17 Furthermore, no sig-
nificant differences in weight gains appeared
in mice exposed  to 940  Mg/rn3  (0.5  ppm)
NC>2 for 6,  18, or  24 hours  a day, 5 days a
week, for up to 12 months.43  Combined NOX
exposures of 940 to 1,880 Mg/m3 (0.5 to 1.0
ppm) NC>2 plus 245 Mg/m3 (0.2 ppm) NO or
2.8 to 3.8 mg/m3 (1.5 to 2.0 ppm) NO2 plus
245  Mg/m^  (0-2 ppm) NO 16 hours a day for
18 months did not  change weight gain in the
12 beagles in either experiment.1 8
   In  contrast, rats exposed continuously to
22.6 mg/m3  (12 ppm) NO2 for 9 months con-
tinued to grow, but their  body weights re-
mained 20  percent below those  of control
animals.31'34 One observer44 reported a 10
percent weight loss in rabbits exposed to 5.6
Mg/m3 (3 ppm) NO2 for 15  weeks,  while
controls  added 11  percent to  their weight.
The  effect was apparently dose dependent,
for rabbits exposed to 2.5 mg/m3 (1.3  ppm)
NO2  for 17 weeks increased their weight 2
percent compared to an 8 percent increase by
controls.
(3)   Voluntary Behavior.  Six-hour exposures
of 6.9 to 39.3 mg/m3 (3.7 to  20.9 ppm) NO2
depressed  the  voluntary running activity of
male mice, when the concentration was 14.5
mg/m3  (7.7 ppm)  or greater. Although the
threshold  was  not identified, it  could  be
narrowed to between 6.9 and 14.5 mg/m3. In
all cases, activity  returned to normal on the
first post-exposure day.9
(4)  Hematologic  Effects.    Studies of the
effect of NO2 on  the circulating blood have
included both short- and long-term exposures.
In one short-term  study,  a group45 exposed
dogs to 73.3 and  99.6 mg/m3 (39  and  53
ppm) for 60 minutes, 97.8 and 160 mg/m3
9-10

-------
(52 and 85 ppm) for 15 minutes, and 235 and
308 mg/m3  (125  and  164  ppm)  for  5
minutes, and found no changes in the hema-
tocrits and blood platelet counts determined
4, 24, 48, and  72 hours after exposure. The
same  was true  in  dogs exposed  to 1.9 and
47.0 mg/m3 (1  and 25 ppm) NC>2  6 hours a
day for 18 months.'7
   Freeman,  et  al.37  found polycythemia in
rats allowed to breathe 3.8 mg/m3  (2 ppm)
N(>2  or more,  continuously. The concentra-
tion of erythrocytes rose within 2 to 3 weeks
and achieved levels  of  about  40 to  100
percent  above  baseline values.  After some
delay,  both  hematocrit  and  hemoglobin
levels rose, but  to a lesser degree, so that the
fully  developed polycythemia  was  charac-
terized  by a reduction in mean  corpuscular
volume,  mean  corpuscular  hemoglobin,  and
mean corpuscular hemoglobin concentration.
Cellular  diameters  measured on the flat sur-
face of  stained smears  were  normal.  The
blood  of monkeys (M.  speciosa) exposed to
3.8 and 16.9 mg/m3 (2 and  9  ppm) NC>2
followed a similar course.37
   Leukocytosis  occurred  in the peripheral
blood of rabbits exposed to both 2.5 and 5.6
mg/m3 (1.3  and 3.0 ppm) NO2 for 15 to 17
weeks  but  it  receded  with  cessation  of
exposure. The  leukocytic response was ac-
celerated by the presence  of SO2- Phagocytic
activity  was  depressed  by  NO2 and  both
leukocytic and  phagocytic alterations  were
greater at the higher NC>2 exposure.44
   Methemoglobin (MeHb) has been detected
in the blood of animals exposed to NC>2 as
well as to NO. Cats, rabbits, guinea pigs, mice,
and rats were  exposed  to 56.4  to  1,880
mg/m3  (30  to  1,000  ppm) NC>2 for un-
specified periods of time. MeHb  was  not
detected in   animals receiving  up  to  103
mg/m3 (55 ppm) NC^. Cats exposed to  188
mg/m3  (100   ppm)  showed  detectable
amounts of MeHb after approximately 1 hour
of exposure,  however.  At concentrations of
282 mg/m3  (150 ppm) NO2 and above, the
blood of all  animals tested contained MeHb
after 1 hour of exposure, but the MeHb was
no  longer  measurable  1  to 2  hours after
animals were returned to clean air. In these
experiments  the  NC<2  values reflected the
total  oxides  of nitrogen present in the ex-
posure chamber; the concentration of NO was
not evaluated separately.46
  In  other experiments, MeHb  appeared in
the blood  of rabbits  and  rats  exposed to
fumes produced in  the electric-arc welding
process.47 Rats exposed for 6 hours a  day, 5
days a week,  for 43 days, to fumes containing
45.1 mg/m3 (24 ppm) NO2 developed  4.5 to
21.7 percent  MeHb (mean of 13.6 percent).
Detectable  levels  were  still  present on the
eleventh  day  of  the post-exposure period.
Rabbits similarly exposed for 45 days, formed
an average of 2.8 percent MeHb (range  0.9 to
4.5  percent). On  the sixth  day of the post-
exposure period, detectable amounts re-
mained  in  male  rabbits only.  Male  rats
exposed to 132 mg/m3 (70 ppm) NO2 for 6
hours a day formed  2.6 percent  MeHb after
the  first day of exposure and 3 percent by the
third  day.  Three  days  after exposure, the
levels  were within the control range. In these
experiments,  it was  estimated that between
12  and  17   percent  of the  nitrous  fumes
existed as NO.47
  A mixture of 180  ppm  NO2 plus
caused death in 12 dogs within 55 to 285
minutes.  MeHb  concentrations  increased
significantly (p < 0.01),  with an average  of
325 mg per 100 g of blood before, and 700
mg per 1 00 g after, exposure.4 8
(5)    Immunologic  Effects.  A  circulating
substance with properties similar to a lung
antibody appeared in the serum of guinea pigs
exposed to 9.4 mg/m3 (5 ppm)  NO2 either 4
hours a day 5 days a week, or 7.5 hours a day
5 days a week, for up to  5.5 months. In a
second group,  exposed to  28.2  mg/m3  (15
ppm)  NO2  continuously  for  1   year,  the
antibody  reacted  in vitro  with  proteins
extracted from the  lung tissue of control
animals.  The titers of reactive substances in-
creased with  the intensity and duration of
exposure,  but  no  absolute  values  were
                                                                                  9-11

-------
ascribed  to  the data because the latex ag-
glutination method employed is not quanti-
tative.16
(6)  Effects on Enzyme Systems. The effect
of exposure to 28.2 mg/m3  (15 ppm) NO2
continuously  for  10 weeks; on oxygen  con-
sumption, lactic dehydrogenase, and aldolase;
in guinea pig  lung, liver,  spleen, and serum
was  determined.49  The following significant
changes  occurred:  oxygen consumption in-
creased in spleen and kidney; lactic dehydro-
genase activity increased in lung, liver, and
kidney;  and  aldolase  activity  increased  in
tissues. Short-term  exposures also caused  a
significant increase in  the activities of the
same enzymes  in some tissues. Exposure  of
guinea pigs to 75.2 mg/m3 (40 ppm) for 30
minutes at 2-hour intervals for a total of 4.5
hours  revealed the  following significant
changes:  oxygen consumption increased  in
liver, spleen, and kidney; lactic dehydrogenase
activity increased in liver, kidney, and serum;
and aldolase activity increased in liver, spleen,
kidney,  and   serum.  The mechanism  that
causes these enzymatic changes is not known,
but  it might  reflect  a general  response  to
stress.
   Following a 2-hour exposure  to 65.8 and
94.0 mg/m3  (35 and  50  ppm) NO2,  lactic
dehydrogenase isoenzymes from lung tissue of
squirrel  monkeys  shifted  to predominantly
anaerobic band 5 when  chromatographed, and
after a 2-hour  exposure to 18.8 mg/m3  (10
ppm) NC>2, a  similar increase in band 5 was
noted  in  serum  lactic  dehydrogenase
isoenzymes. This change in isoenzyme pattern
was  present at 2 days, but  returned to normal
8  days after  termination  of  the NO2 expo-
sure. A similar shift to anaerobic isoenzyme
was  observed in the heart tissue of hamsters
exposed  for 2 hours to 9.4 and 65.8 mg/m3
(5 and 35 ppm) N02-15'50
d.   Susceptibility to Respiratory Infection
(1) Bacteria and Influenza. Study of rabbits
exposed  to  concentrations of  NO2 ranging
from ambient Cincinnati levels to 113 mg/m3
(60  ppm) revealed increased   numbers  of
polymorphonuclear leukocytes (heterophiles)
in lung washings; this condition persisted for
more than 72 hours after a single 3-hour ex-
posure. When streptococci were instilled into
lungs of  NC>2-exposed, anesthetized  rabbits
30  minutes before  lavage, a  pronounced in-
hibition of phagocytic activity was observed,
in comparison to controls.51
  Exposure of mice, hamsters,  and squirrel
monkeys to NO2 increases susceptibility  to
bacterial pneumonia and influenza infection.
This susceptibility,  observed  in  both short-
and  long-term exposures,  is  based  on three
parameters:  (1) increased mortality rates, (2)
reduced survival times, and (3) reduced ability
to clear inhaled infectious agents from  the
lungs, as determined by the number of viable
organisms that can be cultured. Mice in these
experiments were exposed for  2 hours to NC>2
in concentrations ranging from  2.8  to 47.0
mg/m3  (1.5 to 25 ppm), then  challenged with
an aerosol  of Klebsietta pneumonia within 1,
6, and 27 hours after the NC>2 exposure.5 2<53
The  minimum NC>2 concentrations required
to produce a  statistically significant rise in
subsequent  mortality  was 6.6 mg/m3  (3.5
ppm)  for  2  hours  when  the  infectious
challenge  took place 1 hour  after  the NC>2
exposure. When  the infectious challenge was
delayed,- a statistically significant effect was
noted at 6, but not at  27 hours,  at NO2 con-
centrations of 9.4 mg/m3 (5 ppm) and above.
Exposure to 47.0 mg/m3 (25 ppm) NO2,  6
and  14 days prior  to  the  challenge with  K.
pneumoniae did not increase mortality. The
same authors reported that when mice were
infected with K. pneumoniae first,  and then
exposed for 2  hours to 47.0 mg/m3 (25 ppm)
NO2, within 1, 6, 27, 48,  and 72 hours after
the  infectious  challenge,  the mortality in-
crease was statistically  significant. This effect
was  not observed  at 4.7  mg/m3 (2.5 ppm)
NO2.
  A 2-hour exposure to 9.4 mg/m3 (5 ppm)
NO2 in various inbred  strains  of mice (BDFj,
BALB/c, C57BL, and LA?!)  either before or
after infectious challenge with an aerosol of
K. pneumoniae favored increased mortalities.
A similar  enhancement  in  mortality  was
9-12

-------
observed in hamsters exposed for 2 hours to
65.8 mg/m3 (35 ppm) N(>) and above, and
challenged with K. pneumoniae.5 °
   The rate of clearance  of inhaled bacteria
from  the lungs of mice  and hamsters was
reduced upon short-term exposure to NC^-5 °
The animals were exposed to 9.4 mg/m3 (5
ppm) NC>2 for 2 hours,  and within 1 hour
challenged with K. pneumoniae aerosols. Im-
mediately, and at 1, 3, 5, 6, 7, and 8 hours
after  the infectious challenge, the animals
were  sacrificed and K. pneumoniae in lungs
was quantitatively assayed. In control animals
the  bacterial population  was  markedly
reduced  during the  6 hours  following the
challenge.  Thereafter, the  population  in-
creased, reaching its initial concentration after
approximately 8 hours. In mice and hamsters
exposed  to  9.4 mg/m3  (5  ppm)  NC«2, the
period of initial clearance was reduced to 4.5
and 5 hours,  respectively, and the original
concentration was re-established in less than 7
hours.4 8
   In  a  long-term  study,  Ehrlich  and
Henry43-54 exposed  four groups of mice to
940 Mg/m3 (0.5 ppm) NO2 for 6, 18, and 24
hours a day,  7 days a week,  for  up to  12
months.  Challenge  with  an  aerosol  of  K.
pneumoniae took place after 1, 3, 6, 9, and
12 months  of exposure.  Statistically  sig-
nificant increases in mortality were observed
after continuous exposure to 940 Mg/m3 (0.5
ppm)  for 3 months, and after 6- and 18-hour-
a-day  exposures  for 6  months.  After  12
months  exposure,  mortality could  only  be
increased significantly  when the  NC>2 ex-
posure was continuous. The clearance rate of
viable bacteria  from  the  lungs of  mice was
also affected  by long-term exposure  to 940
Mg/m3 (0.5 ppm) NO2- In mice exposed to
NC>2 for 6 and 18 hours a day for  9 months
some reduced capacity to  clear bacteria from
lungs  was observed, and after 12 months ex-
posure a significant  inhibition of bacterial
clearance was apparent. Mice exposed to NO2
fot 24  hours  a  day  showed  significantly
reduced capacity to clear viable bacteria after
6, 9, and  12 months of exposure.
   Henry et al.15 exposed squirrel monkeys to
 18.8 to 94.0 mg/m3 (10  to 50 ppm) NC«2 for
 2  hours and  challenged  them by  the  intra-
 tracheal route with K. pneumoniae. Exposure
 to 94.0  mg/m3  (50 ppm) NC>2  for 2  hours
 was  not fatal,  whereas  the  same exposure
 followed by challenge with K. pneumoniae
 was  fatal to three  out  of three  monkeys.
 Squirrel monkeys exposed to 18.8 mg/m3 (10
 ppm) NC>2 for 2 hours and then challenged
 with K. pneumoniae had bacteria present in
 their lungs 19 to 51 days after challenge. In
 many   bronchioles,  epithelial  cilia  were
 missing;  cells had proliferated; and lympho-
 cytes and  polymorphonuclear cells had in-
 filtrated the collapsed areas.
   The same group5 5 exposed male squirrel
 monkeys continuously to NO2 and challenged
 them with K,  pneumoniae aerosols.  Out of
 four monkeys exposed  to 18.8  mg/m3 (10
 ppm) NC>2 for  1 month, one died and two
 had the infectious agent  present  in the lungs
 at  autopsy.  Two  out  of seven monkeys
 exposed to 9.4  mg/m3 (5 ppm) NC>2  for 2
 months died, and five had the infectious agent
 present in the lungs at autopsy.
   Squirrel monkeys were also experimentally
 infected with  influenza  A/PR-8  virus,  24
 hours before continuous exposure to 18.8 and
 9.4 mg/m3 (10  and 5  ppm) NO2-  All six
 monkeys exposed to 18.8  mg/m3  (10  ppm)
 died within 3 days, while at 9.4 mg/m3 (5
 ppm) one  out  of  three  succumbed to the
 disease. There were no deaths in the control
 group challenged with influenza virus.5 3
  When  the squirrel monkeys were exposed
to 9.4 mg/m3 (5 ppm)  NC>2 for  5 months
and, during the  exposure, challenged  three
times by  the  intractracheal route with the
influenza  A/PR-8 virus,  the  formation of
protective  serum  neutralization  antibody
appeared to be  depressed as evidenced  by
serum  neutralization  antibody   titers  and
hemagglutination-inhibition titers.5 6
(2)  Bacteria and Tobacco Smoke.  Combined
exposures to  NC>2, tobacco smoke, and K.
pneumoniae were reported for hamsters.42-57
A  2-hour exposure to 28.2 mg/m3  (15  ppm)
                                                                                   9-13

-------
NC>2  followed by  1-hour exposure to 3 per-
cent   (v/v)  tobacco  smoke  significantly
decreased the resistance to  bacterial pneu-
monia, as evidenced by enhanced mortalities.
Furthermore the combined exposures reduced
the rate  of clearance  of viable bacteria from
the lungs of hamsters  to a greater extent than
exposures to the  individual  pollutants.  The
combined  exposure  produced  irreversible,
marked alterations in  surface morphology of
lung tissue.  The typical structure of mucus-
secreting cells was not  apparent, and a marked
loss of cilia was noted 2 days after exposure.
(3)  Interferon  Formation.  Impairment   of
interferon formation  by NC>2 exposure  has
been  demonstrated by challenging rabbit  al-
veolar   monocytes  with  parainfluenza-3
virus.58 Rabbits exposed to 49.0 mg/m3 (25
ppm)  NC>2 for 3  hours, either  immediately
after,  or at  0, 3, 6, 12,  or 24 hours  before
innoculation with rabbit  pox virus,  failed to
develop  the  anticipated  resistance  to
infection. This inhibition of resistance per-
sisted  for at least  96 hours after  NC>2 ex-
posures.  During this  period, alveolar mono-
cytes from  NC>2—exposed  animals  were un-
able to produce the antiviral substance, inter-
feron,  in vitro when innoculated  with the
parainfluenza-3 virus.

2.   Effects in Man
a.   Experimental Exposures
(1) Odor Perception. In a series of experiments
with healthy male  volunteers  between  the
ages of 20  and  35 years, individuals were
exposed to varying concentrations of NC>2 in
a  specially  designed  chamber,  and  the
olfactory  threshold  was  measured under
various conditions  (Table  9-6).  At a  con-
centration of 225 Mg/m3 (0.12 ppm) NO2,
only a few  subjects perceived  the  odor im-
mediately. Perception was immediate in over
                     Table 9-6. RECOGNITION OF N02 ODOR BY HEALTHY
                    YOUNG ADULT MEN AFTER VARIED EXPOSURES IN AN
                                EXPERIMENTAL CHAMBER59

Concentration,3

mg/m^
0.225

0.415

0.835

8.0

19.9

39.9

56.8


ppm
0.12
(0.1-0.15)
0.22
(0.16-0.28)
0.42
(0.39-0.46)
4.0
(3.0-4.2)
10.6
(10.1-11.2)
19.7
(19.3-20.1)
30.2
(29.1-31.4)



Subjects
9

13

8

12

4

8

3

Odor perception on entry
to chamber
Number of
subjects
3

8

8

12

4

8

3

Duration,3
min
5
(0.5-10)
3
(MO)
2.5
(1-5)
5.5
(3-10)
8
(6-13)
12
(5-24)
32
(25-40)
               3The range of actual measured values is included in parentheses.
9-14

-------
half of the  subjects exposed  to 415
(0.22 ppm)  NO2- At higher concentrations,
beginning with 835  Mg/m^ (0.42 ppm) NC>2,
all subjects recognized the odor immediately.
Persistent recognition  of  the odor varied
considerably, ranging from a few minutes at
the lower exposure levels to over 30 minutes
at some of the higher concentrations.5 9
   In another study60 of 14 subjects, the N©2
olfactory  threshold  was 230   jug/m3  (0.12
ppm).  The effect of NC>2  and SC>2 was ad-
ditive,  i.e., a lower concentration of each gas
led to odor perception  if both gases were
present simultaneously.
   Recently  diesel exhaust odor has been the
subject of considerable study and it has been
found  that NC>2 may contribute to the odor
of diesel exhaust. An assessment of the major
industrial odorants in diesel exhaust revealed
that the  mixture  with the  greatest odor
contained  about  940   Mg/m3  (0.5  ppm)
NQ2.61

(2)  Pulmonary  Function. In  one  experi-
ment,  exposure  to  NC>2 was controlled in a
special room equipped with an air filter.62
Six  normal  subjects and  four  patients with
"moderate  to  marked" pulmonary  disease
were exposed to concentrations of 940 to
5,640  jug/m3 (0.5 to 3 ppm) NC>2 on several
occasions for 2  to 3 hours at a time. Several
physiological  parameters were   measured
before, during, and after exposure,  when the
subjects  were either  at  rest or exercising.
Smokers were not excluded from  the study,
but  were required to  abstain from  smoking
for the  8-hour period prior to the chamber
study.  The data showed no consistent changes
in airway resistance, pulse rate,  respiratory
rate, or  subjective complaints that could be
related to the NO2 challenge.
   NC>2 and  SC>2 produced additive effects on
pulmonary  function in  five healthy males,
ages 21  to  40,  who were judged  to be  free
from  respiratory   disease.63  One  subject
smoked  5 to 6 cigarettes a day. Each subject
was exposed on separate  occasions to 7.5 to
9.4  mg/m3  (4 to 5 ppm) NO2 and  10.5 to
13.1 mg/m3 (4 to  5  ppm) SO2. Each ex-
posure  was for  10 minutes,  with  2-week
intervals between  exposures.  The  subjects
wore  nose clips  and  inhaled  the  gas-air
mixtures through a  mouthpiece. Inspiratory
and expiratory flow resistance and pulmonary
compliance  were measured before, immedi-
ately after, and 10, 20, and 30 minutes after
exposure.  Vital capacity,  FEV} (1 - second
forced  expiratory  volume),  maximal  mid-
expiratory  flow rate, and peak flow rate were
measured   prior to  and  30  minutes  after
exposure.  Inhalation of NO2 caused an in-
crease  in both inspiratory and expiratory flow
resistance,  the  maximum  recorded  increase
occurring  30 minutes  after the  end of ex-
posure. No data were given for the recovery
time. Mean pulmonary compliance was un-
changed, immediately after exposure,  but was
slightly decreased 30 minutes after exposure
(significant  at p  <0.10 but not at  <0.05).
Lung volumes and  peak  flow  rates  did not
change significantly. Inhalation of SC>2 caused
an increase in both inspiratory and expiratory
flow resistance that was maximal immediately
after inhalation and restored to pre-exposure
levels 30 minutes later.
  Exposure to  a mixture of 4.7 mg/m3 (2.5
ppm) NX>2  and 6.6 mg/m3  (2.5 ppm) SC>2
produced a bimodal increase in both inspira-
tory and expiratory flow resistance. The first
increase, which corresponded to the effect of
SC>2 alone, occurred immediately after ex-
posure. The second increase  in resistance
corresponded to the effect of NC>2 alone, and
was maximal 30 minutes after exposure. SC>2
is known  to  cause  an  immediate reflex
increase  in airway resistance  along  para-
sympathetic pathways.64  The mechanism by
which  NC>2 induces resistance is unknown,
but the delay suggests that it is different from
that of SO2.6 3

b.   Occupational exposures
  There are numerous possibilities  for oc-
cupational  exposure to NOX. Hazardous oc-
cupations include the  manufacture  of NO,
nitration  of cellulose,  and  other  organic
                                                                                    9-15

-------
materials; electric-arc welding; and  photo-
engraving. Accidental exposures have resulted
from the combustion of celluloid and nitro-
cellulose  films, and  from inhalation of silage
gas.
  Four clinical types of poisonings  due to
occupational exposures65 have been encoun-
tered: (1) irritant gas, (2) reversible, (3) shock,
and  (4)  combinations of  the  other three.
Irritant gas exposure is characterized initially
by severe irritation that results in pain, burn-
ing,  and  choking in the  throat and chest;
violent  cough; and the  expectoration  of
yellow-tinged sputum. The  reversible  type is
characterized by dyspnea, cyanosis, vomiting,
vertigo, somnolence, a feeling of intoxication,
fainting,  loss of  consciousness,  and  methe-
moglobinemia. Persons suffering this  type of
poisoning do not  develop pulmonary edema,
and  if removed  from  the exposure early
enough, may recover completely. Otherwise,
the  poisoning  may rapidly become fatal.
Shock-type patients immediately show severe
symptoms of asphyxiation, convulsions,  and
respiratory arrest; death is presumably due to
stasis in the pulmonary circulation.  This form
is exceptional and may result from the sudden
inhalation of high concentrations. In the com-
bined  type  the  patient immediately  shows
symptoms referable to  the central nervous
system, such as  vertigo,  somnolence, and a
staggering gait. There may be cyanosis. After
apparent  recovery, this stage may be followed
some  hours later by  progressive  dyspnea,
marked cyanosis, and pulmonary edema.
  Several  accidental  exposures  to silage
gas,66'67  which may contain very  high con-
centrations of NC>2, have been reported. In
several instances the exposure proved fatal.
One man died 20 hours after being exposed to
this type  of gas for 5 to  8 minutes.64 One
person exposed at the same time for 2 to 3
minutes survived,  although  he suffered acute
pneumonitis. Exposure to NO2 has been  ex-
plained in six  degrees,  with  corresponding
clinical syndromes:  (1)  at 940mg/m3  (500
ppm) or  higher victims develop acute pulmo-
nary edema  and die within 48 hours; (2) at
564 to 752 mg/m3 (300 to 400 ppm) pulmo-
nary edema with broncho-pneumonia devel-
ops and  death  ensues in  2  to  10 days; (3)
those exposed to 282 to 376 mg/m3 (150 to
200 ppm) develop bronchiolitis fibrosa obli-
terans,  which is fatal in 3 to  5  weeks; (4)'
exposure to 94  to  188 mg/m3 (50  to 100
ppm)  produces  bronchiolitis with  focal
pneumonitis lasting 6 to 8 weeks, followed by
spontaneous recovery; (5) individuals exposed
to NO2 in the range of 47 to 141 mg/m3 (25
to 75 ppm) develop varying  degrees of bron-
chitis and  broncho-pneumonia, but  recover
completely;  and (6)  chronic  intermittent
exposure to  concentrations  of NO2  in the
order of  18.8 to 75.2 mg/m3 (10 to 40 ppm)
may produce chronic pulmonary fibrosis and
emphysema.
   Occupational exposure to  NO2 is frequent-
ly  encountered  in  a  variety  of  welding
processes. Evaluation of responses to  inhala-
tion of  such fumes is  complicated by the
presence  of other contaminants. Predominant
among these are NO, ozone, manganese, ferric
oxide  fumes, and  particulates of  various
types.  In a case of acute poisoning from NOX
from an  oxyacetylene  torch, a welder died
from chemical pneumonitis  10  days  later.68
Other,  workers  present  during  the exposure
developed  coughs  and  other  respiratory
symptoms.  Random  air  samples obtained
under simulated conditions revealed oxides of
nitrogen  in the range of 71.4 to 662 mg/m3
(38 to 352 ppm), occurring as  NO2- After 3
minutes exposure to fumes containing 395 to
714 mg/m3 (210 to 380 ppm)  NO2, welders
developed a dry cough and tightness in their
chests.  These symptoms  disappeared  when
they returned to clean air.
   Pulmonary edema developed 18 hours after
a 30-minute accidental exposure to the fumes
of an  oxyacetylene torch estimated  to have
produced approximately  169  mg/m3 (90
ppm) NO2- In pulmonary function tests con-
ducted at that time the vital capacity was 50
percent of the  expected level,  but the 0.5-,
1.0-, and 2.0-second FEV's  were not dimin-
ished.69
9-16

-------
   The  effects of electric-arc  welding ex-
posures from both bare and coated rods have
also been examined.  In an experiment with
bare  rods, two men  were exposed  to the
fumes of the  electric-arc welding process for
190 minutes.47 '70 The average concentration
of nitrous fumes, expressed as NC>2, was 158
mg/m^ (84 ppm), with a  maximum  of 175
mg/m^ (93  ppm)  that lasted for 40 minutes
and a peak of 194  mg/m^ (103 ppm) lasting 5
minutes.  These concentrations measured  by
the  total nitrate method actually  include an
unspecified mixture of NO-NO2-N2O4- Sub-
jects were also exposed to approximately 785
/ug/m^   (0.4  ppm)  ozone throughout  the
period. They reported no headache, eye irrita-
tion, throat  irritation, or  other  ill  effects,
however.
   In welders  exposed  to  the  fumes  from
electric-arc welding using  coated rods47 for
the  same length  of time,  the average MeHb
concentration was 2.5 percent and the maxi-
mum  was 3  percent.  The  concentration of
"nitrous  gas"  in  the  working area  did not
exceed 25 mg/m^ (13.3 ppm),  expressed as
NC>2, of which not more than 20 percent was
estimated to be NO.
   The possible long-term effects of occupa-
tional exposure to  nitrous fumes was reported
by Becklake  et al.71  in a study of seven
miners  who  were  accidentally  exposed  to
nitrous fumes for periods of 5 to 75 minutes
and  who developed pulmonary edema  3 to 27
hours after  exposure  (see  Table  9-7). The
investigators studied pulmonary function at
the time of each patient's discharge from the
hospital  and  compared yearly follow-up ex-
amination data to the mean values for the
same data from 16 normal persons.
   Among the seven patients, the most com-
mon  effects  were a  reduction  in maximal
breathing capacity and an increase in  expira-
tory  resistance; the latter functional  impair-
ment was common to all  the patients who
complained of exertional dyspnea.  In the two
patiejits who felt that  their capacity for work
was unaffected by  the accident, tests showed
thatf pulmonary function  had  returned  to
normal  during the study period. Chest roent-
genograms of those patients with more than 6
years of underground mining service, taken
within a year prior to the exposure, had been
normal.  Following the  accident, the arterial
oxygen saturation was not affected  either at
rest  or  during exercise. The mixing  index, a
measure of efficiency in clearing nitrogen gas
from the lungs, was  significantly different
from predicted  values  in  four  patients,  al-
though  it returned to normal in the course of
the study period in the three who could  be
followed. The  investigators  concluded that
the exposed individuals sustained a degree of
bronchial and bronchiolar narrowing due to
fibrosis, secondary to various degrees of bron-
chiolitis obliterans.
   A study of the survivors of the Cleveland
Clinic fire in  1967, who also were exposed to
high concentrations of nitrogen oxides, is dis-
cussed in chapter 10, Section C.I.
   Multiple clinical  symptoms, biochemical,
and  hematologic changes were described  in
workers engaged in the manufacture of sulfur-
ic acid and hence exposed to an average of 4.9
mg/m^  (2.6 ppm)  NC<2 for 3 to 5  years.72
Because the data are not  supported  by diag-
nostic  criteria,  this  report cannot be evalu-
ated. In contrast, Italian workers employed in
the manufacture of nitric acid and exposed to
an average of 56.4 to 65.8 mg/m^ (30 to 35
ppm) NC»2 for an unspecified number of years
exhibited no  signs or symptoms of injury.73
D.   OTHER OXIDES OF NITROGEN
   Although  several other  oxides of  nitrogen
besides NO and NO2 exist in the atmosphere,
they are generally present in very small quan-
tities and except for nitrous oxide (^O) have
not  been of  toxicological  interest. Informa-
tion  concerning their effects is likewise scarce.
   Considerable  attention has been focused on
N2O for the past century because of its anes-
thetic and analgesic properties. At concentra-
tions of 80 percent (1.44 x  106 mg/m3), N2O
is  an effective general anesthetic.  Inhalation
of 10 to 20 percent  ^O (1.8  x  105 to 3.6 x
10^  mg/m^) provides effective  analgesia.74
Although  extensive  information relevant  to
                                                                                    9-17

-------







rt

.a
c
U
o
(Summary












i
-i >,
|I
J
c
0

•£"
2 =
•a |
£ C
'







ng volume
J








c S
x a
O
l|
- O
Sni
?
c C.
« u
11
Compliance,
liters/cm

o
0
13
00
c
.2
S
1"
>
KJ
a
P£
LL.

X
u
VI
IH
-
1
O
on
I
«
fS
?
s
C5

5 5 55
•— ' 00 »O
c™* ^o oo

«
>sD CM fl f*-
iO <*o ro rs
C4 O O
ON OO — "

V£J ^H M
— * OO OS
S 55 5

4> U 0)
^P- 00
NO OO OO
C4 m m

•0 VO NO
s
0
(N
O
3
TT
0

f) ^ >o O ^
>O 00 ON ON i""1
ON --^ 00 vO ^*
OO Tt "0 «f OO



p- m (N (N (N
(N ON 00 ^t" f^
fi ~i
ON NO

r*~ oo oo
O *N (N
cn vo rf

a>
WJ TJ- OS
fO fS (N
o rf i>


(P OS »O
O< O ^
O 00 OS


O O ON
K lo 2


(N
m
m
2
oo
vn
fN
O

OO NO »O

-------
these properties of N2O is available, its inclu-
sion is beyond the scope of this review.
E.   FUTURE RESEARCH NEEDS
   To cope with the problem of NOX pollu-
tion, future research should include:
   1. Further  delineation of  in vivo  bio-
chemical and biophysical effects of nitrogen
oxide exposures relative to:  (a) oxidation of
fatty acid double bonds in cellular lipids and
in lung surfactants; (b) denaturation or altera-
tion of lung proteins  (collagen  and elastin,
enzymes, and cellular membranes).
   2. A definition  of  the  relationship  of
metabolic tissue changes to NC>2 dosage in
terms of a concentration-time response. The
relative  importance  of  low-concentration,
long-term  exposures versus  short-term  peak
concentration should be studied.
   3. The  determination  of interaction of
NOX with particulate pollutants  in terms of
biochemical,  biophysical,  infectious,  and
ultrastructural responses.
   4. The  assessment  of the biologic impor-
tance of tolerance  and cross tolerance with
other oxidant pollutants.
   5. Expanded studies  to determine the
effects of  NO2  on the  infectious  defense
mechanisms and immunological processes.
F.   SUMMARY
   The  two  oxides of nitrogen  present in
ambient air in greatest quantities, nitric oxide
(NO) and nitrogen dioxide (NC>2), are poten-
tial health hazards. Summaries of laboratory
studies  of animal  exposures  are given in
Tables  9-8, 9-9, and 9-10. The toxicology of
nitrous  oxide  (N2O)  and other oxides of
nitrogen does not appear to be relevant to the
problems  of  ambient  air  pollution at the
present time.  N2O is analgesic at concentra-
tions of about 1.8 x 105 to 3.6 x 105 mg/m3
(1.0 to 2.0 x  10^ ppm) and a general  anes-
thetic at concentrations of 1.4 x  10^ mg/m3
(8.0 x 10^ ppm) and above.
Nitric Oxide
   At/ concentrations  found in  the  atmos-
phere, NO is  not  an irritant and is not con-
sidered  to  have adverse health  effects.  Its
main  toxic potential  at ambient  concentra-
tions  results from  its oxidation  to NO2-  A
12-minute exposure to 3,075 mg/m3 (2,500
ppm) NO was lethal  to mice. Lower doses,
from  24.6  mg/m3  (20 ppm) to  1.23 x 104
mg/m3 (104 ppm), of NO produced reversible
inhibition of bacterial hydrogenase activity.

Nitrogen Dioxide
Animal Studies
  NO2 exerts its primary toxic effect on the
lungs.  Most  of the  available  information
comes from studies with animals.  Concentra-
tions greater than 188 mg/m3 (100 ppm) are
lethal to  most animal species, and 90 percent
of  the   deaths are  caused  by   pulmonary
edema. The mortality rate may be modified
by  varying the exposure  product concentra-
tion x time (Ct),  the  temperature, and the
presence  of other irritants.

Pulmonary function.  Short-term exposures to
nonlethal  concentrations  of   NO2  have
produced   transient  pulmonary-function
changes in  the lungs of animals.  Guinea pigs
have shown increased respiratory rates and de-
creased tidal volumes after exposure to con-
centrations of 9.8 mg/m3  (5.2 ppm)  for 2
hours, and  24.4  mg/m3  (13.0 ppm)  for 1
hour. Pulmonary function returned to normal
when animals were returned to  clean air.
Similar changes in pulmonary function were
observed in squirrel monkeys exposed  for 2
hours to  18.8 to 94.0 mg/m3 (10  to 50 ppm)
NO2  and in other monkeys exposed contin-
uously for  2 months to 9.4 mg/m3 (5  ppm)
NO2- Monkeys  (M. speciosa) continuously
breathing 3.8  or  16.9 mg/m3 (2  or  9  ppm)
NO2 developed tachypnea which has persisted
for almost 2 years.

  Rats continuously exposed to  1.5 mg/m3
(0.8 ppm)  NO2 maintained  a 20  percent in-
crease in respiratory  frequency   throughout
their  lifetimes.  Dogs exposed to  0.9 to 3.8
mg/m3 (0.5 to 2.0 ppm) NO2 plus 245 jug/m3
(0.2  ppm)  NO for  18  months, however,
showed no  signs of  changes in  pulmonary
function.
                                                                                   9-19

-------
Metabolism. The structure of lung collagen
and elastin was altered in rabbits exposed to
1.9 mg/m3 (1 ppm) NC>2 for 1 to 4 hours, but
the alteration appeared to be reversible within
24 hours.  Similar changes were observed in
rabbits exposed  to 470  jug/m3  (0.25 ppm)
NC>2, 4 hours a day, for 6 days, but recovery
was delayed and some damage persisted after
7 days. It  has been  suggested that the meta-
bolic  activity  of the lung tissue is reduced
following exposure to NC>2, and that repeated
exposure,  with  associated  denaturation  of
collagen  and elastin, may be a factor in the
pathogenesis of pulmonary emphysema.

  Lung lipids extracted from rats, exposed to
1.9 mg/m3 (1 ppm) NC>2 for 4  hours were
peroxidated. The process was delayed and sus-
tained, reached a maximum at 24 hours after
exposure,  and was maintained for  an addi-
tional 24 hours. Rats fed a vitamin E-deficient
diet and  exposed to NC>2 had more peroxida-
tion  of surfactant and tissue lipids than simi-
larly exposed rats receiving vitamin E  supple-
mentation.

Pathology.  Pathological  lesions  have been
observed  in the  lungs of animals following
both short- and long-term NC>2 exposures. At
exposures to 141 to 188  mg/m3 (75  to  100
ppm) for 2 hours, rabbits, rats, and  guinea
pigs  developed acute inflammation  in bron-
chiolar epithelium, but appeared  to recover 2
weeks after exposure. At 1.9 mg/m3 (1 ppm)
for 1 hour, or 940 Mg/m3  (0.5 ppm) for 4
hours, mast cells of rat lungs became degranu-
lated, possibly signifying the onset of an acute
inflammatory  reaction.  These cells  returned
to normal 24 hours after the end  of exposure.
More  serious damage was found  in lungs of
squirrel monkeys exposed for 2 hours to 18.8
to 94.0 mg/m3 (10 to 50 ppm) NO2; the pri-
mary  lesions  found  at  low concentrations
became progressive alveolar expansion. Hyper-
plasia of the  respiratory  bronchiolar  epithe-
lium was seen in  hamsters exposed to  84.6 to
103  mg/m3 (45  to 55 ppm) for  10 weeks. A
similar response  was noted in the major bron-
chi and distal portions of the respiratory tract
in hamsters  exposed to  188  mg/m3 (100
ppm) for 6 hours.
Chronic Changes. Since certain  pathological
changes seen in animals after experimental
NO2 exposure  are similar to changes occur-
ring in  the pathogenesis of chronic obstructive
pulmonary disease in man, it is suggested that
long-term, low-level exposures to this pollut-
ant may play a significant role in the develop-
ment of chronic lung disease.
  Improved  histochemical and  electron
microscopic  techniques have  shown  that
long-term  exposure of rats to NC>2 at concen-
trations that do not produce acute inflamma-
tory responses  have a  cumulative  and sus-
tained  effect. Emphysema-like  lesions were
produced in the rat lungs with concentrations
of 18.8 to 47.0 mg/m3 (10 to 25 ppm). Rats
exposed  to 3.8 mg/m3  (2 ppm)  for their
natural lifetime had less cilia; less-than-normal
bronchiolar, epithelial blebbing; and crystal-
loid, rod-shaped, intracytoplasmic  inclusion
bodies  in their bronchiolar epithelium.  Similar
effects have  been seen  occasionally  in rats
continuously exposed  to  1.5 mg/m3  (0.8
ppm).  Mice exposed  to 940 Mg/m3 (0.5 ppm)
for 3 to 12 months  on 6-, 18-, and 24-hour
daily schedules have shown  increase in  the
size  of alveoli, due to alveolar  distension
rather  than  septal  breakage.  Additionally,
inflammation of the bronchiolar epithelium
with a reduction in distal  airway  size sug-
gested  the development of early focal emphy-
sema.  Rats  exposed chronically to 18.8  to
47.0 mg/m3 (10  to  25 ppm) NC>2 were
observed  to  develop compensatory changes
such as polycythemia, thoracic kyphosis, and
a lateral flaring of the ribs.  Early evidence of
pulmonary emphysema was observed in dogs
exposed  continuously  to 47.0 mg/m3  (25
ppm) NC>2 for 6 months.
Susceptibility  to   Infection.  Exposure  of
mice, hamsters, and squirrel monkeys to NC>2
causes  increased  susceptibility  to  bacterial
pneumonia and influenza infections, demon-
strated by  increased  mortality,  decreased
survival time, and a reduction  in ability to
clear infectious  agents from the lungs.
 9-20

-------


















CD
O
C
d>


w
'o ?>

C 3
"+-» crt
eo o
«< x"
1— 1 CD
e~
.2
w
t-H
a
o
O
0





m
^_

p


a,
ex


o
'o
SI
CO




o











2
T3
Reversible inhibition of h

^
'g
o
vo
tN




0



^5 &0
S'|
11

















genase activity






















o
c

,2
XJ
o
•eh
O
E
0>
s

•5
1o
NO-hemoglobin and NO-

Os
•a
c
aj
CO
d




o



c/3
cS




OO






T3
^C

X
0)

O
3
T3
1
O
O
"oh
1


>>
T3
















(^1











00
CD
a
No effect on respiratory r
and tidal volumes


M
J3
r~
OS




o
1
V3
00
'5.
CS
u
1
O

















Irreversible inhibition of

£
'g
O
o
o

^


o
o
o"



1 1
if




0












o
CD
cs
C
OO
o
•o































T3
:overe

-------
-s:
S
^M


o
a
1




+J
a
4i
o
0





VI
tj

w






tL>
p
o
D-
X
w
c" -g.
.2 °°
8
§ e
QJ g<
1>
'o
8,
00




in

^
8 >.
G 0
0 "S
"» 2
 G
o ™
 2
Xl 3
'* £
t/a en
£-0
^B
_o
"3
"S,
|
"a>
o
bn

g
.s
c
^o
•^
en
3
^i
5S

^





C3
C)


00
(2 «



in -H ON m o

3 « G "3 1
•S .§ §86^ S-n-SS
^t" x i-i ^ 3 £^ ^n co ^ 5*
^ 'S ^^^  WH *~lCNi3
52 ^^ ^-, -^ -S «> ^^^-^
1 >>£ 1st -§H§ l^i-
t> ^S ^2o«S, § 2 *c3 ^* SJu'g'S)
£> S "S Q2S d 1 3 « fS^&l

O .y'O3ot3 ^S>
O -^H O O "trt 6^ *O *+^
•^5 Q ^1 £X< 2 ^ . QJ CT3 ^— i

S ??&ou'»-iSc'S ^ai^
I &!§•£ o^l ||i
"33 sa.Sg^^^a, S^-°
* 3g|^ ^.g|. |-|

g c '^O|U-S-§e°
.S -B Sj^S'S 8"C"S ^'££.3
6"* C_"*H 'u S* 5J § «3 t5 *j ^ Ssr-iS
coo '-•"a, ^S^^S •C'°Q§ o1^ o rt
5o 3 3 .S^e.s'i (§!ss§. £^z-2

•— •* — (N CM





- -H ~ vo ON
q q p in q
^—i ^H ^—l fT^ l^j

CO
-4-1
co co *7? ai d>
•t-j +-j *O Q u
^ ^ (2 5 5
^^ ^*^ ^*^ F^I ^^i



ro ro










.C°

^ &
2 S |

o -2 *o
2 ° G
'&, "c3 '—
W3 ^Q ^
fl>  >%
gp § £ J ;|
^— ) ^S "O 0

CN q r-
in co r->
s>
E.

J u
3 •*-<
0 S
9-22

-------
^3
o
|

-4-1
CD

-4-* &
•o

s
3

D

So

2
u


3
*^
CD

•|
(D
G
^O
-V^
^2
5








£
r^
r^*
o
CD
3
1 ><
•* -* o
• - ₯
"« .5
Lung:
Respiratory rate unchange
decreased; septal breaks, a
j-
•~t«
^
sfl
rt
13
'<%
2
i->
*c3 x»
§>£
S £
E°
li
«!
ta 13
G" "S-
o «
II
CD
_3
"3
13
"3
'-4-»
T3"
Lung:
Respiratory rate unchange
o
'55
1
X
CD
"o
> G
0 —
Liver:
Cell ballooning with clear
and nuclear displacement;





congestion


*a
c
ca
o
tj
Decreased interferon prod


0
•Z!
o
«s
C
eri
decreased resistance to vin


CD
-*-»
ca
t_i
X
O
ca
Lung:
Marked increased in respir

. ~

1
13
•a
with marked decrease in ti


8
d
ca
=3
8
ca
"o
CD
"3
13
-4-1
O
-4->
O
C8
I-H
•6
"5
si
* S
t-i O
i! '53
« c
e «
edema, and lymphocytic ij
other areas of alveolar exp

"O
|
.S
i
s
2
XI
OD
G
1
13
§*


<4_i
0
^o
i
ca
epithelial surface erosion,
cilia




a
S
Heart:
Interstitial fibrosis and ede
J3
"3
:=)
8
CD
-4-»
T3
C
Kidney:
Glomerular-tuft swelling a
lymphocytic infiltration

-------






•2
^.

2g
CJ
<
£

w
u,
UH
UJ
— 1

ri

ga
•J C3
ARY OF STUDIES OF TOXICO
OF SHORT-TERM NO2 EXPOS
P
Cfl


•o
o>
3
'S
o
o

i
2
5
H








CD
O
1




^j

CD
s
o
u











tS
«
i
g
V)
O
cx

w


ro
. s
G Gift
•2 6
VH
a
u
o
§ 6
U PH


-2
"o
&

















•o
c
ta

_O
r/l
Liver:
Lymphocytic infiltration, conge:















































centrolobular necrosis
























ON





















~
Enzyme systems:
Increased ©2 consumption— livei
t
cd ^-i
tn d
c* '^^
'G i-.
in J3
O fS




*
in
r-




O
o
CA
&P
'5x
ca
O
O




















.s

6 g
P ^
spleen, kidney; increased lactic
dehydrogenase-liver, kidney, se:
creased aldolase— liver, spleen, ki
serum

00


f- "^3 '"^
— *j in


















ro

(N



O 
ca ,_
-*
CA
K


















T3
CD

ca ^

o ca
G CD
'*" (j
mucopolysaccharide production.
phospholipid production, and in
lysozymal enzyme content
























^o

•^ q
CM ca
CD
° "o
-H O
*J
73 t=
CD j""
(H CJ
ca 2
|8

5 *S -a










Blood:
Methemoglobin present


^^







q
od
00




O
O



CA
























Blood:
Methemoglobin present


T**H







q
CN
oo
(N




0


„
.t^ CTJ
^o o
jjQ C
^ 3
OS 60















































CA
op
'S.
9-24

-------
I

V)
O
w

 r<
O
z
o

I

u.
O
Z


z



o


to
o
O
CD


5


I

to
O

>•
OS

 
S





-t_) >-• i x
a £ ha "3
S O ^3 c O
i! lit
<0 e "tr 2 -3
0 42 o ^
g £> oo .S ™
CO --~- O w
S ^ C! £ c
rt 
^ kj

'x-S

73 *
cd g
infection as demonsti
mortality after challe
pneumoniae
>
nP O
f- §• 0
-ft; ^
















m




"3
0
e
o
^

"3
^c
E

'S
Reduction in rate of
3
O
3
e
"S
0

m
(N




ro
' — '

-w
jg
.2







o
J-H

o
o

.s
^
oo
VJ
crease in exposed ver


1
' '















[^

o
00
Approximate doublin
red cell number with
lesser increases in
hematocrit and hemo
globin









Blood:
Polycythemia
3
O
3
U m

oo
ft)




O
oi
CO
CO
*> c
£ 0
c2 e


OS






"J3
c

E
D
*->
C
fli C
Morphological changi
bronchiolar epitheliu


•a
"*"

oo
rn




O
(N

to
§
(2
                                                                                                               9-25

-------
1
 o
 o
O
C
01
t-(— i
0>
&
Comment








o
W
3
1
CX
X
w
CO
§ "^
3 £
ca
C
o
C ~
O P !
cj n,
a.

IU
'o
CD
(X
CXI

CO
m



Possibly pre-
emphysematous lesion.





(4-4
o
V?
o
1
o
00
c
JS
Lung:
Bronchiolar epithelial c
tinuous,
.me
r- -t-»
li

00
ro



q
oi


oo
-t-t
ca
Qi






1 -§.
•i 1°
OJ .„
o  o
00 CO
art
£
01 O
II
.S c3
O
is S
Q a)
1-5 T3
C/l
0
!*
c *
a^

^
>0



q
en
on
.^
^
cd
oi

^0
^^



Amounts increased with
length of exposure






1 c
8 «
c op
-2 3
t/5 (-(
•S"6
s 2 - T3 ?
° ~t7 « ""

T)-
ON



q
wi
on
00
'S<
ca
.g
'3
o

in
W)



Minute volume unchanged.
One of three monkeys




•o
CD
ca
£
o
T3
ca
£
Lung:
Increased respiratory ra
inuous,
li
U (N

Tt
Ov



q
in
on
O>
C
o
S






(N
ca 0
N Z
| £ «>
.r* O ^*
^- ^ cd
0 "S ^
>-» r^ VH
T3 T± •»-»
o /-vj ^^
C/3 v ^ C9
0 w _
5-2!
D > T3








CD
3
"3
•3
S















10
10



1
Minute volume increased.








ob
3
inuous,
0
O

00.
00



o
6
CO
CD*
c
o







Three monkeys exposed
to influenza virus 24 hr
before N02 died
within 3 days






£
"o
"c8
•O
'-M
T3
Both respiratory rate an
increased

0













M5
ro



Associated thoracic








£?
3
inuous,
c
o
o

00
00


in

-------
-



ob
3
„
3
O
3

-4_1
O
U

od
(N

q
in
a
'a
cU
CD
O


1

3
Formation of circulating






>>
^H















possibly lung antibody



















S



Crt
>
N
^




**<
J>
O

CN
00
(N

q
m
a
OH
ca
c
'3
O


c"
(D
"o.
VI
1
c
Increased 62 consumptio
kidney;





















% ft
« X
T3 0)
ir epithelium
_TO
^-1
Inflammation of bronchic
S
„
^

"5: °
"^ C
•a _
in  rf


in


CO
-*-»
IS
1
t-

Animals exposed to
welding fumes, which
contained 3,075 to
6,150 g/m3 (2.5-5.0
ppm) NO as well as other
contaminants; methemo-
globin disappeared during


0
o
S


*•

•o

X
VO
r— (
in
*

O
c^



C/3
d
2
c
03
U
E
E
Methemoglobin formatioi
2.8% and rats, 13.6%
^_J
* X^ S
fe M .0
^ 13 t 03
^TCO ° ^
^^ ^" 'tfl W3
"O >- to ,&•
>n ^ ^ T3








•^
Jg
2
o"
o



O Q>
I 3l'S|l_ 1
I S!i|!| 1.
" <"a,2j?
o

q
^-
Ch


0
m

CO
>_,

i
K
"«
0
o
I
fT*
Increase in volume, with ;



















exposure;
"ation in
4>
2 wk after cessation of th



















0)
transitory epithelial prolil
respiratory bronchioles



















9-27

-------
  In mice, the threshold of increased suscept-
ibility  to  Klebsiella pneumoniae occurred at
6.6  mg/m3 (3.5 ppm) NC>2 for 2 hours, if the
infectious challenge was given within  1 hour
after the  NC>2 exposure. Mice infected with
K. pneumoniae and exposed  up to  72 hours
later to 9.4 or 47.0 mg/m3  (5 or  25 ppm)
NC>2 for 2 hours exhibit similarly enhanced
susceptibility to  the respiratory  infection.
Squirrel monkeys exposed to 18.8 mg/m3  (10
ppm) NC>2 for 2 hours, then  challenged with
K.  pneumoniae  aerosol,  retained the infec-
tious agent in the lungs for extended periods
of time. Two-hour exposure  to 94.0  mg/m3
(50 ppm) NC>2 followed by  infectious chal-
lenge, led to a markedly increased mortality
in squirrel monkeys.
  Continuous  exposure  reduces  the NC>2
threshold  concentration. In long-term  studies
of mice, significantly increased susceptibility
to  infection occurred after continuous daily
exposure to 940 jug/m3 (0.5 ppm) NC>2 for 3
months, or 6- or 18-hour daily exposure for 6
months. A significant increase in susceptibil-
ity  to  influenza virus or K.  pneumoniae  has
also been found in squirrel monkeys  contin-
uously exposed to 18.8 and  9.4 mg/m3  (10
and 5 ppm) NC>2 for 1 and 2 months,  respec-
tively.
  Impairment of  interferon  formation  and
decreased resistance to  viral infection  was
demonstrated  in  rabbits exposed  to 47.0
mg/m3 (25 ppm) NO2 for a 3-hour period.
Systemic  effects.  Inhalation of  NC>2  can
produce systemic effects generally secondary
to  those on the lungs. Monkeys exposed to
28.2 to 94.0 mg/m3  (15 to 50 ppm) NC>2 for
2 hours exhibited cellular changes in heart,
liver, and kidney tissue.  Long-term exposure
of rats to 2.5 to 5.6 mg/m3 (1.3 to 3.0 ppm)
for  15  to  17 months has been associated with
loss of weight and reduced  rates  of weight
gain.
  A circulating  substance,  possibly  a lung
antibody, has been detected  in the blood of
guinea pigs exposed to 9.4 mg/m3 (5.0 ppm)
for  4 hours daily, 5 days per week,  for  5.5
months. Guinea pigs exposed to 28.2  mg/m3
(15 ppm) NC>2 for 10 weeks showed an altera-
tion in  O2 consumption and enzyme activi-
ties (aldolase  and lactic dehydrogenase) in the
serum,  lung,  kidney,  liver,  and  spleen.
Rats and  monkeys  continuously exposed to
3.8 mg/m3 (2.0 ppm)  NC>2 for 3  weeks ex-
hibited marked polycythemia. An increase in
circulating  methemoglobin was  detected in
the blood of several species exposed to con-
centrations greater than 122 mg/m3 (70 ppm)
NC>2 for 1 hour.
Human Studies
   The small amount of information available
concerning the  toxicological effects  of the
oxides of nitrogen in man pertains to levels of
these compounds higher than those normally
found in ambient air. Experimental exposure
of volunteer  subjects to 9.4 mg/m3 (5 ppm)
NC>2 for 10 minutes has produced a substan-
tial, but transient,  increase in airway resist-
ance. Other information derived from occupa-
tional exposure to  higher  concentrations of
NO/NO2  mixtures  is  complicated  by  the
presence of other pollutants. Impaired pulmo-
nary function, indicated by reduced maximal
breathing  capacity,  increased  expiratory  re-
sistance, and  occasional decreased vital capac-
ity, has been  observed in patients accidentally
exposed  to  high concentrations  of nitrous
fumes for a few minutes. In some cases, the
impairment has lasted  for more than 2 years
after the incident. Occupational exposure to
169 mg/m3  (90 ppm) NO2 for 30 minutes
produced  pulmonary edema and decreased
vital capacity  18 hours  later.  Exposure to
very high concentrations for about 5 minutes
has produced death  within 2 days to 5 weeks.
   The human  threshold for perceiving the
odor of NO2 appears to be about 225 jug/m3
(0.12 ppm).

G.   REFERENCES
1.  Gray, E. Le B. Oxides of Nitrogen, Their Oc-
    currence, Toxicity, Hazard. AMA Arch, of Ind.
    Health, 19: 479-586, 1959.
2.  Flury, F. and F. Zernik. Schadliche Gas. Berlin,
    Sprunger, 1931.
3.  Murphy, S.  D.,  et  al.  Altered  Function in
    Animals Inhaling Low Concentrations of Ozone
9-28

-------
    and Nitrogen Dioxide. Amer. Ind. Hyg. Ass. J.,
    25: 246-253, 1964.
4.  Paul, W. D. and C. R. Kemp. Methemoglobin: A
    Normal Constituent of Blood. Proc. Soc.  Exp.
    Biol. Med., 56: 144-196, 1951.
5.  Bodansky,  M.  Methemoglobinemia  and
    Methemoglobin-Producing  Compounds. Pharma-
    col. Rev, 3: 144-196, 1951.
6.  Flury,  F.  The  Role of Nitrogen Monoxide in
    Poisoning  from  Nitrogen  Fumes.  Arch.  Exp.
    Pathol. Pharmakol., 157: 104-106, 1930.
7.  Pflesser, G.  The Importance of  Nitrogen Mon-
    oxide  in Poisoning by Nitrous  Oxide Fumes.
    Arch. Exp.  Pathol.  Pharmakol.,  179:  545-553,
    1935.
8.  Petr,'B. and P.  Schmidt.  The Influence of an
    Atmosphere Contaminated with  Sulfur Dioxide
    and Nitrous Gases on the Health  of Children. Z.
    Gesamte Hyg.  Grenzgeb., 13:  34-48,  January
    1967.
9.  Sancier, K.  M.,  G.  Freeman,  and  J.  S. Mills.
    Electron  Spin  Resonance  of  Nitric Oxide-
    Hemoglobin  Complexes in  Solutions.  Science,
    137: 752-754, 1962.
10.  Krasna,  I. and D. Rittenberg. The Inhibition of
    Hydrogenase by Nitric Oxide. Proc.  Nat. Acad.
    Sci., 40: 225-227, 1954.
11.  Air Quality  Criteria for Photochemical Oxidants.
    National Air Pollution Control Administration,
    Washington,  D.  C. Publication  Number AP-63,
    March 1970.
12.  Gray, E. Le B., et al. Toxicity  of the Oxides of
    Nitrogen. AMA Arch. Ind.  Hyg. Occup. Med.,
    10: 423-425, November 1954.
13.  La Towsky,  L. W., E. L. MacQuiddy,  and J. P.
    Tollman. Toxicology  of  Oxides of Nitrogen:
    Toxic Concentrations. J. Ind. Hyg. Toxicol., 23:
    129-133, 1941.
14.  Gray,  E.  Le  B. Toxicity of  the  Oxides  of
    Nitrogen   II.  Acute Inhalation Toxicity  of
    Nitrogen Dioxide, Red Fuming Nitric Acid and
    White Fuming Nitric Acid. AMA Arch. Ind. Hyg.
    Occup. Med., 10: 418-422, November 1954.
15.  Henry,  M.  C., R. Ehrlich, and W. H. Blair. Effect
    of Nitrogen  Dioxide on Resistance  of Squirrel
    Monkeys to  Klebsiella pneumoniae  Infection.
    Arch. Environ. Health, 18: 580-587, 1969.
16.  Balchum, O. J.,  et  al.  Nitrogen Dioxide  In-
    halation and Lung  Antibodies. Arch.  Environ.
    Health, 10: 274-277, 1965.
17.  Wagner, W.  D.,  et  al.  Experimental Study  of
    Threshold Limit of NO2- Arch. Environ. Health,
    10: 455-466, 1965.
18.  Vaughan, T.  R., L. F. Jennelle,  and T. R. Lewis.
    Long-Term Exposure to Low Levels of Air Pol-
    kitants.  Effects  on Pulmonary  Function in  the
    Beagle.  Arch. Environ. Health, 19: 45-50, 1969.
 19. Buell, G. C., Y. Tokiwa, and P. K. Mueller. Lung
    Collagen and Elastin  Denaturation  in  Vivo
    Following Inhalation of Nitrogen Dioxide. Cali-
    fornia State Dept. of Public Health. (Presented at
    the  59th Annual  Air  Pollution  Control Asso-
    ciation Meeting. San Francisco. Paper No. 66-7,
    June 1966.)
20. Mueller,  P. K. and M. Hitchcock. Air Quality
    Criteria-Toxicological  Appraisal  for   Oxidants,
    Nitrogen  Oxides,  and  Hydrocarbons.  J.  Air
    Pollut. Contr. Ass., 19: 670-676, 1969.
21. Thomas,  H. V., P. K. Mueller, and R.  L. Lyman.
    Lipoperoxidation of  Lung  Lipids  in  Rats
    Exposed  to Nitrogen  Dioxide.  Science,  759:
    532-534, 1968.
22. Roehm, J. N., J. G. Hadley, and  D. B. Menzel.
    Antioxidants versus Lung Disease. (Presented at
    the 10th Annual Hanford Biology Symposium.
    Richland, Washington, June 2-5, 1970).
23. Kleinerman, J.  and G. W. Wright. The Reparative
    Capacity  of Animal  Lungs After Exposure  to
    Various  Single and Multiple  Doses of Nitrite.
    Amer. Rev. Resp. Dis., 83: 423-424, 1961.
24. Kleinerman, J.  and G. W. Wright. The Effects of
    Prolonged and Repeated Nitrogen Dioxide In-
    halation on the Lungs of Rabbits and  Guinea
    Pigs. (Read Before the Fifth  Annual Air Pol-
    lution Medical Research  Conference  California
    State Department of Public Health. Los Angeles.
    December 1961.)
25. Thomas,  H. V., P. K.  Mueller, and G. Wright.
    Response of Rat  Lung Mast Cells to Nitrogen
    Dioxide Inhalation. J. Air Pollut. Contr. Ass., 17:
    33-35, 1967.
26. Freeman,  G.  Discussion: Ultrastructural  Early
    Lesion Induced in  the Rat Lung by NO 2 in In-
    halation  Carcinogenesis. (Proc. of Oak  Ridge
    National   Laboratory  Conference. Gatlinburg.
    October  8-11,  1969.)  AEC  Symp.  Ser., 18:
    267-269,  1970.
27. Freeman, G., et al. Lesion of the Lung in Rats
    Continuously Exposed to Two  Parts per Million
    of Nitrogen Dioxide. Arch. Environ. Health, 17:
    181-192,  1968.
28. Stephens, R. J., et al. Ultrastructural Changes in
    the  Terminal  Bronchiole  of  the  Rat  During
    Continuous Low-Level Exposure to  Nitrogen
    Dioxide. Exp. Mol. Pathol., In Press, 1970.
29. Evans, M. J.,  R. J. Stephens, and G. Freeman.
    Effects of Subacute Concentrations of NO2 on a
    Cell  Renewal  in the  Lung.  (Presented  at  10th
    Annual Hanford Biology Symposium. Richland,
    Washington June  2-5,  1970.)  Arch,  of Intern.
    Med., In Press,  1970.
30. Kleinerman, J.  and C.  R. Cowdrey. The Effects
    of Continuous  High Level Nitrogen Dioxide on
    Hamsters. Yale J. Biol. Med. 40: 579-585, 1968.
                                                                                                9-29

-------
31.  Kleinerman, J.  Effects  of NC>2  in  Hamsters:
    Autoradiographic  and  Electron  Microscopic
    Aspects.   (Inhalation  Carcinogenesis,  AEC
    Symposium  Series  18.  Gatlinburg.  CONF-69
    1001. October 1969.) 10 pages. National Tech-
    nical Information Service. Springfield, Virginia,
    April 1970.271-281.
32.  Blair, W.  H., M.  C. Henry,  and R.  Ehrlich.
    Chronic Toxicity of Nitrogen Dioxide: II. Effects
    on   Histopathology of  Lung  Tissue.  Arch.
    Environ. Health, 18: 186-192, 1969.
33.  Freeman,  G. and G. B. Haydon. Emphysema
    After  Low-Level  Exposure  to  NO2-  Arch.
    Environ. Health, 8: 125-128, 1964.
34.  Freeman, G.  B.,  et al. Environmental Factors in
    Emphysema and  a Model System with NC^. Yale
    J. Biol. Med., 40: 566-575, 1968.
35.  Haydon, G. B., G. Freeman, and N. H. Furiosi.
    Covert  Pathogenesis  of  NC<2  Induced
    Emphysema in the Rat. Arch.  Environ. Health,
    11: 776-783, 1965.
36. Freeman,  G., S.  C.  Crane, and R. J. Stephens.
    Pathogenesis of  the  Nitrogen Dioxide - Induced
    Lesion in  the Rat Lung: A Review and Presenta-
    tion of New Observations. Am. Rev.  Resp. Dis.,
    98: 429-443, 1968.
37. Freeman,   G., et  al.  The  Subacute Nitrogen
    Dioxide -  Induced Lesion of the Rat Lung. Arch.
    Environ. Health,  18:  609-612, 1969.
38. Freeman,  G., S. C. Crane, and  N.  J. Furiosi.
    Healing in  Rat Lung after Subacute Exposure to
    Nitrogen Dioxide. Amer. Rev. Resp. Dis., 100:
    662-676, 1969.
39. Stephens,  R. J., G.  Freeman,  and M. J.  Evans.
    Connective  Tissue  Changes in Lungs  of Rats
    Exposed  to  NO2  (Presented  at  10th  annual
    Hanford Biology Symposium. Richland, Wash-
    ington.  June 2-5, 1970.) Arch, of Intern. Med.,
    In Press, 1970.
40. Riddick, J. H., K. I. Campbell, and D. L. Coffin.
    Histopathological  Changes  Secondary  to
    Nitrogen  Dioxide   Exposure  to  Dog  Lungs.
    (Presented at the Annual  Scientific Assembly,
    American  Society of  Clinical  Pathologists.
    Chicago. September 22-30, 1967.)
41. Riddick, J. H., K. I. Campbell, and D. L. Coffin.
    The Effects  of  Ferric  Oxide Dust Exposure in
    Nitrogen  Dioxide-Impaired Dog  Lungs. (Pre-
    sented  at the Annual Scientific Assembly, Ameri-
    can  Society  of  Clinical  Pathologists.  Miami
    Beach.  October 11-19, 1968.)
42. Henry,  M. C., and C. Aranyi. Scanning Electron
    Microscopic Observations of  the   Effects  of
    Atmospheric Pollutants.  Arch. Environ. Health,
    To Be Published, 1970.
43. Ehrlich, R. and M. C. Henry. Chronic Toxicity of
    Nitrogen Dioxide: I. Effects on  Resistance  to
    Bacterial Pneumonia. Arch. Environ. Health, 17:
    860-865, 1968.
44. Mitina,  L.  S.  The Combined  Effects of Small
    Concentrations of Nitrogen Dioxide and  Sulfur
    Dioxide Upon  an  Organism. Gig.  Sanit., 27: 3-8,
    October 1962.
45. Carson, T.  R., et  al. The Responses of Animals
    Inhaling Nitrogen  Dioxide for Single, Short-Term
    Exposures.   Aeronautical  Systems Div.,  Tech.
    Report 61-525. October 1961. Amer. Ind. Hyg.
    Ass. J. 23: 457-462, 1962.
46. MacQuiddy, E. L., et al. Toxicology of Oxides of
    Nitrogen: II. Physiological Effects and Sympto-
    matology. J. Ind.  Hyg.  Toxicol., 23: 134-140,
    1941.
47. McCord, C. P., G. C. Harrold, and S. F. Meek. A
    Chemical  and Physiological  Investigation  of
    Electric Arc Welding: III. Coated Welding Rods.
    J. Ind. Hyg. Toxicol., 23: 200-215, 1941.
48. Pryor, W. H., H. L. Bitter,  and R. J. Bertler. The
    Effect of Nitrogen Dioxide-Nitrogen  Tetroxide
    on Oxyhemoglobin Dissociation. SAM-TR-67-33.
    U. S.  A. F.  Sch. Aerosp. Med., 1-5, April 1967.
49. Buckley, R. D., and  O. J.  Balchum.  Acute and
    Chronic Exposures to Nitrogen Dioxide Effects
    on  Oxygen  Consumption and  Enzyme Activity
    on  Guinea  Pig Tissues. Arch.  Environ. Health,
    10: 220-223, 1965.
50. Ehrlich, R. Effect  of Nitrogen Dioxide on Resist-
    ance to Respiratory Infection. Bacterial Rev., 30:
    604-614, 1966.
51. Gardner, D. E., R. C. Holyman, and D.  L. Coffin.
    Effects of Nitrogen Dioxide on Pulmonary Cell
    Population.  J.  Bacteriol. 98:  1041-1043,  1969.
52. Ehrlich, R. Effect of Air Pollutants on Respira-
    tory   Infection.  Arch.   Environ.  Health,  6:
    638-742, 1963.
53. Purvis, M. R. and  R. Ehrlich. Effect  of Atmos-
    pheric Pollutants  on Susceptibility to Respira-
    tory Infection: II. Effect of Nitrogen Dioxide. J.
    Infec. Dis., 113: 72-75, 1963.
54. Ehrlich, R.,  M. C. Henry,  and J. Fenters. Influ-
    ence  of Nitrogen  Dioxide on  Resistance  to
    Respiratory  Infections.  (Inhalation  Carcino-
    genesis.  AEC   Symposium Series 18.  CONF-
    691001.  Gatlinburg. October  1969.)  14  pages.
    National Technical Information Service. Spring-
    field,  Virginia.  April 1970. 243-257.
55. Henry, M. C., et al. Chronic Toxicity  of NO2 in
    Squirrel  Monkeys III.  Effect  of Resistance to
    Bacterial and  Viral  Infection. Arch.  Environ.
    Health, 20: 566-570, 1970.
56. Fenters, J., et  al.  Chronic  Toxicity of Nitrogen
    Dioxide. IV.  Serologic Response in Squirrel
    Monkeys Exposed to 5 ppm NO2 and Influenza
9-30

-------
    Virus. Arch. Environ. Health, To Be Published,
    1970.
57. Ehrlich, R., et al.  Effects of Nitrogen Dioxide
    and Tobacco Smoke on  Retention  of  Inhaled
    Bacteria.  (Third  International Symposium  on
    Inhaled  Particles.  London.  September  16-23,
    1970.) pp. 7.6.1-7.6.4.
58. Valand, S.  B., J. D. Acton,  and Q.  N. Myrvik.
    Nitrogen  Dioxide Inhibition  of Viral Induced
    Resistance  in  Alveolar  Monocytes.  Arch.
    Environ. Health, 20: 303-309, 1970.
59. Henschler,  D., et  al. Olfactory Threshold of
    Some Important Irritant  Gases and Manifesta-
    tions in  Man by  How  Concentrations. Arch.
    Gewerbepathol.  Gewerbehyg.,  Berlin,  17:
    547-570,  1960.
60. Shalamberidze, O. P. Reflex Effects of Mixtures
    of Sulfur and Nitrogen Dioxides. Hyg. Sanit., 32:
    10-14, 1967.
61. Vogh,  J.  W.  Nature  of  Odor Components in
    Diesel Exhaust. J.  Air Pollut.  Contr. Ass., 19:
    773-777,  1969.
62. Rokaw, S. N., et  al. Human Exposures to Single
    Pollutants - NO2 in a Controlled  Environment
    Facility. (Preprint of presentation  at the Ninth
    AMA Air Pollution Medical  Research Confer-
    ence. Denver. July 24, 1968.)
63. Abe, M.  Effects  of Mixed NO2 - SO2  Gas on
    Human Pulmonary Functions. Bull. Tokyo Med.
    Dent. Univ., 14: 415-433, 1967.
64. Nadel,  J.  A., et  al. Mechanism of Bronchocon-
    striction During Inhalation of Sulfur Dioxide. J.
    Appl. Physiol., 20: 164-167, 1965.
65. Von Oettingen, W. F. The Toxicity and Pollution
    Dangers of  Nitrogen Fumes. U. S. Public Health
    Service. Washington, D. C. Public Health Bulletin
    No. 272. 1941.34 pages.
66. Grayson, R.  R.  Silage Gas Poisoning: Nitrogen
    Dioxide Penumonia, a New Disease in Agricul-
    tural Workers. Ann. Intern. Med., 45: 393-408,
    1956.
67. Lowry,  T.  and  L.  M.  Schuman.  Silo-Filler's
    Disease.  A Syndrome  Caused  by  Nitrogen
    Dioxide. J.  Amer. Med. Ass., 162:  153-160,
    1956.
68. Adley, F. E. Exposures  to Oxides of Nitrogen
    Accompanying  Shrinking Operations.  J.  Ind.
    Hyg. Toxicol., 28: 17-20,  1946.
69. Norwood, W. D., et al. Nitrogen Dioxide Poison-
    ing Due to  Metal  Cutting with  Oxyacetylene
    Torch. J. Occup. Med., 8:  201-305, 1966.
70. Harrold, G. C., S. F. Meek, and C. P. McCord. A
    Chemical and  Physiological  Investigation of
    Electric   Arc Welding: Bare, Washed Welding
    Rods. J. Ind. Hyg. Toxicol., 22: 347-378, 1940.
71. Becklake, M. R.,  et al. The Long-Term Effects of
    Exposure  to Nitrogen  Fumes.  Amer.   Rev.
    Tuberc. Pulm. Dis.,  76: 398-409, 1957.
72. Vigdortschik, N. A., et al. The Symptomatology
    of Chronic Poisoning with Oxides of Nitrogen. J.
    Ind. Hyg. Toxicol., 19: 469-473,  1937.
73. Vigliana, B. C. and  N.  Zurlo. Experiences of the
    Clinical  del  Lavoro with Maximum Allowable
    Concentrations  of Industrial  Poisons.  Arch.
    Gewerbepathol.  Gewerbehyg., 13:   528-535,
    1955.
74. Goodman, L.  S.  and  A. Gelman (eds.).  The
    Pharmacological  Basis of  Therapeutics.   New
    York: Macmillan, 1965.
75. Freeman, G., N.  J.  Furiosi, and G. B. Haydon.
    Effects of Continuous Exposure to 0.8 ppm NO2
    on Respiration of Rats. Arch. Environ. Health,
    13: 454-456, 1966.
                                                                                               9-31

-------

-------
                                   CHAPTER 10.
                    EPIDEMIOLOGICAL APPRAISAL
                           OF NITROGEIN  OXIDES
 A.   INTRODUCTION
   In contrast to the large volume of toxico-
 logical  data,  epidemiological  data  on  the
 effects of exposure to nitrogen  oxides (NOX)
 are meager. Epidemiological studies have been
 primarily  concerned  with  nitrogen  dioxide
 (NC>2) because it is the oxide of nitrogen that
 has been most closely associated with effects
 on  health. Generalizations  drawn from  the
 toxicological  data in chapter 9 form the basis
 for several of the epidemiology studies pre-
 sented in  this chapter; that  is to say, NO2
 exerts its  primary toxic effect on the lungs
 where, in  animals, exposure can be associated
 with  increased susceptibility to  respiratory
 infection and emphysematous changes.
  Exposure  to  other  pollutants,  such  as
 ozone, SC>2,  metal fumes, and  aerosols, pre-
 sents a major problem in the interpretation of
 all  information  collected.   The  studies  re-
 ported here have incorporated some measures
 to minimize interference by these factors.
 B.   EPIDEMIOLOGIC EVIDENCE OF
     LONG TERM EFFECTS
 1.   The Cleveland Clinic Study1 -
     An Accidental High Exposure
  On May 15, 1929,  the X-ray room of the
 Cleveland  Clinic  was  engulfed  in  flames;
 50,000 nitrocellulose films  were ignited; and
 three explosions occurred. The  air was filled
 with  concentrations  of  nitric  oxide  (NO),
 carbon monoxide (CO), and hydrocyanic acid
 (HCN) estimated at 63.3 g/m3 (51,500 ppm),
 45.9 g/m3 (40,000 ppm), and 6 g/m3 (5,400
 ppm),  respectively. Ninety-seven  individuals
 succumbed within 2 hours.  Neither CO nor
 HCN  was  considered  the primary cause  of
death in the  26 additional  individuals who
died in less than 1 month, nor were either CO
or HCN  implicated in 92 nonfatal cases of
serious physical injury.
  Much  later, an epidemiological study was
conducted  to determine whether survivors of
the disaster incurred  a risk  of dying sooner
than nonexposed individuals.
  Study  groups consisted of individuals who:
     1. Were present in the building at the
       time of the explosion.
     2. Entered the building  that afternoon.
     3. Were exposed to smoke in an adjacent
       building.
     4. Assisted rescue and first-aid workers.
  An estimated 98 to 99 percent of the indi-
viduals present at the fire were enumerated by
searching records and 87 percent  of  these
were actually studied. Individuals,  who were
present at the fire, but who were not in any
of the four exposure groups were used as
controls.
  When  cumulative  observed survival  rates
and  cumulative expected survival rates were
compared,  their  ratios,  calculated for the
years 1929 to  1965, showed no statistically
significant differences between mortality rates
of exposed and nonexposed groups.
2.   The Chattanooga  Studies - Ambient
     NO£ and Respiratory Illness
a.   Community effects
  Shy, et al.2'3 studied the effects of com-
munity exposure  to nitrogen dioxide in four
residential areas in greater Chattanooga. One
area, in close proximity to a large TNT plant,
had high-NO2  and low-particulate exposure.
Another  had high-suspended particulate and
low-NO2  exposure.  The  two other  areas
served as  "clean" controls.
                                          10-1

-------
(1.)  Pollutant  monitoring.  Each  area  con-
tained three elementary schools. In the high-
NC>2  area  surrounding  the  TNT plant,
monitoring  sites  were selected  to  represent
the  average pollution level near each of the
three schools. Sites for pollution monitoring
stations  in  the  high-particulate  and  the
control  areas were  selected  to  represent
average pollutant exposure  on an  area-wide
basis.  The  effects of  wind,  topography,
vegetation, and local pollution sources were
considered in selecting a site.
   Twenty-four-hour total-suspended-particu-
late  matter,  suspended  nitrates,  suspended
sulfates, and gaseous nitrogen dioxide concen-
trations were measured daily at  each site, in
November  1968 and November  1969.  From
December 1968 through February  1969, and
again in April  1969, these 24-hour concentra-
tions were  measured once every 4 days on a
systematic  sampling  schedule (Table  10-1).
Twenty-four-hour sulfur  dioxide concentra-
tions were  measured once weekly in Novem-
ber  1968,  but measurements  were discon-
tinued when concentrations were found to be
less than 2.8 jug/m^ (0.015 ppm) at each site.
  Integrated  24-hour NC>2  samples  were
collected  by  gas-bubbler  techniques and
Table 10-1. ARITHMETIC MEAN AND 90TH PERCENTILE CONCENTRATIONS OF POLLUTANTS SAMPLED
                                 FOR 24 HOURS AT VARIOUS SITES2
Pollutant
NO2, ppm


Suspended
nitrate,
Mg/m3
Suspended
sulfate,
Mg/m3
Total
suspended
particulates,
Mg/m3
Soiling
index,c
Coh/1000
lineal ft
Level
of
exposure
Meana
90 percentile"
Standard deviation
Mean
90 percentile
Standard deviation
Mean
90 percentile
Standard deviation
Mean
90 percentile
Standard deviation

Mean
90 percentile
Standard deviation

High NO2
School
1
0.109
0.242
0.098
7.2
14.8
9.1
13.2
22.6
6.8
96
183
63

0.80
1.46
0.51

School
2
0.078
0.141
0.054
6.3
13.4
5.8
11.4
19.2
6.4
83
138
46

0.89
1.73
0.64

School
3
0.062
0.098
0.040
3.8
8.0
4.6
10.0
19.5
4.9
63
108
42

0.91
1.84
0.68

High
parti-
culate
0.055
0.087
0.024
2.4
4.6
1.7
10.7
17.3
4.6
99
181
58

2.09
4.37
1.67

Control
1
0.063
0.096
0.030
2.6
5.9
2.6
9.8
15.8
4.5
72
128
45

1.39
3.29
1.20

Control
2
0.043
0.069
0.021
1.6
3.1
1.0
10.0
15.6
4.5
62
112
35

1.23
2.53
0.89

aMean = arithmetic mean of all samples collected.
"90 f ercentile = concentration exceeded by only 10 percent of samples.
c4-h9ur measurements.
10-2

-------
analyzed by the Jacobs-Hochheiser method.
Total-suspended-particulate  matter,  nitrate,
and sulfate were collected with a high-volume
sampler and subsequently, analyzed.
(2.) Health-effect  evaluation.   Two  possible
health effects  of NO2  exposure  were  in-
vestigated  in  this  community during  the
1968-69 school year: (1) impaired ventilatory
function in elementary school children  and
(2) increased frequency  of acute respiratory
illness  in  family groups. A  total  of  987
second-grade school  children  participated in
the ventilatory  testing.  Weekly  ventilatory
function (Forced Expiratory Volume in 0.75
seconds, FEVg 75) of each child was adjusted
for differences in individual standing height,
for boys and girls, separately. These tests were
made  during  November 1968 and March
1969.
   The  socioeconomic  factors  considered
were: house value or rent, education of the
head of the household,  and  the number of
people in  a household (crowding).  In  each
case, the high-NC^—low-particulate exposure
area exibited the highest socioeconomic level,
followed in rank by  control areas. The low-
NC>2—high-particulate  exposure area had  a
distinctly  lower socioeconomic level. Home
cigarette smoking in  the control and high-
NC>2  areas  differed  only by  1 percent. No
differences in the duration of residence at the
current address were  found  in any of the
areas.
   Four  ventilatory  tests  with Stead-Wells
 spirometers were made weekly  in November
 1968 and again in March 1969. Analysis of
 variance was used to determine  the statistical
 significance of factors such as sex, month of
 test,  study  area,  schools  within  high-NC>2
 area, and the concentration of NC>2 on day of
 test (see Tables 10-2 and 10-3).
   At the beginning of the study, all members
 of the household of each participating second-
 grade child were asked  to volunteer for the
 second  phase of the study —  a prospective
 study,  designed to  assess  the  frequency of
 acute respiratory  disease. A total of 4,043
 individuals  in  871  families participated. At
 bi-weekly  intervals from  November  1968
 through April 1969, each family was asked to
 report  the  incidence  of new colds  or  sore
 throats,  in terms of the age and sex of sick
 household members and the severity of the
 illness. Severity indices included: the presence
 of fever, length of home confinement, and
 consultation with a doctor for treatment.
   Area differences in respiratory illness rates
 were analyzed in the following groupings: the
 entire 24-week period,  the influenza-A inter-
 val, the  interim interval,  and the influenza-B
 interval.
(3.)  Conclusions.  The  ventilatory  per-
formance of second-grade school children in
the high-NO2 exposure area was significantly
lower than  performance  of children in  the
control areas.
    Table 10-2. ANALYSIS OF VARIANCE OF EFFECTS ON HEIGHT-AD JUSTED FEV0 75 OF SEX
                     OF CHILD, MONTH OF TESTING, AND STUDY AREA 2

Factor
Sex of child
Month of test
Study area
High-NC>2 vs. Controls 1 and 2
High-TSPa vs. Controls 1 and 2
Control 1 vs. Control 2
Mean
square
0.3852
0.0318

0.0098
0.001 1
0.0004
F
value
189.3
15.6

4.8
0.6
0.2
Probability of
significant difference
p<0.01
p<0.01

p<0.05
not significant
not significant
aTotal suspended particulate matter.
                                                                                     10-3

-------
        Table 10-3. ANALYSIS OF VARIANCE OF EFFECTS ON HEIGHT-ADJUSTED FEV0 75
         OF NO2 CONCENTRATIONS ON DAY OF TEST, SCHOOL WITHIN HIGH-NO2 AREA,
                          SEX OF CHILD, AND MONTH OF TEST2
Factor
Sex of child
Month of test
NO2 on day of test
Schools within high-N02 area
Schools 1 vs. 2 and 3
School 3 vs. 1 and 2
School 1 vs. 2
Mean
square
0.0952
0.0178
0.0009

0.0019
0.0007
0.0030
F
value
90.3
16.9
0.9

1.8
0.6
2.8
Probability of
significant difference
p<0.01
p<0.01
not significant

not significant
not significant
not significant
  Illness-incidence  rates for each  family
segment in the  high-NO2 area were consist-
ently and  significantly higher than indicence
rates in the two  control areas throughout the
entire study period.  This increased incidence
of acute  respiratory disease was observed
when the  24-hour NC>2 concentration, meas-
ured over a 6-month  period, was between  117
and 205 jug/m3 (0.062 and 0.109 ppm) and
the mean suspended nitrate level was 3.8
Mg/m3 or greater (Tables 10-1 and 10-4).
  School 3 in the high-NO2/low-particulate
area had consistently higher  incidence rates
than the  control  areas,  even though the
average NC>2 level was the same as in control
1  (Table  10-4). The  NC>2  and  suspended
    Table 10-4. AVERAGE BIWEEKLY RESPIRATORY ILLNESS RATES PER 100 FOR EACH FAMILY
               SEGMENT ACCORDING TO EXPOSURE TO OXIDES OF NITROGEN3
Rank of Population
by N02 exposure


Average
24-hour
NO2, ppm
0.109

0.078

0.062

0.063
0.043
Average
24-hour
suspended
nitrate,
Mg/m3
7.2

6.3

3.8

2.6
1.6





Study
population
School 1
(high-N02)
School 2
(high-N02)
School 3
(high-N02)
Control 1
Control 2



Family segment


All family
members
17.7

17.5

16.3

13.9
15.0

Second
graders
23.4

23.4

20.4

18.0
20.1


Siblings
19.9

18.0

19.1

15.6
17.0


Mothers
15.3

14.4

13.4

11.8
12.3


Fathers
11.0

12.8

12.1

8.8
9.6
10-4

-------
nitrate levels at school 3 were, however, more
variable than control  area levels.  This  varia-
bility was  associated with higher  peak-levels
of these pollutants (Table 10-1).
   Illness-incidence  rates peaked during the
A2/Hong Kong-influenza epidemic and again
during  an  influenza-B  outbreak.  Families
residing in the two polluted areas reported a
consistent  excess of respiratory illness over
the control area, particularly  during the A2
epidemic  and  the  interim period  between
epidemics.
   Control-2 area appeared to  be hard hit by
the influenza-B  epidemic.  During  this  time,
illness  rates  among  second  graders  and
mothers in the Control-2  area were equal to
rates of comparable family segments in the
high-NO2 area. Rates in Control-1  area, how-
ever, remained well below those of the high-
NC>2 area in  all intervals  and in  all family
segments.
  In the high-NO2 area and in the high-partic-
ulate area excess illness rates occurred  in all
family  segments: second-grade  children,
siblings, fathers,  and mothers.  The  relative
excess was  18.3 percent in  the high-NO2 area
and 10.4 percent in the high-particulate area
(Table 10-5). Area differences in illness rates
could  not  be  explained  by  difference  in
family composition,  economic  level,  demo-
graphic  characteristics,  or  prevalence of
chronic conditions. Parental smoking  habits
did not appear to influence respiratory illness
rates  in second-grade school children.  Expo-
sure to NC<2  and to suspended particulates
appear to be the most  probable explanation
for the observed excess in respiratory illness
rates  for  the population. Severity  of  illness
was essentially the same among all  the areas;
however,  a  precise NC>2  dose-response rela-
tionship cannot be established from the data.

b.   Acute Lower Respiratory Illness
     in Children
   Later, Pearlman, et al.4  made a  retrospec-
tive  study of acute lower respiratory  illness
among infant cohorts and  first and second
graders living in three of  the  Chattanooga
neighborhoods  previously  described,  (all,
except the  high-particulate area).  Pearlman
felt, on the basis of aerometric  observations,
that one of the control areas (Control 1) in
the Shy study  could  properly be designated
an intermediate-NC»2-exposure area.  The high-
                 Table 10-5. PERCENT RELATIVE EXCESS* OF RESPIRATORY
                  ILLNESS AMONG FAMILY SEGMENTS IN EXPOSED VERSUS
                      CONTROL AREAS DURING 24 WEEKS OF STUDY3
Family segment
All family members
Second graders
Siblings
Mothers
Fathers
Study areas
High-N02
18.8
16.8
16.0
18.3
31.5
High-particulate
10.4
12.6
-0.06
34.2
22.8
                  a /-Illness rate of exposed group   _ n
                     Illness rate of control groups
                                                                                     10-5

-------
NC>2 exposure area and the other control area
(Control 2)  completed  the  pollutant-dose
gradient for the Pearlman study.
  Parents were asked to complete a question-
naire  for the study-children, reporting  the
frequency with which  a physician had diag-
nosed  croup,  bronchitis,  pneumonia,  and
asthma, during the 3-year period  from  July
1966  through June  1969. Completed ques-
tionnaries were returned by 95 percent of the
1820 school  children and 84 percent of the
1311 infants still residing in the study areas.
Questionnaire responses  were  validated by
physicians'  office and hospital records. Sensi-
tivity and specificity exceeded 67 percent for
each clinical  diagnosis. Parents could usually
approximate  the  number of episodes of each
illness  but  were  often unable  to  date each,
precisely, within the 3-year study period.
  About 7  percent of the respondents in each
area had a history of asthma  and were ex-
cluded from  the  main analysis. There were
too few asthmatics in  each  dose-duration
category to  permit  firm conclusions about
this subpopulation, but excess bronchitis was
noted  among the ones  residing in the high-
exposure area.
  Exposure to intermediate and high levels of
NC>2 was associated with a significant increase
in acute  bronchitis among the infant cohort,
exposed  for 3 years, and  school  children,
exposed for 2 and 3  years.  This greater fre-
quency  of  acute bronchitis  was  observed
when the  mean 24-hour NC<2 concentration,
measured over a 6-month period, was between
118  and 156 Mg/m3 (0.063  and 0.083 ppm)
(Table 10-6).
  Pearlman  suggested a threshold relationship
for NC>2 exposure  and bronchitis  since the
morbidity excess followed the pollutant gradi-
ent only among school children exposed for 3
years. Area  differences  in total acute lower
respiratory illness could be accounted for by
bronchitis  excess.  Croup,  pneumonia,  and
hospitalizations for acute lower respiratory ill-
ness did not show significant area differences.
  Since the high-NO2 exposure area of this
study included  a school with  NC<2 levels
about the same as the levels in the intermedi-
ate  area,  the  difference in  NC>2  exposure
levels between the high and intermediate areas
was  diminished. This may explain why the ill-
ness rates of  some  intermediate  exposure
groups exceeded  the rates of high exposure
groups.  Groups within each  school  of the
high-NO2 area were  too  small, however, to
allow separate analysis of illness rates.
  The Pearlman study replicated Shy's obser-
vation that excess acute respiratory illness can
be found  among  children living in  NC>2-pol-
            Table 10-6. DISTRIBUTION OF CHILDREN REPORTING ONE OR MORE
                  EPISODES OF BRONCHITIS BY LENGTH OF EXPOSURE4

6-month mean NC>2
High-N02,
156 Mg/m3 (0.083 ppm)
Intermediate-N02,
118 Mg/m3 (0.063 ppm)
Low-NO2 (control),
81 Mg/m^ (0.043 ppm)
School children
exposure, yr
1
20.9
31.6
25.1
2
34 .7a
45 .5a
20.3
3
32 .2a
31.2a
23.2
Infant exposure,
yr
1
33.3
26.2
21.1
2
37.5
29.5
34.0
3
46.8a
50.5a
36.3
        aDiffers significantly from low-N02(control) area.
10-6

-------
luted  areas. He  suggested that acute lower-
respiratory illness may be  a  more selective,
sensitive  indicator of pollutant effect upon
host response than is acute general respiratory
illness.

3.  Czechoslovakia!! Study - Ambient
    N(>2 and SC>2 on Peripheral Blood
  Studies of Czechoslovakian  children5  who
lived near a factory  that emits NOX and SC>2
have produced some positive information not
usually sought in the United States. Data are
available  from three towns: (1) Ohrazenice,
which experiences  air  pollution predomi-
nately from SC>2  [concentrations  ranging
from  30  to 320 Mg/m3 (0.01 to 0.12 ppm)],
with  slight  admixtures of  NOX  [concentra-
tions  ranging from 5  to  50  Mg/m3*]; (2)
Rosice, which experiences a larger amount of
pollution  by NOX  (concentrations  ranging
from  20  to  70 /Lig/m3), and only  a small
amount of SC>2 pollution [concentrations up
to 12 /zg/m3 (0.005 ppm)]; and  (3) Bohda-
nec,  which  has  no  nearby  air pollution
sources.
  Petr and Schmidt5  observed a  statistically
significant difference in  the configuration of
lymphocytes and monocytes in the smears of
peripheral blood  from children in the three
areas. Although  they failed to indicate the
size of the study group, they noted that the
children  did not differ from  one another in
age or socioeconomic grouping. The  authors
consider the results  to be a sensitive indicator
of the children's reaction to SC>2 and nitrous
gases in the environment.
  When they studied  the blood  of ten  8- to
10-year-old children from each of the three
towns, Petr and Schmidt5  observed relative
increases in resistance to hemolysis  and in-
creases in number of immature red blood cells
in children from Rosice (high-NO2/low-SO2)
and  Ohrazenice  (high-SO2/low-NO2) as
compared to children from Bohdanec (con-
trol).  They interpreted this as a compensatory
response  of  tne children  to noxious  sub-
*It is not possible to calculate ppm from total oxides of
 nitrogen.
stances  in the environment.  Methemoglobin
levels were determined in these same children,
and the investigators observed an average of
2.5 percent  methemoglobin in the  blood of
the   Rosice  (high-NO2/low-SO2)  children,
compared to an average of 0.86 percent in the
children from Bohdanec (control). The differ-
ence was statistically significant (p  <0.025),
but   the possible contribution  of large
amounts of nitrates in the  drinking water
could not be overlooked. When the study  was
repeated in the  Rosice (high-NO2/low-SO2)
children  at a later  date, after the source of
atmospheric nitrous gases  had  been con-
trolled,  they  found no abnormal  levels  of
methemoglobin.  The authors  implicated  the
atmospheric NOX as an etiologic  agent in
methemoglobinemia.
   It is difficult to interpret the significance of
hematological  changes. More  experimental
and epidemiological work are needed to con-
firm these data. For example, no information
is  given  about  antecedent  viral or bacterial
infections, which are known to infect children
seasonally, and could explain peripheral blood
changes.  Controlled studies of red-cell  age,
and of lymphocyte and monocyte  morphol-
ogy are also needed.
4.    Occupational Exposure
   While  occupational exposures to  NOX  are
fairly common, either in connection with the
use of  rocket  fuel  (nitrogen tetroxide)  or
dynamite, most  occupational exposures  in-
clude mixed pollutants and many are  poorly
documented.  Among  the  frequently  cited
studies are those  of Vigdortschik,6  who pre-
sented evidence that nitrogen oxide exposure
over  many  years is  capable of producing
pulmonary fibrosis;  and of  McCord,7 who
suggested that  some  methemoglobinemias
occur as a result  of nitrogen oxide  exposure
among welders. Such studies, although sugges-
tive, need further confirmation.
C.   IMPLICATIONS OF
     CHATTANOOGA STUDIES
   The most  substantial  studies of ambient
exposure to NO2 in the United States,  the
                                                                                    10-7

-------
Chattanooga studies, have several implications
in regard to  respiratory illness, implications
that can be extended to other cities.
   The  National  Air  Surveillance Network
(NASN)  has  measured NC<2 levels for the
years  1967, 1968, and 1969, by the Jacobs-
Hochheiser 24-hour  method (see  chapter 6,
Table 6-7), the same  method  that  was em-
ployed in the Chattanooga studies. Although
the Chattanooga analyses  were based  on a
6-month  NC»2 average,  the  yearly average
would not be substantially different,  since
NC»2  does not  exhibit marked seasonal varia-
tions  (See discussion chapter 6, Section  B.2),
and direct comparison of  the NASN yearly
averages with the lower limit at which health
effects were noted in the Chattanooga studies
is, therefore, possible. Any  site that exhibits a
concentration of 113 pig/m3 (0.06  ppm) or
greater exceeds the Chattanooga health-effect-
related NC<2 value. As might be expected, an
analysis  of the NASN data shows  that the
yearly NC»2 averages reflect variations accord-
ing to population densities. Ten  percent of
cities with populations of  less than 50,000
show a yearly  average equal to or exceeding
113 jug/m3 (0.06 ppm). In the  population
range from 50,000 to 500,000, 54 percent of
the cities equal or exceed a yearly average of
113 Mg/m3 (0.06 ppm). In the over-500,000
population class,  85  percent of the  cities
equal or exceed 113 Mg/m3 NC«2 (0.06  ppm)
on a yearly average.
   It is important to realize that the relation-
ship to population is a very general one and
that the exact location  of the sampling site
within each city or general area plays a domi-
nant  role in  determining the concentrations
measured. This is illustrated by an examina-
tion of those  locations in Table 6-7 (chapter
6) that have  more than one station in opera-
tion.  The  levels reported   in 1969 by two
Denver  and two Chicago stations show that
within any one city the NC>2 yearly averages
can differ by factors of 2 to 3, depending on
the site.
D.  FUTURE RESEARCH NEEDS
   The dearth  of information relating  com-
munity health  effects to ambient concentra-

10-8
tion of NOX is apparent, but  the following
studies are particularly needed:
   1. Studies to determine which segment of
the population  is  most  susceptible  to ab-
normal levels of NOX.
   2. Studies to precisely delineate the rela-
tionship among methemoglobin levels, periph-
eral  blood  alterations,  and  NOX  concen-
trations.
   3. Replication of studies  of the enhanced
susceptibility  to  respiratory  infection  at
various NOX levels.
   4. Studies to  determine the relationship of
other pollutants to the oxides of nitrogen and
their combined effect on human health.

E.   SUMMARY
   There  is a paucity of well-controlled epi-
demiological  studies  involving  exposure of
human populations to ambient levels of nitro-
gen oxides. The  Cleveland Clinic Fire of 1929
exposed  many  employees to extremely high
levels of HCN,  CO, and  NOX.  A follow-up
study  of survivors revealed no  increase in
mortality between  the  exposed group  of
employees and a control group.
   Effects of community exposure to NC>2 in
four residential  areas of greater Chattanooga
were studied.  In one  area,  NC>2 concentra-
tions were high and  particulate  matter was
low; in another,  NO2 concentrations were low
and particulate  matter was  high. Two  other
areas served as "clean" controls.
   The ventilatory performance (FEVg.vs) of
children  in  the high-NC>2  area  was  signifi-
cantly  reduced, when compared to the per-
formance of children in the control areas. In
addition, an  18.8 percent relative  excess of
respiratory illness  occurred among families
living in the elevated-NC>2 area.  This increased
incidence of  acute  respiratory disease in
family groups was observed when the  mean
24-hour NC<2 concentration, measured over a
6-month period, was between 117 and 205
f/g/m3 (0.062 and 0.109 ppm) and the  mean
suspended-nitrate level  was 3.8  jug/m3* or
greater.
 *Suspended nitrate is a solid, and solids are not reported in
 PPm.

-------
   In a retrospective study of the same Chatta-
nooga area, exposure to intermediate and high
levels  of NO2 in ambient  air was  associated
with a significant increase in the frequency of
acute bronchitis. This occurred among infants,
exposed  for  3 years, and school  children,
exposed for 2 and 3 years.  This increased fre-
quency  of  acute bronchitis  was  observed
when the mean 24-hour N(>2 concentration,
measured over a 6-month period, was between
118 and  156 Mg/m3  (0.063 and 0.083 ppm)
and the mean suspended nitrate level was 2.6
Mg/m3 or greater.
   A  report  from Czechoslovakia  indicates
that NOX produced several alterations in the
peripheral blood. Increased levels  of methe-
moglobin  were observed in  school children
residing  in a town that had  relatively  high
ambient levels of NOX; however, the findings
in this report require further clarifying investi-
gation.
   The Chattanooga studies have several impli-
cations in regard to respiratory illness—implica-
tions that can  be extended  to other cities.
Since NC>2 does not exhibit marked seasonal
variations, it is possible  to make direct  com-
parison of the NASN yearly averages with the
lower limit at which health  effects were noted
in the Chattanooga studies. Any site where a
concentration  of 113 Mg/m->  (0.06 ppm) or
greater is measured exceeds the Chattanooga
health-effect-related NC<2 value. Ten percent
of cities in the United States with populations
of less  than  50,000  have a  yearly average
equal to or exceeding 113 jug/m3 (0.06 ppm).
In  the  population  range  from  50,000  to
500,000, the yearly average NO 2 equals or
exceeds  113 Mg/m3 (0.06 ppm) in 54 percent
of the cities. In the over-500,000  population
class, 85 percent of the cities equal or exceed
113 Mg/m3 (0.06 ppm) yearly average NO2-
F.   REFERENCES
1.   Gregory, K. L., V. F. Malinoski, and C. R. Sharp.
    Cleveland  Clinic  Fire:  Survivorship  Study,
    1929-1965. Arch. Environ. Health, 18: 508-515,
    April 1969.
2.   Shy,  et al.  The Chattanooga School  Study:
    Effects  of Community Exposure  to  Nitrogen
    Dioxide.  Methods,  Description  ot Pollutant
    Exposure  and Results  of Ventilatory Function
    Testing. To be Published in J. Air Pollut. Contr.
    Ass. (1970).
3.   Shy,  et al.  The Chattanooga School  Study:
    Effects  of Community Exposure  to  Nitrogen
    Dioxide. Incidence of Acute Respiratory Illness.
    To be published in  J.  Air Pollut.  Contr. Ass.,
    (1970).
4.   Pearlman,  M. E., et al. Nitrogen Dioxide and
    Lower  Respiratory   Illness.  Submitted to
    Pedatrics,  1970.
5.   Petr, B. and P. Schmidt. The Influence of an
    Atmosphere  Contaminated with Sulfur  Dioxide
    and Nitrous Gases on the Health of Children. Z.
    Gesamte Hyg. Grenzgeb., 13: 34-48, 1967.
6.   Vigdortschik, N. A., et  al. The Symptomatology.
    of Chronic Poisoning with Oxides of Nitrogen. J.
    Ind. Hyg. Toxicol., 19: 469-473, 1937.
7.   McCore, C. P., G. C.  Harrold, and S. F. Meek. A
    Chemical  and  Physiological  Investigation of
    Electric Arc Welding. J. Ing. Hyg. Toxicol., 23:
    200-215, 1941.
                                                                                         10-9

-------

-------
                                  CHAPTER 11.
                     SUMMARY  AND CONCLUSIONS
A.   INTRODUCTION
  This  document  contains  a  consolidation
and  an  assessment  of  the  current state of
knowledge regarding the group of air pollut-
ants  known  as  the  oxides of nitrogen. This
chapter  provides a  concise  presentation of
that  information, including  reasonable con-
clusions for evaluating the concentrations of
nitrogen oxides  (NOX) and the accompanying
situations that have an effect on either health
or welfare. The studies and  data cited com-
prise the best available basis for developing
specific standards for NOX in the ambient air,
aimed at  protecting public  health and  the
environment.
  Although the essential  role of NOX in the
production  of photochemical  oxidants is
treated from the physical-chemical standpoint
in this  document,  little  research has been
done to demonstrate the significance of the
indirect  effects of NOX on health, vegetation,
and  materials  through  the photochemical
reaction mechanism; thus,  only the  direct
effects of NOX are treated here. A P C O pub-
lication AP-63, Air Quality Criteria for Photo-
chemical Oxidants,  provides a comprehensive
review of photochemical oxidant effects.
  Units of pollution  concentration used in
this document are expressed  as both mass per
unit  volume (e.g., micrograms  per  cubic
meter,  jug/m^)  and as  volume-ratios (e.g.,
parts per million, ppm). Conversion between
these units requires a knowledge of  the gas
density, which  varies  with  temperature  and
pressure measurement.  In this document 25°C
(77°  F) has  been taken as standard tempera-
ture, and 760 mm  Hg (atmospheric pressure
at «ea level)  as standard pressure.  All refer-
ences to NOX are expressed in terms of NC>2
mass per unit volume on the basis of the con-
version formula: ppm x 1880 = ng/m^ at 25°
C,  760 mm Hg, unless otherwise  specified.
Similarly, hydrocarbons and oxidant concen-
trations are expressed as mass of methane and
ozone per unit volume, respectively.

B.  PROPERTIES OF NITROGEN OXIDES
    AND PHYSICAL EFFECTS ON
    LIGHT TRANSMISSION
  Of the  oxides of nitrogen known to exist,
only two, nitric oxide (NO) and nitrogen
dioxide (NO2) are emitted to the atmosphere
in  significant quantities.  Nitric  oxide is
formed  during all  atmospheric  combustion
processes in a  spontaneous chemical reaction
between the nitrogen and  oxygen in the air.
The amount formed depends on the combus-
tion temperature, the  concentration of both
reactants and products, and the length of time
favorable  conditions persist for the reaction.
  Both  NO and NO2 are formed when com-
bustion  temperatures  exceed approximately
1093°C (2000° F), but usually less than 0.5
percent  is NO2- More NO2 is formed when
atmospheric oxygen (02) reacts with NO, but
at the dilute  concentrations  of  NO  charac-
teristically found in  ambient atmospheres,this
reaction  proceeds very slowly.  During the
initial phases of exhaust gas dilution,  how-
ever, the  concentration of NO is high, and
forces the reaction  to proceed more  rapidly
until the exhaust has been sufficiently diluted
(to  1  ppm or less). At that time the major
process for converting NO to NO2 reverts to
the photochemical cycle.
  Visibility reduction is common in polluted
atmospheres.  Scattering and absorption  of
light rays by  particles  and gases reduce the
                                         11-1

-------
brightness and contrast of distant objects. The
degree  of reduction depends on the concen-
tration  and properties  of  the  pollutants.
Nitrogen dioxide absorbs light energy over the
entire visible spectrum, although primarily in
the shorter, blue-wavelength regions; thus,
NO2 can by  itself reduce visibility. At pre-
sent, however,  under most  ambient condi-
tions, aerosols make  the major contributions
to visibility reduction.

C.   SOURCES AND CONTROL OF
     ATMOSPHERIC NITROGEN OXIDES
   On a  global basis, the total amount  of
nitrogen  oxides generated by natural sources
exceeds the amount from man-made, techno-
logical  sources. Natural  scavenging processes
keep  background  levels  in  nonurban areas
low, on the order  of 8 ng/m? (4  ppb) NC>2
and  2 Mg/m3  (2  ppb) NO.  In  urban areas,
however, where 60  percent of the techno-
logical sources are located, the levels are fre-
quently higher because pollutants are  added
faster than  scavenging processes control them.
   Fuel  combustion  is the major source of
technological  NOX air  pollution.  Chemical
processing  is responsible for high, but local-
ized emissions.
   Control  of NOX  emissions has  been  di-
rected at both combustion sources and chemi-
cal  processes.  For  stationary combustion
sources,  the control principle has been based
on reducing either the flame temperature or
the  availability of oxygen,  to prevent  NO
formation.  Similar principles of control are
applicable  to motor vehicles. Catalytic prin-
ciples,  which  have been applied  to reduce
NOX emissions from chemical processes, are
also  being  investigated  for  possible use in
control of NOX in motor-vehicle exhaust.
D.   CHEMICAL INTERACTIONS
     OF NITROGEN OXIDES
     IN THE ATMOSPHERE
   The role of NOX in the generation of pho-
tochemical  oxidants is a complex function of
the interaction of certain hydrocarbons (HC)
with the NO2 photolytic cycle, which is dis-
cussed here as well as in the document AP-63,
Air Quality Criteria for Photochemical Oxi-
dants and the document  AP-64, Air Quality
Criteria for Hydrocarbons.
   In order to fully describe the HC-NOX-OX
interrelationship a comprehensive  simulation
model  that takes into account emission rates,
chemical reactions,  and atmospheric disper-
sion factors,  is required. In the absence of
such an applicable model an observation-
based model  was developed and  applied to
ambient aerometric data. This latter model is
restricted to defining the maximum daily oxi-
dant that may be reached from  a given early-
morning precursor level and, therefore, the
model  results in definition of the upper-level
oxidant curve, as a function of precursor con-
centrations. The model for the NOX-OX rela-
tionship indicates that  an NOX 6- to  9-a.m.
value of 80 Mg/m3 (0.04 ppm) is associated
with  the  reference  concentration  of  200
Mg/nrP   (0.1  ppm) maximum  daily  1-hour-
average oxidant.
  The  reference concentration of 200 /ug/m3
OX used here was selected on the basis of
convenience  and  does not  represent  the
lowest  health-related value (130 Mg/m3 OX)
expressed  in  APCO publication AP-63, Air
Quality Criteria for Photochemical Oxidants.
   Application of the observation-based model
to ambient NOX, HC, and oxidant interrela-
tionships showed that the peak oxidant level
is dependent on  the concentration of both
reactants. Analysis of data from three urban
areas indicates that a reference concentration
of  200  fjig/m3  (0.1 ppm) maximum  daily
1-hour-average  oxidant is associated with an
HC range of 200 to 930 Mg/m3 (0.3 to  1.4
ppm C) 6- to  9-a.m.  nonmethane hydrocar-
bon, when the 6- to  9-a.m.   average NOX,
expressed as NO2, was  below 80 ng/m^ (0.04
ppm).   Similarly,  observation  of the  200
Mg/m3  (0.3 ppm C)  nonmethane HC level
showed NOX in the range of 80  to  320#g/m3
(0.04 to 0.16 ppm), expressed as NO2- These
conclusions are supported by  the predomi-
nance  of weekend data near the low-concen-
tration end of the upper-limit oxidant curve,
which  reflects the lower oxidant values from
lower emissions on weekends.
 11-2

-------
E.   METHODS FOR MEASURING
     NITROGEN OXIDES
   Research  is  still  needed to  develop  and
thoroughly evaluate more sensitive, reliable,
and practical methods for measuring ambient
levels of NO, NO2, and NOX.  All of the field
techniques in use at present can measure only
NO2  directly;  NO must be oxidized to NO2,
then measured. NOX can be determined either
by summing NO and NO2 concentrations that
have  been measured  independently or  by
oxidizing  NO  to  NO2, then measuring the
total as NO2-
   Any method used for measuring NO2 in
the ambient air should be calibrated against
atmospheres containing known amounts of
NO2- The use of permeation tubes to generate
the test atmospheres is recommended.
   Two techniques  are  currently  used  in
atmospheric monitoring programs.  For sam-
pling  periods of 30 minutes or less, the most
suitable currently available method for meas-
uring NO2 is the colorimetric Griess-Saltzman
method. This method can also be automated
for continuous measurement.  The Jacobs-
Hochheiser method is the most suitable of the
available methods  for  long-term (up to  24
hours) sampling, or  for situations requiring a
delay of the analysis for  more than 4 hours
after  sampling. The  Griess-Saltzman  and
Jacobs-Hochheiser  methods  are not  inter-
changeable,  can yield different results, and
must  be chosen carefully, according to the
purposes of the sampling to be done.
   When used in conjunction with an oxidiz-
ing prescrubber to  convert NO  to NO2, the
continuous Griess-Saltzman method can  be
used  to measure NO in ambient air in either a
series or  parallel  mode,  with the  same  or
separate  samples of air.  Problems  exist  in
obtaining  complete  NO to NO2  oxidation,
and researchers disagree as to which of the
two modes is more satisfactory.
F.  ATMOSPHERIC LEVELS OF
     NITROGEN OXIDES
   Continuous measurement of the oxides of
nitrogen by various monitoring networks has
made  it possible to compile tables of mean
concentrations averaged over different  time
periods and to relate various  temporal  pat-
terns to photochemical  and meteorological
parameters.
   Both NO and NO2 concentrations display
distinct diurnal variations dependent on both
the  intensity of the solar ultraviolet energy
and  the amount  of atmospheric  mixing. In
many sampling areas, these variations are also
associated with the traffic patterns.
   Nitric oxide shows an additional seasonal
variation, with higher values occurring during
the  late fall and  winter months. Nitrogen
dioxide,  however, does not  display  any
distinct seasonal patterns.
   The effect of meteorological parameters on
NO and NO2 concentrations has been reason-
ably well documented. As might be expected,
periods of stagnation and high  traffic volume
in urban areas have resulted  in  high peak
levels of NOX.
   Continuous  measurement  has  indicated
that  peak values of NO  above  1.2 mg/m3 (1
ppm) are widespread, but NO2 concentrations
have rarely been measured at this  level. Peak
concentrations of NO2 in urban areas rarely
exceed 0.94 mg/rn3 (0.5 ppm).
   Considerable differences were found among
NO2  data collected at the same site, at the
same time,  but by different methods. The
methods of NO, NO2, and NOX measurement
are  still in  need of refining  and must  be
judged accordingly.
G.  EFFECTS OF NITROGEN OXIDES
     ON MATERIALS
   Significant effects of  NOX  have been ob-
served  and  studied on  three classes of ma-
terials:  textile  dyes and  additives, natural and
synthetic textile fibers, and metals.
   The  most pronounced problem is associ-
ated with textile dyes and additives. Fading of
sensitive disperse  dyes  used  on cellulose
acetate fibers  has been attributed to NO2
levels below 188 mg/m3  (<100 ppm). Loss of
color, particularly in blue-  and  green-dyed
cotton and viscose rayon, has occurred in gas
dryers where NOX concentrations range from
1.1 to 3.7 mg/m3 (0.6 to 2 ppm). Yellow dis-
coloration in undyed white and pastel-colored
                                      11-3

-------
fabrics has recently been  attributed, to NOX
by controlled laboratory experiments.
  Laboratory  and  field  observations have
shown that  cotton  and Nylon textile fibers
can be deteriorated by the presence of NOX,
but specific reactants and threshold levels are
undetermined at this time.
  Failure of nickel-brass wire springs  on re-
lays  has  been related  to high  particulate
nitrate levels. This type of stress corrosion has
been observed  when surface concentrations of
particulate   nitrates  have  exceeded   2.4
Mg/cm^ and  relative  humidity  was  greater
than 50 percent. Another  type of this corro-
sion has  been  associated with annual average
particulate nitrate concentrations of 3.0 and
3.4 Mg/m3 with corresponding NOX levels of
124  and  158 Mg/m3 (0.066 and 0.084 ppm).

H.   EFFECTS OF NITROGEN  OXIDES
     ON VEGETATION
   The degree of injury occurring  with the
lower  concentrations of NC>2 present in the
atmosphere  remains to be determined. Expo-
sure of many kinds of plants to concentra-
tions of  NC>2 above 47 mg/m3 (25  ppm) for
any extended period  causes acute necrotic
leaf injury.  Such lesions  are usually  charac-
teristic for each plant,  but their nonspecific
character in  relation  to  other  toxicants
renders these symptoms of little value  in diag-
nosing NC>2 damage in the field.
   The 1-hour visible-injury-threshold value
 for  NC>2 can  be achieved by exposing plants
 to  18.8  to  28.2 mg/m3 (10 to 15  ppm). In-
 creasing  the exposure time, however, obviates
 the  threshold  level; 4.3  to 6.6 mg/m3 (2.3 to
3.5 ppm) NC>2 administered for 8 to 21 hours
and 1.9  mg/m3 (1  ppm) NC>2 for  48 hours
cause equivalent leaf injury. Continuous fumi-
gation with 940 Mg/m3 (0.5 ppm) NC>2 for 35
days resulted in leaf drop and chlorosis in
citrus, but no actual necrotic lesions devel-
oped.
  The effects of exposure  to low  levels of
NO2 for  extended  periods  are less evident.
Recently  completed  studies  suggested  that
470 Mg/m3 (0.25 ppm) or less of NC>2, sup-
plied continuously  for  8  months will cause
increased leaf drop and reduced yield in navel
oranges.
  The mechanism(s)  by which NOX causes
direct injury to plants can only be postulated
at this time. Evidence of diurnal fluctuation
in sensitivity to NC>2 has been presented, and
could indicate that the pollutant is reacting
with a particular plant metabolite, which only
accumulates at certain periods during the day.
The absence of a protective metabolite within
the plant at certain periods would also cause a
diurnal sensitivity.
  Limited information regarding the effect of
nitric oxide  on photosynthesis indicates that
NO would reduce  the  growth  of plants if
concentrations in the  range  of 3.8 to 7.5
mg/m3 (2.0 to  4.0 ppm) persisted contin-
uously.
I.   TOXICOLOGICAL EFFECTS OF
     NITROGEN OXIDES
   Both of the prominent  oxides of nitrogen
present in ambient air are potential  health
hazards.  At ambient concentrations, NO pre-
sents no direct threat to general health; NO2
does, however. Effects of NO2 determined in
extensive studies  are  summarized in Table
 11-1.
   The toxicology of nitrous oxide (N2O) and
other oxides of nitrogen does not appear to
be  relevant  to the problems  of ambient air
pollution at the present time.
1.   Nitric Oxide
  NO is not an  irritant and is not considered
to have adverse health effects at concentra-
tions  found in  the atmosphere.  Its  greatest
toxic potential  at ambient concentrations  is
related to its tendency to undergo oxidation
to  NO2-  A  12-minute  exposure to  3,075
mg/m3 (2,500 ppm) NO has proved lethal to
mice. In  addition, NO has been observed to
inhibit bacterial  hydrogenase activity at lower
concentrations—24.6 mg/m3  (20 ppm). This
inhibition was reversible,  however, until the
exposure reached about  12,300  mg/m3
(10,000 ppm).
2.   Nitrogen Dioxide
  NO2 exerts its primary  toxic effect on the
lungs. High concentrations, greater than 188
mg/m3 (100 ppm), are lethal to most animal
 114

-------
Table 11-1. SUMMARY OF REPRESENTATIVE N02 EFFECTS
Effect
Lowest level associated
with reference oxidant
production of 200 fig/m?
(0.1 ppm)
Increased incidence of
acute respiratory disease
in families
Increased incidence of
acute bronchitis in infants
and school children
Human
olfactory threshold
Rabbits -
structural changes in
lung collagen
Nave] orange -
leaf abscission;
decreased yield
Rats-
morphological changes
in lung mast cells
characterized by
degranulation
Mice -
pneumonitis; alveolar
distension
Mice -
increased susceptibility
to respiratory infection
Navel orange -
leaf abscission,
chlorosis
Rats-
tachypnea, terminal
bronchiolar hypertrophy
NO-? concentration,
ppm
0.04
0.062
to
0.109
0.063
to
0.083
0.12
0.25
0.25
0.5
1.0
0.5
0.5
0.5
0.8
iUg/m-5
80
117 to
205
118 to
156
225
470
470
940
1880
940
940
940
1504
Duration
3hr
(6 to 9 a.m.)
2 to 3 yr
2 to 3 yr
...
4 hr/day
for 6 days
8 mo,
continuously
4hr
Ihr
6 to 24 hr/day
for 3 to 12 mo
6 to 24 hr/day
up to 12 mo
35 days,
continuously
Lifetime,
continuously
Comment

Chattanooga study - 6-mo
mean concentration range
Chattanooga study - 6-mo
mean concentration range
Immediate perception
Still apparent 7 days after
final exposure

Possibly precedes onset of
acute inflammatory reaction
Possibly emphysematous
condition
Based on mortality following
challenge with K. pneumoniae


Reference
1
2
3
4
5
6
7
8
9
6
10
                                                          11-5

-------
             Table 11-1 (Continued). SUMMARY OF REPRESENTATIVE N02 EFFECTS
Effect
Rats-
bronchiolar epithelial
changes, loss of cilia,
reduced cytoplasmic blebbing
crystalloid inclusion bodies
Rabbits -
structural changes in lung
collagen
Sensitive plants -
visible leaf damage
Rats, monkeys -
polycythemia

Man -
increase in airway
resistance
Monkeys -
tissue changes in lungs,
heart, liver, and kidneys
NO^ concentration,
ppm
0.8
to
2.0
>
1.0

1.0
2.0

5

15
to
50
Mg/m3
1504
to
3760
1880

1880
3760

9,400

28,200
to
94,000
Duration
Lifetime,
continuously
Ihr

21 to 48
hr
3wk,
continuously

10 min

2hr
Comment
Possibly pre-emphysematous
lesion
Denaturation of structural
protein suggested

Approximate doubling of red
cell number with lesser in-
creases in hematocrit and
hemoglobin
Transient

Degree of damage directly
related to concentration
ofN02
Reference
11
12

13
14

15

16
species; 90 percent  of the deaths are caused
by pulmonary edema.
  The concentration time product determines
nonlethal  morbidity effects of  NC>2 expo-
sures. At 940 Mg/m3 (0.5 ppm) for 4 hours or
1.9 mg/m3 (1.0 ppm) for 1 hour, mast cells of
rat lungs became degranulated, possibly signi-
fying the onset of an  acute  inflammatory
reaction. These cells returned  to normal 24
hours after  exposure  was terminated.  Lung
proteins, collagen and elastin, were found to
be altered structurally in rabbits exposed to
1.9 mg/m3  (1 ppm) NC>2 for  1  to  4 hours.
The  condition was  also reversible within 24
hours. Similar changes were observed in rab-
bits exposed to 470 Mg/m3  (0.25 ppm)  NC>2,
4 hours a day for 6 days, except that recovery
was delayed and some  denaturation was still
 apparent  7 days  after  the  final exposure.
 Denaturation of collagen and elastin associ-
 ated with repeated exposure to NC>2 has been
 suggested as a  possible factor in the patho-
 genesis of pulmonary emphysema.
  Early pulmonary emphysema-type lesions
have been observed in dogs exposed contin-
 uously  to  47.0  mg/m3  (25 ppm)  for  6
months. In lung tissue from monkeys exposed
to 18.8 to 94.0 mg/m3 (10 to 50 ppm) NO2
for 2 hours, alveoli were expanded and had thin
septal walls. This response involved increas-
ing numbers of alveoli as  the NC>2 concentra-
tion was increased. Hyperplasia has been ob-
served in  respiratory  bronchiolar epithelium
of hamsters exposed to 94.0 nig/m3 (50 ppm)
for  10  weeks,  and  a similar response was
noted in major bronchi and distal portions of
11-6

-------
the respiratory tract of hamsters exposed to
18.8 mg/m3 (100 ppm) for 6 hours.
   Long-term  exposures  to  NC>2  concentra-
tions that do not produce acute inflammatory
responses have a cumulative, sustained effect,
suggestive of a pre-emphysematous condition.
Examination  of lung tissue  from  rats ex-
posed to 3.8 mg/m3 (2 ppm) for their natural
lifetimes showed loss of cilia; decreased bron-
chiolar blebbing; and intercellular, crystalloid,
rod-shaped, inclusion bodies.  Similar effects
have  been seen  in lungs  of rats continuously
exposed  to 1.5  mg/m3 (0.8 ppm). Alveoli in
lungs of  mice exposed  to  940 Mg/ni3 (0.5
ppm) for  3  to  12 months on 6-,  18-,  and
24-hour daily schedules have shown increase
in size from distension rather than from septal
breakage. The accompanying inflammation of
the bronchiolar  epithelium and reduction in
distal airway size suggested  the development
of early focal emphysema.
   Rats chronically  exposed  to 18.8 to  47.0
mg/m3 (10 to 25 ppm) N(>2 developed com-
pensatory changes, such as polycythemia and
thoracic kyphosis, with  lateral  flaring of the
ribs.
   Since certain pathological changes seen in
animals after experimental NC>2 exposure are
similar to  changes  that occur  in the patho-
genesis  of chronic  obstructive  pulmonary
disease in  man, it is suggested that long-term,
low-level exposures to NC>2 may play a signifi-
cant role in the development of chronic lung
disease.
   Exposure of mice, hamsters, and squirrel
monkeys to NC>2 increased  susceptibility to
bacterial  pneumonia and influenza infection.
The susceptibility has been demonstrated by a
significantly  increased mortality,  decreased
survival  time, and a reduction in ability to
clear  infectious  agents  from  the lungs.  In
mice, threshold for increased susceptibility to
Klebsiella  pneumoniae occurred after expo-
sure  to  6.6  mg/m3 (3.5 ppm) NC>2 for  2
hours, if the  infectious  challenge was given
within 1 hour after  the NC>2 exposure. Squir-
rel monkeys  exposed to 18.8 mg/m3  (10
ppm) NC>2 for 2 hours and then challenged
with K.  pneumonia aerosol retained the in-
fectious  agent  in their  lungs  for  extended
periods of time.
   In long-term studies of mice, significantly
increased susceptibility to infection occurred
after continuous daily exposure to 940 Mg/m3
(0.5 ppm) NC>2 for 3 months, and after 6- and
18-hour  daily  exposures for   6 months.  A
significant increase  in susceptibility to influ-
enza virus or K. pneumoniae was also seen in
squirrel  monkeys continuously exposed to
18.8 and 9.4 mg/m3 (10 and 5 ppm) NC>2 for
1  and 2 months, respectively. In  addition,
interferon formation has been impaired and
resistance to  viral   infection  has  decreased
following exposure of rabbits to 47.0 mg/m3
(25 ppm) NC>2 for 3 hours. Researchers con-
jecture that such increased susceptibility to
infection may also be significant in the patho-
genesis of human lung disease.
   Inhalation of NC>2   can  produce other
systemic effects,  although these are generally
secondary to  the effects  on   the  lungs. In
monkeys exposed to 28.2 to 94.0 mg/m3 (15
to 50 ppm)  NC>2 for 2 hours, cellular changes
appeared in heart, liver, and kidney tissue. A
circulating  substance, possibly a lung  anti-
body,  has  been detected  in  the  blood of
guinea pigs exposed  to 9.4  mg/m3 (5.0 ppm)
for 4 hours daily,  5 days  per week  for 5.5
months.  Rats  and  monkeys  continuously
exposed  to  3.8 mg/m3 (2.0 ppm) NO2 for 3
weeks   developed marked  polycythemia.
Methemoglobin  has been   detected  in  the
blood of several species exposed to NC>2 con-
centrations greater than 122 mg/m3  (70 ppm)
for 1 hour.
   The small amount of information available
concerning  the  toxicological  effects  of the
oxides of nitrogen in man  pertains to levels
higher  than those   found  in  ambient  air.
Experimental exposure of volunteer subjects
to 9.4 mg/m3 (5 ppm) NC>2 for 10 minutes
has produced a substantial, but transient, in-
crease  in airway  resistance. Other  data, de-
rived from occupational exposure to  high-con-
centration mixtures of NO  and NO2,are com-
plicated  by  the presence of other pollutants.
                                                                                    11-7

-------
Impaired  pulmonary  function,  evidenced
by reduced  maximal  breathing capacity, in-
creased  expiratory resistance, and occasional
decreased vital capacity, has been observed in
patients accidentally exposed to high concen-
trations of nitrous fumes for a few minutes.
Such evidence has persisted for more than 2
years after the exposure, in some cases. In one
case, occupational exposure to 169 mg/m3
(90 ppm)  NC>2  for  30 minutes  produced
pulmonary edema and  a  vital capacity  50
percent  lower than expected 18 hours  later.
Exposure to very high concentrations  for
about 5 minutes has produced  death within 2
days to 5 weeks.
  The threshold for odor perception of NC>2
is about 225 jug/m3 (0.12 ppm).

J.    EPIDEMIOLOGICAL APPRAISAL OF
     NITROGEN OXIDES
  Nitrogen dioxide, the only oxide of  nitro-
gen examined in epidemiological surveys,  can
be  significantly  correlated with  increased
respiratory disease at mean 24-hour concen-
trations between 117 and 205 jug/m3 (0.062
and 0.109 ppm).
  Effects of community  exposure to NC>2
were  studied in  four  residential  areas of
greater  Chattanooga. The  ventilatory per-
formance (FEVg.vs)  of children  in a  high-
NO2  area  was  significantly reduced,  when
compared to the performance of  children in
control  areas. In addition,  an 18.8 percent
relative  excess of respiratory illness occurred
among families exposed  to high NC>2 concen-
trations. A  10.4 percent  excess  occurred
among families in an elevated-particulate area.
The increased incidence of acute respiratory
disease was observed  when the mean 24-hour
NC>2 concentration, measured over a 6-month
period,  was  between  117  and 205 iug/m3
(0.062 and  0.109 ppm) and  the mean sus-
pended  nitrate level was 3.8 Mg/m3 or greater.
  In a retrospective study of the same Chatta-
nooga area, exposure to intermediate and high
levels of NC»2 in ambient air  was associated
with a significant increase in the frequency of
acute bronchitis among infants exposed for 3
years and school children exposed for 2 and 3
years. When increase was observed, the mean
24-hour NC«2 concentration, measured over a
6-month period, had ranged between 118 and
156  Aig/m3 (0.063 and  0.083  ppm) and  the
mean suspended nitrate level  had been  2.6
Mg/m3 or greater.
  A report from  Czechoslovakia indicates
that  NOX has  produced several alterations in
the  peripheral blood.  Increased  levels  of
methemoglobin were observed in school  chil-
dren residing in a town that  had relatively
high ambient  levels of  nitrogen oxides.  The
findings in that report  require further clari-
fying investigation,  however, before  conclu-
sions can be drawn.
  The Chattanooga studies have several impli-
cations in regard to respiratory illness-implica-
tions that can be  extended to other cities.
Since NC>2 does not exhibit marked seasonal
variations (See discussion chapter 6,  Section
B,2), direct comparison of the NASN yearly
averages with the lower  limit at which health
effects were noted  in the Chattanooga studies
is, therefore, possible. Any site that exhibits a
concentration  of 113 jug/m3 (0.06 ppm) or
greater  exceeds  the  Chattanooga   health-
effect-related NC»2 value. Ten percent of cities
with populations of less than 50,000 show a
yearly  average equal to  or exceeding   113
jug/m3  (0.06  ppm). In  the population range
from 50,000  to 500,000, 54 percent of the
cities in  the United States equal or exceed a
yearly  average of  113  Mg/m3  (0.06 ppm)
NO2- In the  over-500,000 population class,
85 percent of the  cities equal  or exceed  113
Mg/m3  (0.06 ppm) NO2 on a yearly average.

K.   AREAS FOR FUTURE RESEARCH
1.   Environmental Aspects of Oxides of
     Nitrogen
   The fate of as much as 50 percent of the
nitrogen  oxides that become incorporated
into the photochemical  complex is still unde-
termined, for  many  of the nitrogen oxide
end  products remain unidentified.
   Even for the identified nitrogen oxides the
relationship between emissions and air quality
11-8

-------
needs  further  definition through improved
instrumentation, expansion of the number of
monitoring stations, and more accurate deter-
mination of the location and distribution of
sources.
   A  model  for  predicting  upper limits  of
photochemical  oxidant  pollutants from  ob-
served HC and NOX levels has been presented,
but needs  further  definition, sophistication,
and revision  before it  can  be applied on  a
practical basis.

2.   Effects on Vegetation and Materials
a.   Materials
   Further  research is needed to  define reli-
able dose-response relationships for vulnerable
materials. The  effects  of variables such  as
temperature,  relative  humidity, sunlight, and
other pollutants on the damage potential  of
the nitrogen oxides must also be determined.
b.   Vegetation
   The  biochemical,  enzymatic,  and other
metabolic  responses  of plants to  ambient
levels of the nitrogen oxides are  in need  of
research-based delineation. Evidence of diur-
nal variations in sensitivity suggests the exist-
ence  of either extra-sensitive  or protective
metabolites in some plants. Evidence of syner-
gistic  effects  of NOX in mixtures containing
other air pollutants  should  be investigated
further.

3.  Toxicity of Oxides of Nitrogen
   In  order to ascribe toxicity to a specific
concentration range of NOX, the  relation of
metabolic tissue changes to  NO 2  concentra-
tion-time  responses and the relative  impor-
tance  of low-concentration,  long-time expo-
sures versus short-time,  peak  ambient concen-
trations should be studied. The interactions of
the oxides of nitrogen with particulate  pollut-
ants  in  relation to  biochemical, biophysical,
infectious, immunological, and ultrastructural
response parameters require  further research
aimed  at elucidating  possible  synergistic
damage or protection.  Tolerance  to NC>2  in
the presence  of oxidant pollutants has been
suggested  as  a result of exploratory studies,
but  the biologic importance of such protec-
tion needs to be defined.
   Further examination of in vivo biochemical
and biophysical effects of exposure to typical
ambient concentrations of the oxides of nitro-
gen  relative to:  (1) qxidation of  fatty acid
double  bonds  in lung surfactants; and (2)
denaturation or  alteration  of lung proteins
(collagen and elastin,  enzymes,  and cellular
membranes) is needed before  optimal  treat-
ment for, or protection from  exposures can
be developed.
4.   Epidemiology of Oxides of Nitrogen
   In order to determine the effect of NOX on
the health of the general population, epidemi-
ological research  must be  expanded to in-
clude:  (1) studies to  determine which  seg-
ments of the population are most susceptible
to the oxides of nitrogen; (2)  studies to pre-
cisely  delineate  the   relationship  between
methemoglobin levels, peripheral blood altera-
tions, and nitrogen oxide concentrations; (3)
replication of studies of the enhanced suscep-
tibility  to  respiratory  infection  that occurs
with exposure to ambient levels of NOX; and
(4)  studies  to   determine  the  relationship
between  other pollutants and  the oxides of
nitrogen and their material effect on human
health.
L.   CONCLUSIONS
   Derived from  a careful  evaluation of the
studies cited in  this document,  the conclu-
sions given below represent  the best judgment
of  the  scientific  staff of  the Air Pollution
Control  Office  of  EPA  regarding  the ef-
fects that may  occur when  various  levels
of nitrogen oxides are reached  in the ambient
air.  More detailed  information  from which
the  conclusions  were  derived, and the  qual-
ifications  that  entered into  the  considera-
tions of these  data, can be found in appro-
priate chapters of this document.
1.   Nitric Oxide
a.   Effects on Humans
  No evidence shows  that  NO produces sig-
nificant adverse health effects at  the ambient
                                                                                      11-9

-------
aimospheric concentrations thus far meas-
ured (chapter 9, section B.).
b.    Effects on Materials and Vegetation
  Damaging effects to materials at ambient
pollutant levels of nitrogen oxides have been
observed;  however,  concentrations  of NO
producing these effects have not been  pre-
cisely determined (chapter 7, sections C and
D).
  When beans were exposed to concentra-
tions of  12.3  mg/m3 (10 ppm), apparent
photosynthesis was reduced 50 to 70 percent;
when exposed to 4.9 mg/nr5 (4 ppm), a 10
percent reduction  occurred (chapter 8, sec-
tion B).
c.    Effects on Laboratory Animals
  A concentration  of  3,075 mg/m3 (2,500
ppm) is lethal to mice after a 12-minute expo-
sure. Fully reversible inhibition of bacterial
hydrogenase  activity occurs  at  a  concen-
tration of 24.6  mg/m3 (20 ppm) (chapter 9,
section B).
2.    Nitrogen Dioxide
a.    Effects on Humans
(1)  Short-Term  Exposure. Limited studies
show that exposure to NC>2 for less than 24
hours continuously can have several concen-
tration-dependent effects.
   1. The olfactory threshold value of NC>2 is
     about 225 jug/m3 (0.12 ppm) (chapter 9,
     section C.2.a.l).
  2. Exposure to 9.4 mg/m3 (5 ppm) for 10
     minutes  has produced transient  increase
     in airway resistance  (chapter 9, section
     C.2.a.2).
  3. Occupational exposure to 162.2 mg/m3
     (90 ppm) for 30 minutes has produced
     pulmonary  edema 18 hours later, accom-
     panied by an observed vital capacity that
     was 50 percent of the value predicted for
     normal  function  (chapter  9,  section
     C.2.b).
(2)  Long-Term Exposure. An increased inci-
dence of acute respiratory disease was observed
in family  groups when the mean range of 24-
hour NO? concentrations, measured  over a 6-
month period, was between 117 and 205
(0.062 and  0.109 ppm) and the mean sus-
pended  nitrate  level during the same period
was 3.8 Mg/m3 or greater.
  The  frequency of  acute  bronchitis in-
creased  among   infants  and school  children
when the range  of mean 24-hour NO2 concen-
trations, measured over a 6-month period, was
between  118 and  156 Mg/m3  (0.063 and
0.083 ppm) and the mean suspended nitrate
level during the same period was 2.6 Mg/m3 or
greater (chapter  10, section C. 1).
  Yearly average NO2 concentrations exceed
the Chattanooga health-effect-related value of
113 Mg/m3 (0.06 ppm) in  10 percent of cities
in the United States with  populations of less
than 50,000, 54 percent of cities with popula-
tions  between  50,000  and 500,000, and  85
percent   of  cities  with   populations  over
500,000 (chapter 10, section d.).

b.   Effects on Materials and Vegetation
  Although  damage to materials has  been
attributed to the oxides of nitrogen in ambient
atmospheres, the  precise  air-concentrations
producing these effects have not been deter-
mined (chapter  7, sections C and D).

  Crops and ornamental plants can be classi-
fied  into  three  groups  with respect to NOX
sensitivity: sensitive, low sensitive, and resist-
ant.  Several  characteristic  effects  have been
observed among the sensitive plants studied
with regard to direct NO2 exposure.

   1. Exposure to 470  Mg/m3  (0.25  ppm) of
     NO2 for 8  months caused leaf abscission
     and decreased yield among navel oranges
     (chapter 8, section  G).

  2. Exposure to NO2  concentrations of 940
     Mg/m3 (0.5 ppm) for 35 days resulted in
     leaf abscission and chlorosis on citrus
     fruit trees (chapter 8, section G).

  3. Exposure to NO2  concentrations of 1.9
     mg/m3  (1  ppm)  for 1  day can cause
     overt  leaf injury to sensitive  plants
     (chapter 8, section  G).
11-10

-------
c.    Effects on Laboratory Animals
(1)  Short-Term Exposure. Short-term effects
of NC>2 on animals can be summarized by the
analyses of five salient experiments.
   1. Exposure of rats to either  940 Mg/m3
     (0.5 ppm) for 4 hours, or  1.9 mg/m3
     (1.0  ppm)  for  1  hour has  produced
     degranulation of lung mast cells (chapter
     9, section C.I.b.3).
   2. Structural changes in collagen were ob-
     served in rabbits exposed to 1.9 mg/m3
     (1.0 ppm) for 1  hour (chapter 9, section
     C.l.b.2).
   3. The threshold  for increased suscepti-
     bility of mice to respiratory infection by
     K. pneumoniae is 6.6 mg/m3 (3.5 ppm)
     for 2 hours  (chapter 9, section C.l.d).
   4. Exposure of monkeys to 28.2 to 94.0
     mg/m3 (15 to  50 ppm) for  2 hours has
     produced damage to their lungs, heart,
     liver,  and  kidneys  and  pulmonary
     changes  that resemble those seen  in
     human  emphysema (chapter "9, sections
     C.l.b.3 and C.I.c.l).
   5. In  rabbits  exposed  to  47.0  mg/m3 (25
     ppm) for 3 hours interferon formation
     and resistance  to  viral infection de-
     creased (chapter 9, section C.l.d).
(2)  Long-Term Exposure.  Long-term expo-
sure to NO2 altered several functions in animal
circulatory and respiratory systems.
   1. Structural  changes  were found in lung
     tissue collagen  from rabbits  exposed to
     470 Mg/m3 (0.25 ppm) 4 hours a day for
     6 days (chapter 9, section C.l.b.2).
   2. Enhanced  susceptibility  of mice  to
     respiratory infection by K. pneumoniae
     was observed after  3 months of contin-
     uous exposure  to 940 Mg/m3 (0.5 ppm)
     (chapter 9, section C.l.d).
   3. Polycythemia has been  reported in rats
     and monkeys exposed  continuously to
     3.8  mg/m3  (2.0  ppm)  for 3  weeks
     (chapter 9, section C.l.c.3).
   4. Changes resembling those seen in human
     emphysema were reported in the follow-
     ing: mice exposed 6 to 24 hours daily,
     for a period of 3 to  12 months to 940
     Mg/m3  (0.5  ppm) (chapter 9, section
     C.l.b.3); rats continuously exposed to
     18.8 to 47.0 mg/m3 (10 to 25 ppm) for
     4  to  12  months  (chapter  9, section
     C.l.b.3);  and  dogs  continuously  ex-
     posed to 47.0 mg/m3  (25 ppm) for 6
     months (chapter 9, section C.l.b.3).
3.   Other Nitrogen Oxide Effects
a.   Photochemical Relationships
  An  observation-based  model applied  to
ambient NOX, HC, and oxidant interrelation-
ships  showed  that peak oxidant yield was
dependent on the concentration of both reac-
tants. Analysis of data from three urban areas
indicated  that a reference concentration of
200  jug/m3  (0.1 ppm)  maximum  daily
1-hour-average OX could be associated with a
hydrocarbon range of 200 to 930 Mg/m3 (0.3
to  1.4  ppm  C)  6-  to 9-a.m.  nonmethane
hydrocarbon,  expressed as methane, when the
6- to  9-a.m. average NOX, expressed as NO2,
was below 80 Mg/m3  (0.04 ppm). A similar
observation related an  NOX range of 80 to
320 Mg/m3 (0.04 to 0.16 ppm), expressed as
NO2, with 200 Mg/m3 (0.3 ppm C) nonmeth-
ane hydrocarbon.
b.   Stress Corrosion
  Nitrogen oxide reaction products have been
associated with corrosion and failure of elec-
trical  components. In two cities  where this
problem has been observed, the  1965 average
airborne   particulate  nitrate  concentration
were  3.0 and  3.4 Mg/m3  with  associated
average  NOX  levels of  124 and  158 Mg/m3
(0.066 to 0.084 ppm).
 M.  RESUME
   Adverse health  effects, as evidenced by a
greater  incidence of acute bronchitis  among
infants  and school children,  have been ob-
served, under the conditions prevailing in the
areas where studies were conducted, when the
mean 24-hour NO2 concentration, measured
by  the Jacobs-Hochheiser  method, over a
6-month period,  varied  from  118 to  156
Mg/m3 (0.063 to  0.083  ppm). On an  annual
basis, a maximum 24-hour average as low as
284 Mg/m3 (0.15 ppm) would be expected to
be associated  with a  6-month mean of 118
                                                                                 11-11

-------
 Mg/m3.  Adverse health effects, as 6videnced
 by an increased incidence of acute respiratory
 disease,  have been observed in family groups
 when the mean 24-hour NO2  concentration
 measured over a 6-month period was between
 117  and 205 Mg/m3 (0.062 and 0.109 ppm)
 and the mean suspended nitrate level was 3.8
      3 or greater.
   An  analysis of 3 years of data collected in
three  American  cities shows that on  those
several days a year when meteorological con-
ditions are  most conducive  to the formation
of photochemical oxidant, and the 6- to 9-a.m.
nonmethane  hydrocarbon  concentration is
200 fjig/m3  (0.3  ppm C), a 6- to 9-a.m. NOX
concentration (measured by the continuous
Saltzman Method and expressed as NO2 ) that
ranges between 80 and 320 Mg/m3 (0.04  and
0.16 ppm)  would be expected to produce a
1-hour photochemical oxidant  level of 200
Mg/m3  (0. 1  ppm) 2  to 4  hours later.  If  this
same functional relationship exists at the low-
est levels at which photochemical oxidant has
been  observed  to  adversely affect human
health, the  corresponding nonmethane hydro-
carbon  concentration  would  be  approx-
imately  130 Mg/m3 (0.2 ppm C) and the 6- to
9-a.m.  NOX level would be as high as  214
Mg/m3 (0. 1  1 ppm).

   Adverse  effects on vegetation such as  leaf
abscission  and  decreased   yield  of  navel
oranges have been observed during fumigation
studies  when  the NO2 concentration  (mea-
sured  by the continuous Saltzman  Method)
was  470  Mg/m3  (0.25   ppm)  during   an
8-month period.

  Nitrate compounds have  been identified
with corrosion and failure  of electrical com-
ponents.  In two cities where these effects
were  observed, the  average  airborne nitrate
particulate  concentrations were 3.0 and  3.4
Mg/m3  with associated average NOX  levels of
124 and  158 Mg/m3 (0.066 and 0.084 ppm).

  It is reasonable and prudent to conclude
that, when  promulgating ambient air quality
 standards, consideration  should be  given  to
 requirements for margins of safety that would
 take into account  possible effects on health,
 vegetation,  and  materials  that might  occur
 below the lowest of the above levels.

 N.   REFERENCES
  1.  Air Quality Criteria for Photochemical Oxidants.
     National  Air Pollution Control Administration.
     Washington, D.C. Publication No. AP-63. March
     1970.

 2. Shy, et al. The  Chattanooga  School Study:
    Effects  of Community Exposure to  Nitrogen
    Dioxide. Incidence of Acute Respiratory Illness.
    To  be published in J. Air  Pollut. Contr. Ass.,
    1970.

 3.  Pearlman,  M. E.,  et al. Nitrogen Dioxide and
    Lower  Respiratory  Illness.  Submitted  to
    Pediatrics, 1970.

 4.  Henschler, P, et al. Olfactory Threshold of Some
    Important Irritant Gases  and  Manifestations in
    Man  by  Low   Concentrations. Arch.
    Gewerbepathol.   Gewerbehgy., Berlin,  17:
    547-570, 1960.

 5. Mueller, P. K. and M. Hitchcock. Air Quality
   Criteria-Toxicological  Appraisal  for  Oxidants,
   Nitrogen Oxides,  and Hydrocarbons.  J.  Air
   Pollut. Control Ass., 19: 670-676, 1969.

 6. Thompson, C. R., et al. Effects of  Continuous
    Exposure  of  Navel Oranges to NO2- Atmos.
    Environ. In Press, 1970.

 7. Thomas, H. V., P. K. Mueller,  and G. Wright.
    Response  of Rat  Lung Mast Cells to Nitrogen
    Dioxide Inhalation. J. Air Pollut. Contr. Ass., 17:
    33-35, 1967.

 8.  Blair, W.  H., M.   C. Henry, and R. Ehrlich.
    Chronic Toxicity of Nitrogen Dioxide: II. Effects
    on  Histopathology of Lung  Tissue.  Arch.
    Environ. Health, 18: 186-192, 1969.

 9. Ehrlich, R. and M. C. Henry. Chronic Toxicity of
    Nitrogen Dioxide:  I. Effects  on Resistance to
    Bacterial Pneumonia. Arch. Environ. Health, 17:
    860-865, 1968.

10. Freeman, G., N..J. Furiosi and G. B. Haydon.
   Effects of Continuous Exposure to 0.8 ppm N©2
   on Respiration of Rats. Arch. Environ. Health,
   13: 454-456, 1966.
  11-12

-------
 11.  Freeman, G. and  G. B. Haydon. Emphysema
     After  Low-Level  Exposure  to  NC^.  Arch.
     Environ. Health, 8: 125-128, 1964.

12. Buell, G. C., Y.  Tokiwa, and P. K. Mueller. Lung
    Collagen and  Elastin  Denaturation  In  vivo
    Following Inhalation of Nitrogen Dioxide. Cali-
    fornia State Dept. of Public Health. (Presented at
    the  Annual  Air Pollution Control Association
   Meeting.) San Francisco. Paper No. 66-7, June
    1966.

13. Heck.,  W. W. Plant Injury  Induced by Photo-
    chemical  Reaction Products  of Propylene-Nitro-
    gen Dioxide Mixtures. J. Air Pollut. Contr. Ass.
    14: 255-261, July 1964.

14.  Freeman,  G.,  et  al.  The  Subacute  Nitrogen
    Dioxide - Induced Lesion of the Rat Lung. Arch.
    Environ. Health, 18: 609-612, 1969.

15.  Abe, M.  Effects of  Mixed  NO2  - SO2 Gas on
    Human Pulmonary Functions. Bull. Tokyo Med.
    Dent. Univ., 14: 415-433, 1967.

16.  Henry, M. C., R. Ehrlich, and W. H. Blair. Effects
    of  Nitrogen  Dioxide on Resistance of Squirrel
    Monkeys to Klebsiella pneumoniae  Infection.
    Arch. Environ. Health, 18: 580-587, 1969.
                                                                                                 11-13

-------

-------
           APPENDIX.




VOLUME TO MASS CONVERSION TABLES

-------

-------
VOLUME (ppm) TO MASS (Mg/m3) CONVERSION TABLE FOR NO
         (ppm x 1230 = Mg/m3 at 25° C, 760 mm Hg)
ppm
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
0.19
0.20
0.21
0.22
0.23
0.24
0.25
0.26
0.27
0.28
0.29
0.30
0.31
0.32
0.33
0.34
0.35
jug/m3
10
20
40
50
60
70
90
100
110
120
140
150
160
170
180
200
210
220
230
250
260
270
280
300
310
320
330
340
360
370
380
390
410
420
430
ppm
0.36
0.37
0.38
0.39
0.40
0.41
0.42
0.43
0.44
0.45
0.46
0.47
0.48
0.49
0.50
0.51
0.52
0.53
0.54
0.55
0.56
0.57
0.58
0.59
0.60
0.61
0.62
0.63
0.64
0.65
0.66
0.67
0.68
0.69
0.70
Mg/m3
440
460
470
' 480
490
500
520
530
540
550
570
580
590
600
620
630
640
650
660
680
690
700
710
730
740
750
760
770
790
800
810
820
840
850
860
ppm
0.71
0.72
0.73
0.74
0.75
0.76
0.77
0.78
0.79
0.80
0.81
0.82
0.83
0.84
0.85
0.86
0.87
0.88
0.89
0.90
0.91
0.92
0.93
0.94
0.95
0.96
0.97
0.98
0.99
1.00





Mg/m3
870
890
900
910
920
930
950
960
970
980
1000
1010
1020
1030
1050
1060
1070
1080
1090
1110
1120
1130
1140
1160
1170
1180
1190
1210
1220
1230





                       A-l

-------
VOLUME (ppm) TO MASS (Mg/m3) CONVERSION TABLE FOR NO2
         (ppm x 1880 = Mg/m3 at 25° C, 760 mm Hg)
ppm
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
0.19
0.20
0.21
0.22
0.23
0.24
0.25
0.26
0.27
0.28
0.29
0.30
0.31
0.32
0.33
0.34
0.35
Mg/m3
20
40
60
80
90
110
130
150
170
190
210
230
240
260
280
300
320
340
360
380
390
410
430
450
470
490
510
530
550
560
580
600
620
640
660
ppm
0.36
0.37
0.38
0.39
0.40
0.41
0.42
0.43
0.44
0.45
0.46
0.47
0.48
0.49
0.50
0.51
0.52
0.53
0.54
0.55
0.56
0.57
0.58
0.59
0.60
0.61
0.62
0.63
0.64
0.65
0.66
0.67
0.68
0.69
0.70
Mg/m3
680
700
710
730
750
770
790
810
830
850
860
880
900
920
940
960
980
1000
1020
1030
1050
1070
1090
1110
1130
1150
1170
1180
1200
1220
1240
1260
1280
1300
1320
ppm
0.71
0.72
0.73
0.74
0.75
0.76
0.77
0.78
0.79
0.80
0.81
0.82
0.83
0.84
0.85
0.86
0.87
0.88
0.89
0.90
0.91
0.92
0.93
0.94
0.95
0.96
0.97
0.98
0.99
1.00





Mg/m3
1330
1350
1370
1390
1410
1430
1450
1470
1490
1500
1520
1540
1560
1580
1600
1620
1640
1650
1670
1690
1710
1730
1750
1770
1790
1800
1820
1840
1860
1880





                       A-2

-------
                                SUBJECT  INDEX
Acute plant injury, 8-1—8-2
  necrosis (marginal, intercostal), 8-2
  discoloration, 8-2
  sensitivity, 8-2
Alveoli, 9-6-9-9
Anaerobic  isoenzymes,  9-12
Anaesthetic, 9-17, 9-19. See also "Nitrous
  oxide"
Analgesic, 9-17, See also "Nitrous oxide"
Antibody, See "Blood"
Annual variation
  NOX concentrations, 6-4
Asthma, 10-6
                    B
Bacteria, 9-12-9-14
  and tobacco, 9-13,9-14
Behavior, voluntary, 9-10
Blood 9-1, 9-10-9-12
  antibody, 9-11
  changes in cell morphology, 10-7
  cyanosis, 9-1, 9-16
  hemolysis, 10-7
  immature red cells, 10-7
  leucocytes, 9-11
  methemoglobin, 9-1, 9-2, 9-11, 10-7, 10-8
  polycythemia, 9-11
Bronchitis, acute, 10-6
  lung, 9-10,  See "Lung"
  morphology changes, 9-4—9-10, 10-7
Cellulosics, 7-3, 7-4
Chattanooga, 10-1-10-9
Chemical properties of
  Nitric oxides, 2-1
  Nitrogen dioxide, 2-2
  Nitrogen pentoxide, 2-4
  Nitrogen sesquioxide, 2-4
  Nitrogen tetroxide, 2-4
  Nitrous oxide, 2-4
Children, effects on,  10-3-10-9
Chronic plant injury, 8-1, 8-2
Cleveland, 10-1
Cleveland Clinic fire, 9-17, 10-1
Collagen, 9-3-9-5
Concentration
  of nitric oxide, See "NOX concentration"
  of nitrogen dioxide, See "NOX concentra-
  tion"
Control of man-made NOX, 3-4
Cotton fading, 7-3-7-5
Cyanosis, 9-1,9-16
Czechoslovakia, 10-7
                    D
Diurnal variation
  NOX concentration, 6-1—6-2
  plant sensitivity, 8-4—8-11
Dyes, 7-1-7-7
  blue green, 7-1, 7-6
  dye additives, 7-3, See also "Fading"
  resistant, 7-2, 7-4, 7-7
Calibration, 5-1—5-6
   of nitrogen dioxide methods, 5-3
   of nitric oxide methods, 5-5
   systems 5-1, 5-3, 5-5, 5-6
CAMP data, 6-8-6-15
Cells
   blood, See "Blood"
                    E
Eastern air characteristics. 7-6
Edema, pulmonary, 9-2, 9-3, 9-6, 9-16, 9-17
Elastin, 9-3-9-5
Electric-arc welding, 9-16, 9-17
Emphysema, 9-7, See also "Lung"
                                           1-1

-------
Enzymes, 9-2, 9-12
  inhibition, 9-2
  plant, 8-3, 8-10
  systems, 9-12

Epidemiology, 10-1 — 10-9

Exhaust, diesel, 9-15
Exposure
  effect on mortality rate, 10-1
  high, 10-1, See also "Plant injury"
  occupational, See "Occupational exposures"
Fading, 7-1-7-3,7-7
  gas-fume, 7-1—7-3
  resistance, 7-3
  test (AATCC), 7-2
Ferric oxide, 9-10
Fibers, 7-4-7-5
  cellulosics, 7-3, 7-4
  cotton, 7-3-7-5
  Nylon, 7-5
  Rayon, 7-1,7-3
  Spandex, 7-5
  synthetics, 7-5
Formation of:
  Nitric oxide, 2-1
  Nitrogen dioxide, 2-1, 2-2
Fuel combustion, 3-1
Gas, silage, 9-16
Gas-fume fading, 7-1—7-3
Griess-Saltzman  method,   5-1,  See   also
   "Sampling"
                      H
HC-NOX-OX relationship
   concentration variation, 4-10—4-13
Hemolysis, See "Blood"
Humidity, 7-3, 7-6, See afao"Synergism"
Hydrocarbons
   ambient levels and maximum oxidant lev-
     els, 4-3
   concentration variation, 4-10—4-13
  indication of oxidant levels, 4-2
  role in oxidant formation, 4-1
Hyperplasia, 9-7
                    I
Immunologic effects, 9-13—9-15
Infection, 9-13-9-14
  bacterial, 9-13
  viral, 9-14
  respiratory, succeptibility to, 10-1 — 10-6, 10-8
Influenza virus, 9-14, 10-3
Interferon formation, 9-14
                    J
Jacobs-Hochheiser  method,  5-1,  See also
   "Sampling"

                    K

Klebsiella pneumoniae, 9-12

                    L

Light absorption by nitrogen dioxide, 2-4, 2-5
Lipid peroxidation, 9-5
Lung
   alveoli, 9-6-9-8
   antibody, 9-11
   cells, 9-4-9-10
   edema, 9-2, 9-3
   function, 9^, 9-5, 9-15-9-17
   mast cells, 9-6
   proteins, 9-3-9-5
   respiratory rate, 9-5, 9-17
   smoking, 9-13,9-14, 10-3
                                                                  M
Mast cells, 9-6
Materials, effects of NOX on, 7-1-7-7
Maximum concentration
   calculation of, 6-5—6-7
Metals, 7-5-7-7
   Eastern U.S. air damage to, 7-6
   humidity and  damage, 7-6
1-2

-------
   Los Angeles, 7-5
   New York, 7-6
   Nickel-brass alloys, 7-5, 7-6
   nickel-palladium, 7-6
   participate nitrates, synergism, 7-7
   stress-corrosion, 7-5—7-6
   switches, 7-6
   telephone relays, 7-5
   Western U.S. air damage to, 7-5
   wire relays, 7-5—7-6
   zinc, 7-6
Meteorology
   effect on NOX concentration, 6-7—6-8
Methemoglobin, 9-1, 9-2, 9-11, 10-7, 10-8
Methods
   for measuring nitrogen dioxide
     colorometric, 5-1
     continuous, 5-2
     correlation spectrometric, 5-2
     gas chromatographic, 5-2
     Griess-Saltzman, 5-1
     Jacobs-Hochheiser, 5-1
     long-path infrared, 5-2
     manual, 5-1
     oxidation, 5-3
     reduction,  5-2
   for measuring nitric oxide
     chemiluminescent, 5-3—5-5
     oxidation to nitrogen dioxide, 5-3
     parallel mode, 5-4
     series mode, 5-3, 5-4
   for measuring total nitrogen oxides
      chemiluminescent, 5-5,5-6
     oxidation to nitrogen dioxide, 5-5
Monitoring networks, 6-8—6-37
Mortality, 9-2, 10-1
Mothers, effects of NC>2 on health of, 10-5
                    N
Narcosis, 9-1
NASN data, 6-32-6-37
Nickel-brass alloys, 7-5
Nickel-palladium, 7-6
Nitric acid fumes, 9-17, 9-18
Nitric oxide concentration, See "NOX concen-
  tration"
  toxicology, 9-21
 Nitrogen oxides
   ambient levels and  maximum oxidant lev-
     els, 44-4-10
   concentration ,variation of, 4-10—4-13
   definition, 2-1
   indicators of oxidant levels, 4-2
   role in oxidant formation, 4-1
 NC>2 photolytic cycle,  4-1—4-2
 NOX concentration
   annual pattern, 6-4
   calculation of, 6-5, 6-7
   continuous monitoring, 6-8—6-37
   definition, 2-1
   diurnal pattern, 6-1-6-2
   effect of meteorological factors, 6-7—6-8
   seasonal pattern, 6-3—6-4
   urban pattern, 6-8—6-32
 Nylon, 7-5
                    O
Observational model
   HC-NOX-OX relationship, 4-3
   interpretations of, 4-13—4-18
   limitations of, 4-13
   reduction in  photochemical  precursors,
     4-16-4-18
   weekday, weekend effect, 4-14—4-16
Occupational exposures, 9-15-9-17, 10-7, 10-8
Odor perception, 9-14-9-15
Oxidants - See "Photochemical oxidants"
Ozone
   formation of, 4-1
   relationship  to  NO2  photolytic  cycle,
     4-1-4-2
Particulate - nitrates,  7-7, See  also  "Syner-
   gism"
Pathology, 9-3-9-10, 9-20
Permeation tubes, 5-3
Peroxy nitrates
   effects of, 8-1
   formation of, 4-18
Phagocyte inhibition, 9-12
                                                                                      1-3

-------
Photochemical oxidants
   concentration levels, 4-2—4-20
   concentration variation, 4-10—4-13
   maximum daily levels, 4-3-4-12
   rates of formation, 4-1
   relation to ambient hydrocarbon levels, 4-2
   relation to ambient NOX levels, 4-4—4-10
   variables of formation, 4-2
Physiological  effects on  plants  - apparent
   photosynthesis, 8-1, 8-3
   enzyme inactivation, 8-3
   leaf chlorosis, 8-2
   leaf-drop, 8-2
Plant injury, 8-1-8-11, See also "Vegetation"
Plant response factors
  low light intensity, 8-3
  environmental factors, 8-3
Polycythemia, 9-11
Population density, 10-3
Precursor-product relationship, 4-1, 4-10—4-13
Pulmonary edema, 9-2, 9-3, 9-6, 9-16, 9-17
Pulmonary  function, 94, 9-5,  9-15-9-19,
   10-1-10-9

                    R

Rayon, 7-5
Respiratory illness, acute, 10-5
Respiratory infection, 10-1-10-8
Respiratory infection,
  experimental, 9-12-9-14
Respiratory rate, 9-5,
Respiratory tract, 9-3-9-10
Sampling
   calibration of methods, - See "Calibration"
   CAMP, 6-8-6-32
   continuous, 5-2, 5-6
   intermittent, 5-1
   NASN, 5-2
Seasonal variation
   NOX concentration, 6-3—6-4
Silage gas, 9-16
Smoking, See "Bacteria", "Synergism"
Sources and emissions, 3-1, 3-2, 3-3
   man-made, 3-1, 3-2, 3-3
  natural, 3-1
  nonurban, 3-1
Spandex,7-5
Stress-corrosion, 7-5, 7-6
Structural protein, 9-3—9-5
Sulfur dioxide, 10-2
Suspended particulate and NC>2, 10-1, 10-5
Synergism, 9-3-9-5,9-10

  hydrocarbon-NOx, See "Hydrocarbons"
  low-light intensity on plants, 8-3
  moisture, on plants,  8-3, See also "Humid-
     ity"
  ozone, See "Ozone"
  smoking, 9-10,  9-13, 9-14, 10-3
  sulfhydryl groups and nitrous acid, 8-3
  sulfur dioxide,  8-4, 9-15, 10-2-104, 10-6
  suspended particulate-NO2, 10-1, 10-5
  temperature-NO2, 9-3
Synthetics, 7-5
Systemic effects of NC>2
  on animals, 9-10-9-14, 9-28
Tidal volume, 9-4, 9-7
Tissue recovery, 9-7
TNT plant, 10-2
Tobacco, 9-10, 9-13
Toxicology
  NO, 9-21
  NO2, 9-22-9-28
                    U
Urban NOX concentrations, 6-8—6-32
Vegetation
  acute injury, 8-1—8-2
  chronic injury, 8-2
  physiological effects, 8-3
 14

-------
  response factors, 8-3                           Welding,  See "Electric-arc
  dose-injury relationship, 8-4                      welding"
  Vitamin E, 9-5                                Whites, yellowing of, 7-3


                    W                                               Z
Weekend, weekday levels, 4-14—4-16
Weight, 9-10                                    Zinc, 7-6
                                                                                         1-5

-------