OF
Z
\
      \
       <*
       o
       *
NBS SPECIAL PUBLICATION
557
    U.S. DEPARTMENT OF COMMERCE/National Bureau of Standards
   Chemical Kinetic Data Needs



          the Lower  Troposphere

-------
                  NATIONAL  BUREAU OF  STANDARDS


The National Bureau of Standards' was established by an act of Congress on March 3, 1901.
The Bureau's overall goal is to strengthen and  advance the Nation's science and technology
and facilitate their effective application for public benefit. To this end,  the Bureau conducts
research and provides: (1) a basis for the Nation's physical measurement system, (2) scientific
and technological  services for industry and government,  (3)  a  technical basis tor equity in
trade, and (4) technical services to promote public safety. The Bureau's technical work is per-
formed by the National Measurement Laboratory, the National  Engineering Laboratory, and
the Institute for Computer Sciences and Technology.

THE  NATIONAL MEASUREMENT  LABORATORY provides the national system ot
physical and chemical and materials measurement; coordinates the system with measurement
systems of other nations and furnishes essential services leading to accurate and uniform
physical and chemical measurement throughout the Nation's scientific community, industry,
and commerce; conducts materials research leading to  improved methods ot measurement,
standards, and data on the properties of materials needed by industry, commerce, educational
institutions, and Government; provides advisory and research services to other Government
agencies; develops, produces, and distributes  Standard  Reference Materials; and provides
calibration services. The Laboratory consists of the  following  centers:

   Absolute Physical Quantities2 — Radiation  Research—Thermodynamics and
   Molecular  Science — Analytical Chemistry — Materials Science.

THE NATIONAL ENGINEERING LABORATORY provides technology and technical ser-
vices  to  the  public and private sectors  to address  national needs and to solve  national
problems; conducts research in engineering and applied'science in support ot these efforts;
builds  and maintains  competence in the necessary  disciplines required to carry  out this
research  and technical  service;  develops  engineering data and  measurement capabilities;
provides engineering measurement traceability services;  develops test methods and proposes
engineering standards and code changes; develops and  proposes  new engineering practices;
and develops and improves mechanisms to transfer results of its  research to the ultimate user.
The Laboratory consists of the following centers:

   Applied Mathematics—Electronics and Electrical Engineering2 — Mechanical
   Engineering and Process Technology" — Building Technology — Fire Research —
   Consumer  Product  Technology — Field Methods.

THE  INSTITUTE  FOR  COMPUTER  SCIENCES   AND  TECHNOLOGY  conducts
research and provides scientific and technical services  to aid Federal agencies in the selection,
acquisition,  application,  and  use  of computer technology  to improve ellectiveness  and
economy in Government operations in accordance with  Public  Law 89-306 (40 U.S.C. 759),
relevant Executive Orders, and other directives; carries  out  this mission by  managing the
Federal Information  Processing Standards Program, developing Federal ADP standards
guidelines, and managing Federal participation in ADP voluntary standardization activities;
provides scientific and technological advisory services and assistance to Federal agencies; and
provides the technical foundation for computer-related policies  of the Federal Government.
The Institute  consists  of  the following centers:

   Programming  Science and Technology — Computer Systems Engineering.

 Headquarters and Laboratories al Gaithersburg, MD, unless otherwise noted;
mailing address Washington, DC  20234.
!Some divisions within the center are located at Boulder, CO  80303.

-------
Chemical Kinetic Data  Needs  for Modeling
the Lower  Troposphere
Proceedings of a Workshop
held at Reston, Virginia
May 15-17, 1978
John T. Herron, Robert E. Huie,
and Jimmie A. Hodgeson, Editors


National Measurement Laboratory
National Bureau of Standards
Washington, DC 20234


Sponsored in part by
Environmental Protection Agency
Research Triangle Park, NC  27711
U.S. DEPARTMENT OF COMMERCE, Juanita M. Kreps, Secretary

  Luther H. Hodges, Jr., Under Secretary

  Jordan J. Baruch, Assistant Secretary for Science and Technology

NATIONAL BUREAU OF STANDARDS, Ernest Ambler, Director


Issued August 1979

-------
        Library of Congress Catalog  Card  Number: 79-600125
        National Bureau of Standards Special Publication  557
              Nat. Bur. Stand. (U.S.). Spec. Publ. 557, 105 pages (Sept. 1979}
                               CODEN: XNBSAV
                          U.S. GOVERNMENT PRINTING OFFICE

                                WASHINGTON:  1979
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402
                          Stock No. 003-003-02111-3 Price $4

                   (Add 25 percent additional for other than U.S. mailing)

-------
                             Foreword
   It is increasingly recognized that the integrity of the  data input
is one of the most serious limiting factors in modeling complex chemical
systems.  The reliability of the results generated in modeling studies
is critical  considering their application to environmental  regulation
and control.

   The purpose of this workshop was to bring together modelers, chemical
kineticists,  theoreticians and program managers,  in order to define the
critical data needs for modeling the troposphere.   This collection  of
review papers, comments, and recommendations should serve a wide community
of atmospheric scientists in identifying and attacking priority problem
areas.

   The National Bureau of Standards is pleased to be responsible for
this publication, and to have joined with the Environmental Protection
Agency as cosponsors of the workshop.


                                             John D. Hoffman, Director
                                             National Measurement Laboratory
                                             National Bureau of Standards
                                 iii

-------
                               Preface
   It used to be said, "At least the air we breathe is free."  For about
30 years that has  not been the case in the United States.   We are now
paying about 10 billion dollars a year in an increasingly difficult
effort to maintain our atmosphere at a degree of pollution  near current
levels.  That sum will almost certainly increase as we turn towards
alternate energy sources such as coal, shale oil and crude  oils with
high sulfur contents.  It will grow even larger as we learn more about
the impacts of various chemical substances, both natural and man made,
on the complex ecology in which man exists on this planet.

   By contrast one can estimate that about 3 million dollars will be
spent this year (1978) on basic research, by all government agencies for
learning the molecular details of the chemistry of air pollution.  It is
perhaps very flattering to the scientific community involved in this
effort to consider that such a sum will suffice to make significant
progress in the understanding of the chemistry of air pollution.  But
the scientists involved share no such illusions.  In 30 years of research
effort we have learned a great deal about the chemistry of our "dirty"
atmosphere but we can hardly pretend to give quantitative answers to
questions which are increasingly being asked, such as, "What will the
effect be on our atmosphere of removing or adding X tons per week of
substance A?"  Yet important economic decisions rest on the answers to
such questions.  The sums spent on basic research are dwarfed by the
sums now being spent on regulation and abatement.  Research funds are
the least expensive aspect of the problem of air pollution, and without
the answers that they might provide, our expensive efforts  may be largely
squandered.

   The present conference brings together many of the leaders in the
basic research effort directed towards air pollution.   They have examined
the scientific details with a fine microscope and come up with what must
seem to the layman an endless multitude of unanswered questions.  Some
of these may never be answered and some of them need urgently to be
resolved.  Air pollution will not go away.   It will  become  worse even
with current efforts.  No city on earth will  escape its effects.  If we
hope to ameliorate it, we must devote a more profound and longer range
effort to its understanding than we have done so far.   The  recommendations
of this symposium point the direction these efforts must take.   As
chairman, I would like to take this opportunity to express  my appreciation
to my fellow colleagues who have given generously of their  time and
effort towards making this a fruitful meeting.


                                             Sidney W.  Benson
                                             University of  Southern California
                                             Los Angeles, CA  90007
                                  IV

-------
                             Abstract
   This is a report of the proceedings of a workshop on chemical  kinetic
data needs for modeling the lower troposphere, held at Reston,  Virginia,
May 15-17, 1978.  The meeting, sponsored by the Environmental  Protection
Agency and the National Bureau of Standards, focussed on six key problem
areas in tropospheric chemistry: reactions of olefins with hydroxyl
radicals and ozone, reactions of aldehydes, free radical reactions,
reactions of oxides of nitrogen, reactions of aromatic compounds, and
reactions of oxides of sulfur.

   The report includes a summary and list of major recommendations for
further work, review papers, discussion summaries, contributed  comments,
recommendations, and an attendance list.
Key words:  Aldehydes, aromatics, chemical kinetics, data needs, free
            radicals, modeling, NO , olefins, SO , troposphere.
   In order to describe experiments adequately, it has been necessary to
identify commercial materials and equipment in this book.   In no case does
such identification imply recommendation or endorsement by the National
Bureau of Standards, nor does it imply that the material  or equipment is
necessarily the best available for the purpose.

-------
                             Contents
                                                                 Page

Foreword	  iii
Preface	   iv
Abstract 	    v
Introduction 	   ix
Summary and Recommendations of Workshop  	    1
Session I:  Reactions of Olefins with Ozone and Hydroxyl Radicals

An Evaluation of Chemical Kinetic Data Needs for Modeling
the Lower Troposphere:  Reactions of Olefins with Hydroxyl
Radical and with Ozone -- Hiromi Niki	     7

     Summary of Session	    14
     Comments	    15
     Recommendations 	    23
Session II:  Aldehydes

Tropospheric Chemistry of Aldehydes -- Alan C.  Lloyd 	    27

     Summary of Session	    46
     Comments	    46
     Recommendations 	    47


Session III:  Organic Free Radicals

Organic Free Radicals -- David M. Golden 	    51

     Summary of Session	    61
     Comments	    62
     Recommendations 	    66
Session IV:  NOX Chemistry

Tropospheric Chemistry of Nitrogen Oxides   A Summary of the
Status of Chemical Kinetic Data -- Richard A. Cox  	    71

     Summary of Session	    74
     Comments	    74
     Recommendations 	    79
Session V:  Aromatics

Reactions of Aromatic Compounds in the Atmosphere --
Dale G. Hendry	   85

     Summary of Session	   91
     Comments	   92
     Recommendations 	   95
                                 vn

-------
                       Contents (continued)
Session VI:   SO  Chemistry

S0x Chemistry (Abstract only)  --  Jack  G. Calvert  	   99

     Summary of Session	   99
     Comments	100
     Recommendations 	  101
Workshop Attendees 	  103
Subject Index  	  107
Author Index 	  107

-------
                            Introduction
   The Environmental Protection Agency and the National Bureau of Standards'
Office of Environmental Measurements and Center for Thermodynamics and Molec-
ular Science sponsored a workshop entitled "Chemical Kinetic Data Needs for
Modeling the Lower Troposphere," at Reston, Virginia, May 15-17, 1978.  The
objective of the workshop was to assess and make recommendations on mechanis-
tic and kinetic data needs for modeling chemical transformations occurring in
the lower troposphere.

   The workshop was organized around six major topics:   reactions of olefins
with hydroxyl radicals and ozone, the chemistry of aldehydes, free radical
chemistry, the chemistry of oxides of mitrogen, the chemistry of aromatic
compounds, and the chemistry of the oxides of sulfur.  These general topics
cover almost all of the important problem areas in homogeneous chemical
kinetics of interest to the atmospheric scientist.  Heterogeneous processes
were not included.

   Each technical session opened with a review paper followed by a discussion
period.  A set of recommendations was prepared based on the review paper and
subsequent discussions.

   This report of the meeting includes the review papers (with one exception),
discussion summaries, written contributions to the discussion, and recommenda-
tions.  It opens with an overview and summary of the workshop recommendations.

   We want to thank all those involved in organizing and running the workshop,
and all those who through their participation helped make it a success.
                                 ix

-------
                          Summary  and  Recommendations  of  Workshop
   The recommendations of the workshop are  given
in detail  following each session.   Here we
summarize  the major themes of the  workshop  and  list
the most important areas requiring additional
experimental  or theoretical  work.

   Major advances  in the chemistry of the tropo-
sphere depend on understanding the chemistry of
large molecules and free radicals.  Real atmos-
pheres contain large amounts of hydrocarbons C^ and
greater, aromatic  compounds, natural products such
as terpenes, and large aldehydes, ketones,  phenols,
etc, which are their photooxidation products, as
well as a large class of oxygen containing  free
radicals which are the intermediates in these
photooxidation reactions.  These complex molecules
are not only involved in the NO-NO  conversion
process, but almost certainly are important precur-
sors to atmospheric aerosols.   We  will  not  under-
stand either  oxidant or aerosol  formation until we
attack the problem of large  molecules.

   Clearly, moving away from fairly simple
surrogate reactants such as propylene, and
considering the whole range of atmospheric
pollutants creates a problem in scale.  There are
far too many molecules and potential reactions  to
measure everything.  A proper attack on this
problem involves a judicious mix of experiment  and
theory.  We may illustrate this by considering
one of the most pressing problems   the chemistry
of alkoxy radicals.  Large alkoxy radicals  can
isomerize, decompose, or react with oxygen  (as  well
as NO  and SO , see below).   The relative rates of
thesexprocessis must be known.   Even though this
problem was extensively considered at this  work-
shop it is doubtful if a convincing solution was
given.  Measurements are needed to provide  a base
set to allow for the development of theoretical
estimation schemes.

   Another class of reactions of great importance
both in the atmosphere and in laboratory investiga-
tions are alkylperoxy radical reactions. In the
absence of NO, these radicals react with themselves
to produce aldehydes and alcohols.  There is
considerable uncertainty as to the mechanism of
these reactions.  The suggestion made at the
meeting that one of their products might be the
Criegee intermediate emphasizes the need for much
more work in this area.  At the same time the
peroxy radicals formed in the reactions of  OH with
olefins in the presence of 02 need identifying.

   In addition we need data on acetyl  and acetyl-
peroxy type radicals.  There are questions  as to
formation  of acids, particularly from formyl
radical reactions, which cannot be explained on
the basis  of existing data.

   Another major deficiency is in  the area  of
 reactions  of  aromatic compounds.   Recognition  of
 their  importance  is fairly recent.   In particular
 we  need to know about rates and mechanisms of
 reaction of aromatics with OH radicals.  This
 will involve  extension of existing experimental
 approaches and development of new ones.  The
 branching  ratios for different products need to
 be measured and the subsequent chemistry of these
 products needs to be considered.

   The possibility of making the Criegee inter-
mediate from alkylperoxy radicals was noted above.
The Criegee intermediate is presumably a primary
product of an ozone-olefin reaction.   Its
subsequent fate is of great importance.   A crucial
question is whether it decomposes or is stabilized,
and if stabilized what chemical  reactions it can
undergo.  There is some evidence that small  Criegee
intermediates  decompose.   For  the large  ones
however, there is very little  quantitative data.
This question  needs resolution since  the Criegee
intermediate has been postulated to be a potential
oxidizer for NO, S02,  olefins, etc.

   In addition to treating free  radical  reactions
in terms of isomerization, scission,  self-reaction,
and reaction with 02,  we must  consider reactions
with NO  and SO .   Here we are faced  with problems
of rate! and mechanisms and in particular the
problem of the role of association reactions.   For
peroxy radicals the starting point is the H02-N0
reaction.   The new value for the rate constant has
had a dramatic effect on the models.   It needs to
be studied over a wide range of conditions
(temperature,  pressure)  to confirm this  value under
atmospheric conditions.   The use of the rate
constant data  for H02 + NO for R02 +  NO reactions
may be invalid.  Direct measurements  are needed.
In addition, for large alkylperoxy radicals we need
to know if alky! nitrates are  products since  this
is a chain terminating reaction.

   Similar considerations apply in the case of the
reactions of alkylperoxy radicals with N02,
although it is not likely that the peroxynitrates
formed in  simple association reactions would  have
a significant lifetime in the atmosphere.  How-
ever, if the reaction can lead to an  aldehyde and
nitric acid it could be of considerable importance.

   Reactions of alkoxy radicals with NO and N02
can also proceed via channels leading to adducts
or to HNO  or HONO respectively.   The overall  rate
constants  and branching ratios need to be
determined.  Similar considerations apply to OH
reactions  with NO and N02.

   A somewhat different approach to the question
of the importance of adduct formation is to
consider the thermal stability of peroxy radicals.
Work of this kind has been done for some PAN

-------
compounds, but needs to be extended to higher
members of the PAN series.

   The serious gaps in knowledge of radical-NO
reactions are repeated in the case of the radi£al-
SO  reactions.  These reactions are almost certain-
ly involved in sulfate aerosol formation.  At the
outset we need more information on the kinetics of
reaction of OH and H02 with S02 over a wide  range
of temperature and pressure.  What are the products
and what is their subsequent chemistry?  It  will
be necessary to consider the reactions of HS03 and
HSOn with water, NO , and other atmospheric
species.

   Similarly  the reactions of  R02 with S02 need
to be  measured over a range of conditions.   The
products  should  be  identified  and their  subsequent
chemistry  determined.

    In  addition,  we  need  to  know  about the  reaction
of the Criegee  intermediate with S02.  How fast  is
this  reaction and what are  the products?

    In  the  general area of  photochemistry,  the
major  problems  lie  with  the photochemistry of
aldehydes  and ketones.   There  are still  some
serious problems involved  in  the photooxidation  of
formaldehyde.   In particular the origin  of formic
acid  remains  a mystery.  There is not much data
available  on  acetaldehyde  photolysis or  of higher
aldehydes  or  of  the ketones in general.  In
addition  photolysis rates  for  alky! nitrates and
alkyl  peroxnitrates need to be measured.

    Although we now have  a  qualitative  idea of the
 mechanisms of reactions  of olefins  in  the  atmos-
 phere there are still  many areas in which  quantita-
 tive detail is needed.   There is a  need  for  rate
 data on reactions of hydroxyl  radicals  and ozone
 with cyclo-olefins  and natural  products  (isoprene,
 terpenes, etc).   The  rate of  reaction  of OH  with
 ethylene should be  studied at higher  pressure  where
 the possibility exists of the  hot  CzH^QH*  adduct
 reacting with 0^.   Above all  the questions of ozone
 reaction mechanisms remains unresolved,  particularly
 for large olefins,  cyclo-olefins,  terpenes,  etc.
 Studies under atmospheric  conditions  are needed.
 Also we need to know under what  conditions OH
 abstraction vs.  addition become  important.

    The problems of heterogeneous chemical  kinetics
 and aerosol formation came up repeatedly during
 the meeting.  Surface effects in smog chambers
 include free radical  initiation of chamber
 reactions, and absorption and desorption of
 reactive species.  The role of wall  effects  in
 chamber studies will  have to be resolved before
 chamber studies can be used to validate  complex
 chemical kinetic models.

    Aerosol formation initiated by reactions  of
 large organic molecules involves aspects of  both
 homogeneous and heterogeneous chemical  kinetics.
 There  is a whole range of problems  which need
 study.  The  role of hydration of free radicals was
 touched on many times during the meeting.  We do  not
 know  which radicals (if any)  are truly hydrated, and
 what  are the kinetic consequences of hydration.  Are
 condensation nuclei formed in free radical   reac-
 tions?  Thest problems are not only of great
 interest in modeling atmospheric chemistry,  but  are
of great importance in the design and execution of
laboratory studies of the elementary chemical
kinetics to be used in modeling studies.

   Certainly a principal  objective of homogeneous
chemical kinetics should  be a fundamental  under-
standing of the initial  reactions leading  to the
formation of atmospheric  aerosols.

   Finally, since this workshop was directed
toward kinetics data needs for modeling the
troposphere, it is appropriate to include  a general
comment on the meeting by R. J. Cvetanovic:

"Adequate understanding of the chemistry of
photochemical  smog will be made possible only
through comprehensive modeling of the chemical
processes which occur in the  polluted troposphere.
The success of such modeling  will depend very
critically on  the availability of as complete a
list as possible of the elementary chemical
reactions likely to be involved and of reliable
values of their rate constants under tropospheric
conditions.  Incomplete or  unreliable information
could lead to  erroneous conclusions and result in
ultimately very costly misinterpretation of the
pollution.  Accumulation and  continuous updating
nature of the  problems posed  by tropospheric
of the  necessary  information  will require  the
following steps:  1) establishment of a comprehensive
list of chemical  reactions potentially involved
in tropospheric chemistry, preferably in the form
of a reaction grid of the type first used  in the
Climatic Impact Assessment Program (CIAP)  for
modeling the stratosphere; 2) critical selection
of the most reliable values of the rate constants
of these reactions with estimates of their limits
of uncertainty; 3) initiation of measurements of
the rate constants which are at present not
available; 4) updating the rate data through
continuous monitoring of the new values which
become available; 5) initiating any studies
necessary for improved understanding of reaction
mechanisms when this information is not available
or is uncertain".

   The major recommendations of the workshop, in
approximate order of priority are as follows:

Recommendations:

1.  The rate constants for'isomerization,  scission,
and reaction with oxygen, of a base set of alkoxy
radical reactions should be measured.  This
initial set should include methoxy, ethoxy,
propoxy, n-, s-,  t-butoxy, and hydroxy-butoxy.
Both absolute and relative rate measurements  should
be considered.

2.  Continued  efforts should  be devoted to  improv-
ing existing theoretical approaches, based  on RRKM,
to calculate isomerization, and scission rates of
alkoxy  radicals.

3.  The mechanisms and rates  of self-reaction of
alkylperoxy and hydroxy-alkylperoxy radicals  should
be studied.  The question as  to the formation of
Criegee intermediates should  be resolved.

4.  The radicals formed in  the reaction of  OH with
olefins in the presence of  02  should be identified.

-------
5.  The reactions of formyl and higher members of
the acetyl radical series with 02 should be studied.
A careful  search for acid products should be made.

6.  The mechanisms of reaction of OH radicals with
aromatic compounds should be determined under
atmospheric conditions.  A base set should include
benzene, toluene, the xylenes, trimethylbenzene,
and ethyl  benzene.

7.  The chemistry of the Criegee intermediate
should be  determined.  What conditions of molecular
size, pressure, and temperature lead to stabiliza-
tion?  Does it react with other atmospheric
constituents, and if so what are the products?

8.  The kinetics of the reaction H02 + NO should
be studied over a wide range of temperature and
pressure.

9.  The kinetics of the reactions of alkylperoxy
radicals,  starting with methylperoxy, with NO
should be  measured over a range of pressure and
temperature.

10. The extent to which alkyl nitrates are formed
from reaction of long chained alkylperoxy radicals
with NO should be determined.

11. The reactions of alkylperoxy radicals with N02
should be  studied and the extent to which nitric
acid is a  product should be determined.

12. The reactions of alkoxy radicals with NO
should be  studied.  The ratio of formation of
adduct to  HNO + aldehyde should be determined
under atmospheric conditions.  The overall rate
constant should be measured directly or relative
to the rate of reaction with 02.

13. The reactions of alkoxy radicals with N02
should be  studied.  The ratio of formation of
adduct to  HONO + aldehyde should be determined
under atmospheric conditions.  The overall rate
constant should be measured directly or relative
to the rate of reaction with 02.

14. The rate  of reaction of OH with NO and N02
should be  measured over a wide range of temperature
snd pressure.
 15. The thermal  stability of  selected  PAN  compounds
 should be determined.   Compounds which should  have
 first priority are peroxypropionyl nitrate and
 peroxybenzoyl nitrate.

 16. The kinetics of the reactions of OH and H02
 with S02 should  be measured over a wide range of
 temperature and  pressure.

 17. The reactions of HS03 and HSO^ with H20, NOX,
 NH3 and other atmospheric species should be studied.

 18. The kinetics of the reactions of alkylperoxy
 radicals with S02 should be measured over a wide
 range of temperature and pressure.   The products
 of the reactions should be identified and their
 subsequent reactions with H20, NO , NH3, and other
 atmospheric species should be stuoied.

 19. The reactions of Criegee intermediates with S02
 should be studied.

 20. The photooxidation of formaldehyde should be
 studied, and the mechanism of formation of formic
 acid determined.

 21. A quantitative  study of  the  photolysis  of acet-
aldehyde and higher aldehydes  should  be undertaken.

 22. Quantum yields and absorption cross sections
 should be measured for ketones.

 23. Quantum yields and absorption cross sections
 should be measured for alkyl peroxynitrates and
 alkyl nitrates.

 24. Selected ozone-olefin reactions, including
 cycloolefins should be investigated under
 atmospheric conditions.

 25. The kinetics of reaction of OH with ethylene
 should be studied under atmospheric conditions.

 26. The rates of reaction of ozone with large
 olefins, cyclo-olefins, terpenes, etc.  should be
 measured.

 27. The rate of  reaction of hydroxyl  radicals with
 large olefin, cyclo-olefins, terpenes, etc., should
 be measured.

-------
                     Session I
Reactions of Olefins with Ozone and Hydroxyl Radicals

-------
                             AN EVALUATION OF CHEMICAL KINETIC DATA NEEDS
                                 FOR MODELING THE LOWER TROPOSPHERE:
                      REACTIONS OF OLEFINS UITH HYDROXYL RADICAL AND WITH OZONE
                                             Hiromi  Niki
                                          Ford Motor Company
                                      Dearborn, Michigan  48121


        Needs for improved kinetic and mechanistic data for the  reactions  of olefinic hydrocarbons
     with hydroxyl  radical and with ozone have been evaluated from the view point of modeling the
     chemistry of the lower troposphere.   Research priorities for removal  of various uncertainties
     in these reactions have been discussed briefly.

     Key words:  Hydroxyl; kinetics; olefin; ozone; review; troposphere.
             1.   Introduction


   In planning for the abatement and control  of air
pollution, it is essential  to establish the
quantitative chemical  relationship between source
emission and the resulting  air quality.  At present,
concerted modeling efforts  are being made to
achieve this goal.  Chemical  interpretation of smog
chamber data [I]1, prediction of "ozone-isopleth"
[2], and air-shed modeling  [3] of Los Angeles
Reactive Pollutant Program  (LARPP) data [4] are a
few of the notable examples of such endeavor.
Clearly, the degree of success of modeling work is
governed, in large part, by the availability  of
reliable kinetic data.  Despite recent progress,
the knowledge of the chemical reactions taking
place in the lower troposphere is far from
satisfactory.  There exist  numerous critical
uncertainties in the kinetics and mechanisms  of
reactions involving a large variety of atmospheric
constituents.  This paper is intended to assess
the needs for improved experimental data for
olefin reactions with HO and with 03.

   For the purpose of evaluating the existing needs
for improved kinetic and mechanistic information
for these reactions, their  potential role in  the
lower troposphere is discussed briefly.  Further,
the current knowledge of these reactions is
illustrated by some of the  work published within
the last few years.  This paper is not intended to
be an extensive literature  review, but rather, to
convey the author's thoughts on the future
direction and research priorities in the area of
tropospheric chemistry.  In view of the urgency
and long-term interest in establishing firm
scientific bases for the abatement of air pollution
problems, the more systematic data evaluation
efforts, e.g., the "chemical  reaction matrix"
method  [5]  used  for  the  evaluation  of  the  strato-
spheric  chemistry, should  be  made  in the future.

    2.   Atmospheric Role  of Olefin  Reactions
       with  HO and with  03.
Figures  in brackets  indicate  literature references
at the end of this paper.
    Olefins  are among the most  reactive  classes of
organic  compounds  present  in the  lower  troposphere.
In  particular, their possible  role  in the  formation
of  photochemical smog has  been well-recognized
over the  last three decades [6].  As a  result,
olefins  have been  used extensively  as surrogate
hydrocarbons in laboratory smog studies, and have
played the  crucial role in the development of smog
chemistry [1].  It now appears that the atmospheric
fate of  olefins are governed primarily  by  their
reactions with HO  and with Oa.  Conversely, these
reactions are responsible  for  regulating the
atmospheric concentrations of  HO  and 03.   The
latter aspect is of primary importance, since HO is
the major chain carrier of atmospheric  reactions
and determines the role of other  hydrocarbons and
.organic  compounds  in the formation of "oxidant",
e.g., 03.  Whether 03-olefin reactions  can lead to
the formation of "excess"  03 or alternatively serve
as a sink for 03 is another key question to be
answered.

   It must be stressed that a quantitative evalua-
tion of the atmospheric role of the olefin
reactions, or for that matter any other reactions,
can be made only on the basis of  numerical modeling
studies for a given source distribution and
strength under a variety of meteorological
conditions.   Clearly, relative importance  of HO and
03 reactions involving various olefins  can vary
markedly between "fresh" and "aged" air masses
because of the chemically  and meterologically
induced changes in relative and absolute olefin
concentrations.

   To illustrate the relative importance of various
olefinic and other types of hydrocarbons,  table 1
shows the results obtained by Calvert [7]  for the
relative rates of HO-radical  attack on  hydrocarbons

-------
        Table
                                                                            CO present  in the
Compound
Til
*C2Hi»
C2H2
C3H8
*C3H6
*iso-C,1H10
n-C^Hjo
*1-C,,H8
iso-Ci,H8
iso-C5H12
n-C5H12
Cyclo-C5H10
*1-C5H10
*2-Methyl butene
*2,2-Dimethylbutene
2-Methyl pentane
3-Methyl pentane
*1 -Hexene
n-Hexane
*Cyclohexene
2, 2, 3-Tri methyl butane
2-Methyl hexane
(RH), ppb,
mol basis
2.010
49
43
38
37
8.7
12
37
1.5
3.0
44.3
16.2
2.6
4
0.8
0.8
11.0
10.0
1 .7
10.0
10.7
7.7
8.2
6.9
Rel. rate,
HO reaction
2.8
2.2
11.6
1.0
4.4
21.8
4.2
13.3
8.9
27.9
29.2
8.9
2.1
•>- 2.4
•v. 7.4
•v 4.7
9.1
8.3
•^ 10.0
7.1
-v. 10.7
4.4
6.3
6.3
Compound
3-Methyl hexane
*l-Hepteneb
n-C7H16
Methyl cyclohexane
2,2,3- and 2,3,3-
Trimethyl pentane
2, 2, 4-Tri methyl pentane
Tol uene
*1 -Methyl cyclohexene
2, 2, 5-Tri methyl hexane
n-C8H18
EtC6H5
p.m-Xylenes
o-Xylene
n-C^Hjo
n-PrCGH5
sec-BuC6H5
n-C10H22
n-C11H2l)
n-C12H26
CO
Total alkane
Total alkene
(RH), ppb,
mol basis
6.3
4.4
4.3
3.7
1.9
2.5
2.0
4.7
1.0
2.1
4.1
1.4
6.0
1.3
1.0
5.0
1.1
1.0
0.3
1,910
Total aromatic hydrocarbons
Grand total

Rel. rate,
HO reaction
5.8
•v 25.9
3.8
4.7
1.4
1 .8
17.6
•v- 6.0
0.9
2.2
-v 4.2
•v 32.1
11.3
1.6
•v 1 .2
% 7.0
1.5
1.5
0.5
47.8 (12.1%)
128.0 (32.5%)
138.3 (35.1%)
79 . 7 (20. 2%)
393.8
        ^Taken from table IV in  ref. [7].
         Asterisks indicate olefins which contribute significantly to the removal of HO-radical.
in Los Angeles air samples  [4].   It  should  be
noted in this table, that a  variety  of  olefins,
indicated by asterisks, contribute significantly
to the removal of HO-radical.  The olefins  as  a
whole are responsible for 35  percent of the HO
removal by hydrocarbons.  Similarly, relative
removal rates of hydrocarbons  by  03  can be
estimated for the data given  in table 1.  The
results are shown in table  2.  Only  the olefinic
hydrocarbons are listed in  this table,  since 03
reactions with both paraffinic and aromatic
hydrocarbons are negligibly  slow.  Notably, in
this particular case, by far  the  dominant hydro-
carbon-ozone reactions involve cyclic olefins,
i.e., cyclohexene and 1-methylcyclohexene rather
than the straight chain olefins commonly used  as
surrogates.  Calvert [7] also  estimated the
comparable fractional rates  of removal  of C3H§ by
03 (3.2 percent h ') and by  HO (4.8  percent h  l)
for conditions given in table  1.  These examples
amply demonstrate that detailed hydrocarbon
analyses of air samples are  essential to the
understanding of atmospheric  chemistry.

   Because of the close chemical  coupling among 03,
NO and N02 concentrations  via photo-stationary
relationship  [9], the role  of olefins in the
oxidant formation can be assessed only  in  terms of
their effects on NO  chemistry.   Specifically, the
key question  is  "how do the atmospheric reactions
of olefins with HO and 03  control  NO and N02 con-
centrations?"  The relevant  chemistry is shown
schematically in figure  1, where  R represents H
atom or hydrocarbon  radicals  formed, for instance,
from olefin reactions with HO and 03.  Thus, to

    Table  2.  Estimated relative rates  of 0, attack
             on  the olefins  given in table 1.
        Compound
(RH)   Rel.  03a  Rel.  removal'
ppb      rate      rate
C2H< 43
C3H6
l-C,.Hn
iso-C.He
1-C5H,0
2-methylbutene
2,2-dimethylbutene
1 -hexene
cyclo-hexene
1-heptene
1 -methyl cyclohexene
^Taken from ref. [8]
8.
1.
3
4


1.
10.
4.
4.

.7
.5


8
8b
7
.7
4b
,7b

Estimated from analogous
1
1
1

37.
1.

13.

15.

.15
.0
.0
.0
.8
.9
.1
,g
.0
.9
.4

2.
3.

1.
1 .
11 .


51.
1 .
26.

. 3
.2
.5
.2
.2
.2
3
5
4
4
7

reactions.

-------
     Fig.  1.  Reactions controlling Oa level
             in the lower troposphere.
account for the behavior of 03, it is necessary to
know the kinetics and mechanism of various RO and
R02 species interacting with NO  and all other
labile atmospheric constituents.  In fact, the
predictive capability of modeling work on 03
formation hinges on the understanding of free
radical chemistry involving NO [1].  It is fair to
state that at present, there is a serious lack of
reliable information on the question "how
effectively do olefins generate radical species
and what is the subsequent fate of these radicals
in the atmosphere?"  In the following sections
the current knowledge on olefin reactions with HO
and 03 will be discussed to address these questions.

   3.  Current Knowledge of Olefin Reactions
       With HO and With 03.

           A.  HO-olefin reaction

   Kinetics:  As noted in the preceeding section,
the atmospheric life time and radical formation
efficiency of olefins are governed, in large part,
by the HO reactions.  Therefore, these rate
constants should be determined with utmost
accuracy.  In recent years, numerous direct and
relative measurements of these rate constants have
been made over a wide range of temperatures and
diluent pressures [10].  In particular, various
direct experimental methods employing flash photo-
lysis-resonance absorption and fluorescence, and
discharge-flow-laser magnetic resonance have been
successfully used to measure the decay rates of
HO in the absence of interferences from secondary
reactions.   Thus, the overall accuracy of the
rate constants determined by these methods is
determined by the inherent signal-to-noise ratios
in the HO decay curves and by the measurement of
the reactant concentrations.  Therefore, the
highest attainable accuracy should be expected
from these measurements as exemplified by the
excellent agreement between two recent determina-
tions of HO + C3H6 by Atkinson and Pitts [11]
k = 25.1 ± 2.5 x 10~l2 cm3 molecule"1 s"1 and by
Ravishankara et al. [12] (k   25.6 ± 1.2 x 10"12)
at 298 K.

   An extensive, critical review of previous
studies on HO-olefin reactions has been made
recently by Atlkinson et al. [10].  In the case of
the lightest olefin, C2H.,, the pressure dependence
of k has been well established.  The rate constant
at 300 K is in the fall-off region between 2nd
and 3rd order kinetics below ^ 225 Torr of Ar and
below -\> 300 Torr of He.  Table 3 shows a summary
of limiting high pressure data on HO + C2H., taken
from the review by Atkinson et al.  [10].  There is
a spread of a factor of 2 among these values,
although the recent values are more consistent
but not up to the expected accuracy.  The litera-
ture values of k for HO + C3H6 are also summarized
in table 3.  The fall-off region for C3H6 appears
to occur at much reduced pressure of ^ 1 Torr.
Also included in this table are several values
derived from relative decay rates of hydrocarbons
measured in the photolysis of HC-NO  mixtures at
ppm concentrations.  These relative values comple-
ment those of direct studies, but must be
considered to be less precise (by as much as ±
20 percent).

   The most extensive and consistent set of both
direct and relative rate data for a large variety
of olefins including terpenes and haloalkenes have
been obtained recently by Pitts'  group [10].
Their values agree, in general, with those
determined by others to within ± 25 percent or
better.  Therefore, the kinetics of HO-olefin
reactions should be considered to be reasonably
well established.  However, it should be noted that
thus far, no direct measurements of these constants
have been made in the presence of 1 atm air.
Therefore, it is desirable to obtain further
verification and improvement of these rate constants
over wider range of pressures and temperatures.

   Mechanism - Primary Step:  There have been con-
flicting mechanistic interpretation of kinetic and
product data obtained at low diluent pressures
concerning the relative importance  of HO-addition
to olefinic double bond and H-atom  abstraction.
However, recent studies show convincingly that HO
radicals undergo predominantly, if  not exclusively,
addition reactions with C2Hi,, C3H6  and other
methyl-substituted olefins, and that the resulting
adducts are collisionally deactivated in 1  atm of
air.  For example, in the case of C2Hi», Howard [13]
has shown that the rate constant extrapolates to
zero at zero diluent pressure.  Thus, the abstrac-
tion reaction is of negligible importance.   In the
case of C3He, Cvetanovic [14] has made a comprehen-
sive analysis of products formed in the photolyses
of N20-CH,»-C3H6 and N20-H2-C3H6 mixtures, and
concluded that the majority of the  observed
products could be accounted for by  reactions of
radical species arising from the HO adducts.  From
these data, Cvetanovic further deduced the extent
of the terminal addition being approximately 65
percent.  It is interesting to note that Gutman
et al. [15] earlier obtained evidence for
H-abstraction from C3H6 and trans-2-butene under
virtual absence of collisional deactivation in
their cross-molecular beam   mass spectrometric
studies.  Thus, H-abstraction does  occur to a
minor extent.  A somewhat puzzling  study of the
HO-C2Hi, mechanism is that of Meagher and
Heicklen [16].  These investigators photolyzed
H202 in the presence of C2Hi, and excess N2 in
both the presence and absence of added 02 at 298 K.
The observed quantum yields of products such as
CH20, HCOOH and C2H3OH were taken to imply that
the abtraction accounts for 20 percent of the
primary reaction.  This conclusion  is at odds
with those of other workers and must be judged
to be incorrect.

-------
Table 3.  Rate constant data and Arrhenius parameters for the reaction of OH radicals with alkenes.3
1C12 :- A_
Alkenes cm5molec~'s '
Ethene 1.26
(limiting high
pressure data)
—
2.18
1012 '
E cal mol ' cm'molec
-903 +136 5.
5.
6.
7.
8.
-770 ± 300 7.
33C
33 +
23 ±
21 +
1 ±
85 ±
10.0 -
0
0
0
1
0
1
k
-is-,
.65d
.33
.33
.6
.79
.7
At T K Technique
298-301
300
381
416
305 + 2
299
296
FP-KS
FP-RF
PR
Rel ati ve
rate
FP-RF
FP-RA
Temperature
range
Reference covered
Greiner, 1970 (97)a
Davis et al . , 1975 (165)
Gordon & Hulac, 1975 (115)
Lloyd et al . , 1976 (135)
(relative to QH+n-butane
2.82 - 10"12)
Atkinson, Perry & Pitts,
1977 (168)
Overend & Paraskevopoulos,
1977 (103)
298-498 K


299-425 K

 Propene
                   4.1
                             -1080 ± 300
  17  + 4


5.0  + 1.7

14.5  * 2.2

13.4  ± 3.4
14.3 +  0.7

20.0 *  1.0

   5 ±  1

25.1 ±  2.5


25.6 ±  1.29


27.4 *  5.5



20.8



23.5 ±  3.5



23.5 ±  4.7
  300


•>. 300

  298

  298
  381

  416

  300

  298


  298


305 >



  303
DF-MS


DF-ESR

FP-RF

Relative
 rate
PR

DF-RA

FP-RF


FP-RF


Relative
 rate


Relative
 rate
                                                       305 ± 2  Relative
                                                                 rate
              305
                                                             2  Relative
                                                                 rate
Morris, Stedman & Niki,
1971  (78)

Bradley et al., 1978  (151)

Stuhl, 1973  (169)

Gorse & Volman, 1974  (123)
(relative to QH+CO
1.50  10"13)


Gordon & Mulac, 1975  (115)

Pastrana & Carr, 1975 (170)

Atkinson & Pitts, 1975     298-424 K
(155)

Ravishankara et al.,  1978
(105)

Lloyd et al., 1976 (135)
(relative to QH-n-butane
2.82 x 10 ]1)e

Uu, Japar &  Niki, 1976 (121)
(relative to OH+cis-2-butene
5.20 x 10"")"

Winer et al., 1976 (136)
(relative to OH + isobutene
4.80 - 10"")h

Winer et al., 1977 (140)
(relative to OH + isobutene
4.80 * 10"")h
  .Taken from ref. [10].  Reference numbers are indicated as  they appear in  ref. [10].
   Mean Arrhenius preexponential factor.
  jTotal pressure not stated, but stated to be the same as in previous work  (86), i.e.,  100 Torr of helium.
   Essentially the high pressure limits from a Lindeman plot  (175).
  ^Calculated from the Arrhenius pressure expression of reference (153) for  T   305 K.
   Reference  (86).
  9Rate constants at 20 Torr total  pressure with helium as the diluent gas.  Mo pressure effects were
  .observed over the total pressure range 3-20 Torr  (1-butene and cis-2-butene) or 20-200 Torr (oropene)
   Calculated from the Arrhenius expressions of reference (155)  for T   303  K or T = 305 K.
   In the  case of olefins  containing weak  allylic
hydrogens, e.g., 1-butene  and 3-methyl-1-butene,
Atkinson et al. [17] have  postulated,  from the
correlation of HO and  0{3P)-atom reactivities
towards olefins, that  H-atom abstraction can occur
up to  30 percent of the total reaction.   Clearly,
extensive product  studies for larger olefins in
the  presence of an  inert diluent gas at 1  atm
pressure are needed to obtain more  definitive
information on the  questions of HO-addition vs.
abstraction, and position of the HO-addition.

   Secondary Reactions in the Atmosphere:   The
nature  of atmospheric  reactions initiated by HO-
                olefin  reactions  is,  generally,  even less  certain
                than  the corresponding primary  processes.   However,
                in the  case of C2H,,,  C3H6 and 2-C,,H8, for  which
                the HO-addition has  been shown  to be the predomin-
                ant primary step,  some significant progress has
                been  made on the  mechanistic interpretation of
                secondary reactions  involving the HO-adducts in
                the presence of 02 and NO    In  particular,
                computer-aided numerical analyses of smog  chamber
                data  have played  a major role.

                   In computer modeling of smog  chamber data,  the
                degree  of success  in arriving at a unique  chemical
                mechanism may be  judged by the  extent of agreement
                                                       10

-------
       [0+ C=C-C~|
        ADDUCT J"
           0(3P)
                          -C=C-C
              H02, N03, ROX
           (-4%)
     M /(60%)\ (40%)
la
1/2 C2H5CHO
+ 1/2 C-SC-C
(OVERALL
~2%)»
C2H5.

+ HCO
(OVERALL

~ 1%)
                                      (30%)
        FRAGMENTATION
        |_   PATH    J
/
OH
^
/ ' *
* -i C-C-C
C=C-C ADDUCT
/ \
,- * -. OH ONO
TCRIEGEE] \ ^
PATH
/ \ 
-------
departure from conventional  smog chamber-based
studies, and should  be  extended to other olefinic
compounds.

   In conclusion,  there is  a great need for kinetic
and mechanistic data on the oxidation of free
radicals formed from HO-olefin reactions.  The
above-mentioned studies illustrate a classical
photochemical approach  to these problems.  Clearly,
it is highly desirable  to study these chemical
systems with more  direct experimental methods.

           B.  03-olefin reactions

   The atmospheric consumption of olefins proceeds,
to a large extent, by their reactions with 03-
Furthermore, the question of whether the 03-olefin
reactions serve as an 03 source or sink hinges on
the efficiency of  free  radical formation by the
reaction.  Thus, utmost accuracy is required for
the kinetic and mechanistic information on 03-
olefin reactions for modeling purpose.

   Kinetics:  Over the  years, there have been
numerous determinations of  the rate constants for
the reactions of 03  with a  large variety of
olefins.  A summary  of  literature values for
several olefins  is given in table 4.  Generally,
the reported  values  agree reasonably well for
terminal olefins,  but scatter far beyond the
estimated experimental  precision for internally
double-bonded olefins.   For instance, two of the
most recent sets of  extensive measurements by
Huie and Herron  [21] (ref.  [1] in table 4), and
by Niki et a!.  [8] (refs. f and k) disagree by as
much as 48 percent for  trans-2-butene and by an
average of 27 percent for the entire set.  The
individual values  in the latter set are all higher
than those in the  former   Apparently, there are
factors other than systematic measurement errors
affecting one or both experiments.  In these
studies as well  as most of  the others, the rate
constants were derived  from the decay rates of 03
in the presence of excess olefins.  In addition,
the former employed  much higher reactant concentra-
tions and lower total pressures (< 10 Torr) than
the latter (at 1 atm air).   Therefore, the
fundamental question is whether the consumption of
QI is due entirely to the primary process or is
interfered with by secondary reactions.   Unfortu-
nately, there exist  large uncertainties in the
reaction mechanisms, and such effects cannot be
determined reliably  at  present.

   There are several instances of kinetic evidence
for the consumption  of  03 by secondary reactions,
e.g., retardation  of excess 03 decay.rates by
02, [21, 22] and dependence of 03 and olefin decay
rates and reaction stoichiotnetry on reactant
mixing ratios [22].   Therefore, all the reported
values of the rate constants for 03-olefin
reactions must be  considered as upper limit values.
In particular, the higher values obtained by Niki
et al. [8] might reflect the extent of such effect.

   In short, experimental methodologies required
for obtaining the  "true" rate constants for 03-
olefin reactions have not been firmly established
at present, and should  be developed in the future.
Meanwhile, concerted efforts should be made to
minimize systematic  errors  as revealed in table 4.
Concomitantly, kinetic  and  mechanistic studies of
  Table 4.  Rate constants for qas phase ozone-olefin
           reactions at  room temperature.3

                          k,10"'a cm3 molec"1-:"1
                              Literature	
  Ethylene


  Propylene


  1-Butene

  1-Pentene

  1-Hexene



  Dialkylethylenes

   Isobutene


   cis-2-Butene


   trans-2-Butene
  Trialklethylenes

   2-Methyl-2-butene

   cis-3-Hethyl-2-
   pentene

   trans-3-Methyl-2-
   pentene


  Tetraalkylethylenes

   2,3-Dimethyl-2-
   butene


  Others

   Cyclopentene

   Cyclohexene

   1,3-Butadiene
           l-3, 10.6
1.2e, 1.3b, 1-6T, 2.6C,

2.7d, 3.0a, 1.9k, I.?1

6.2a, 7.5C, 8.2b, 11.0C

12.5f

9.0C,

5.3a,

9.2a,

11.Od
10.0°

7.5b,

10.0
            12.3*
           9.0°, 10.7
              10.31
               ,k
        ,a,b
         10.2C,

        ,11.11
6.2b, 8.4C, 15h, 23d,

13.6k, 11.7k

28C, 50h, 140\ 340d, 161k,

1261

35C, 166h, 2601, 275f,

430d, 260k, 1761
29C, 790\ 493k, 4001

456k


563k
39°, 750d, 1510k, 10601
813

301.

8.2b
 59°, 169

, 9.1\ 8.4k
   Taken from ref.  [8].  References (a)  through (k)
   given in  ref. [8]; reference  (1) is ref. [23] of
   this text.

free-radical reactions  involving 03  and olefins
should be made to better  characterize the relevant
secondary reactions as  discussed below.

   Mechanism:  To date, no  direct monitoring of
reactive intermediates  formed in 03-olefin
reactions has been made under atmospheric
conditions.  Therefore, it  is compelling to resort
to the deduction of mechanistic  models from product
studies.  The model will  provide a basis for
establishing experimental priorities to reduce
mechanistic uncertainties.  Significant progress
has been made recently  both in characterizing key
reaction products and in  computer modeling of
these results.

   To illustrate, the reaction of 03 with C2H, has
been studies by Herron  and  Huie  [23], using the

|gaPK ™H 1™rnas%T?trometry (M-s-} meth°d- at
298 K and 8 Torr total  pressure.   From the computer
modeling of the temporal  behaviors of several

rE6™65^6'9',' C2H"» C°2' Hz°' CH'°- HCO°H and
CH3OH, the role of free radical  mechanism involving
27 steps has been proposed.                   "ivuiy
                                                    12

-------
   The initial reactions occurring under their
experimental conditions have been postulated to be
Criegee mechanism [29] followed by the unimolecular
dissociation of the methylene peroxide, CH200, to
serveral products, i.e.,
     CH202 -* H2COO
                    67%
  H20 + CO

  H2 + C02

  2H + C02
                   -2*—>• HCOOH
   The formation of intermediate dioxirane (H2COO)
in the above scheme was predicted by Wadt and
Goddard [24], and was subsequently verified by
Lovas and Sueram [25] using microwave spectroscopy
in the low-temperature reaction of 03 with C2Hi,.

   Clearly, the degree of reliability of the above
postulated mechanism depends on the accuracy of the
input kinetic data for the series of secondary
reactions initiated by the product H atoms.  On the
basis of their modeling work, Herron and Huie [23]
have pointed out the needs for improved data for
several crucial free radical reactions occurring
in 03-C2H., system.  In particular, the mechanistic
knowledge of the HO-C2Hi, reaction is a prerequisite
to unraveling the 03-C2Hi, mechanism.  These
investigators further extended their model to high
pressure conditions and compared the computed
results with the product data on 03-C2H^ reactions
obtained by Scott et al. [26] at ppm reactant
concentrations in 1 atm air.  In general, the
agreement should be considered fair considering
the uncertainty in both measurements and model.

   Very recently, Dodge and Arnts [27] have
developed a sufficiently detailed model including
26 steps for the reactions of 03 with methyl-
substituted olefins, e.g., C3H6 and 2-C<,He, which
involve the formation and subsequent dissociation
of a "hot" CH3 CHOO radical, i.e.,
             r  2™
                ^
CH3
            HCHO
    -c-oo
                  Ozonide
                  80%
and
The above scheme for the decomposition of  the
CH3CHOO radical to several free radical species was
used to model the product studies of Niki  et a!.
[28] in the 03-air-2 butene   HCHO   air system.

   Although these recent experimental and  modeling
studies are by no means definitive, they do shed
new light on the 03-olefin chemistry, in particu-
lar, on the efficiency of radical formation.
Some of the 03-olefin mechanisms adapted for
atmospheric modeling incorporate the formation of
two free radical species for every olefin  molecule
reacted with 03 [9].  This assumption stems from
two types of mechanistic conjectures which appeared
plausible in the past.  One of these was based on
the motion that secondary ozonides are not formed
from small olefins in air because of chemical
instability of the corresponding Criegee interme-
diates, possibly reacting with 02 [26].  The recent
observation of propene ozonide in a mixture of
03-2-C^He-HCHO-air precludes this possibility [28].
The other stems from a theoretical treatment of
gas phase 03-olefin reactions by O'Neal and
Blumstein [30] which deviates from the Criegee
mechanism.  At present, there is little experi-
mental evidence that supports their theory uniquely.
Thus, the new information suggests the production
of fewer radical species than has been assumed
previously.

              4.  Conclusion

   Experimental methodologies for the determination
of the kinetics of HO-olefin reactions appear to
be well established, and should be applied more
extensively to numerous olefins under wider range
of temperature and diluent pressure.   In contrast,
derivation of "true" rate constants for 03-olefin
reactions from the decay rates of the reactants may
suffer from the interferences due to secondary free
radical reactions initiated mainly by "Criegee
intermediates," and thus requires better
mechanistic knowledge.  To elucidate the mechanisms
for 03-olefin reactions, temporal  behavior of the
reactants and products should be better characteriz-
ed first.  These data can then be utilized for
mechanistic modeling.  The HO-initiated oxidation
of olefin is likely to play a major role in the
secondary reactions of 03-olefin system.   Thus,
mechanistic studies of HO-olefin reactions in the
presence of 1 atm air are prerequisite to the
understanding of 03-olefin reactions, and should
assume the highest research priority.
   CH
             07*
                   \°   T
                   [H cocH3J
      50%
 CH3 + CO + HO


 CH3 + C02 + H




  HCO  + CH30


5  H  +  C02  + CH3
References

[1]   Demerjian, K. L., Kerr, J.  A., and Calvert,
      J.  C.,  Adv.  Environ.  Sci. Technol. 4_, 1
      (1974).

[2]   Dodge,  M. C., Effect of Selected Parameters
      on  Predictions of a Photochemical  Model,
      EPA-600/3-77/148 (June, 1977).

[3]   Echenroeder, A., Development of an Air
      Quality Model Based on the LARPP Data, (in
      progress).

[4]   Parker, R. 0. and Martinez, J. R., Los
      Angeles Reactive Pollutant Program (LARPP)
      Data Archiving and Retrieval, Document No.
                                                   13

-------
      P-1464-W, 1974, Environmental Research and
      Technology, Inc.  Concord, Massachusetts,
      (NTIS Accession No. PB 244-298).

[5]   Garvin, D. and Hampson, R. F. eds., Chemical
      Kinetics Data Survey VII, Table of Rate and
      Photochemical Data for Modelling of the
      Stratosphere [Revised], NBSIR 74-430,
      National Bureau of Standards, Washington,
      D.C.  20234 (May, 1973).

[6]   Leighton, P. A., Photochemistry of Air
      Pollution  (Academic Press, New York, 1961).

[7]   Calvert, J. G., Environ. Sci. Techno!. 10.,
      (1976).

[8]   Japar,  S. M., Wu,  C. H., and Niki, H.,
      J.  Phys.  Chem. 78, 2318  (1974).

[9]   Niki,  H.,  Daby, E. E.,  and Weinstock, B.,
      Adv. Chem.  Ser. 113, 16  (1972).

[10]  Atkinson,  R.,  Darnall,  K. R., Lloyd, A. C.,
      Winer,  A.  M.,  and  Pitts,  J.  N., Jr., Adv.
      Photochem.  (in press).

[11]  Atkinson,  R.  and  Pitts,  J. N.,  Jr., J. Chem.
      Phys.  63.,  3591  (1975).

[12]  Ravishankara,  A.  R., Wagner, S.,  Fischer,  S.,
      Smith, G.,  Schiff, R.,  Watson,  R.  T., Tesi,
      G., and Davis,  D.  D.,  Int. J. Chem. Kinet.
       TO., 783 (1978).

 [13]   Howard, C. J.,  J.  Chem.  Phys.,  65_ 4771
       (1976).

 [14]   Cvetanovic, R.  J., 12th Int.  Symp.  Free
       Radicals, Laguna Beach, California, Jan.  4-
       9 (1976).

 [15]   Slagle, I. R., Gilbert, J.  R.,  Graham,  R.  E.,
       and Gutman, D.,  Int.  J. Chem.  Kinetics,
       Symp.  No. 1, 317 (1975).

 [16]   Meagher, J. F. and Heicklen, J.,  J.  Phys.
       Chem.  8J_, 1645 (1976).

 [17]   Atkinson, R., Perry, R. A.,  and Pitts,  J.  N.,
       Jr., J. Chem. Phys.  67., 3170 (1977).

 [18]   Bufalini, J., Walter,  T., and Bufalini,  M.,
       Environ. Sci. Technol. H,  1181 (1977).

 [19]  Carter, W. P. L., Lloyd, A.  C., Sprung,  J.
       R., and Pitts, J. N., Jr.,  Computer Modeling
       of Smog Chamber Data: Progress  in Validation
       of a Detailed Mechanism for the Photooxida-
       tion of Propene and y-Butane in Photochemical
       Smog,  Int. J. Chem.  Kinetics 11,  45 (1979).

  [20]   Niki,  H.,  Marker, P.  D., Savage, C. M., and
        Breitenbach,  L.  P., J.  Phys. Chem. 82, 135
        (1978).                            ~~

  [21]   Huie,  R.  E. and  Herron, J. T., Int. J. Chem.
        Kinetics,  Symp.  1, 165  (1975). ~

  [22]   Japar,  S.  M., Wu, C.  H., and Niki, H., ^_
        Phys.  Chem. 80_,  2057  (1976).
[23]  Herron,  J.  T.  and  Huie,  R.  E.,  J^Am
      Soc.  99_, 5430  (1977).

[24]  Wadt, W. R. and Goddard, W.  A., HI, J.  Am.
      Chem. Soc.  97_, 3004 (1975).

[25]  Lovas, F. J. and Suenram, R. 0., Chem. Phys.
      Lett. 51, 453  (1977).

[26]  Scott, W. E., Stephens,  E.  R.,  Hanst, P. L.,
      and Doerr, R.  C., Proc.  Am.  Petroleum Inst.
      37, 171   (1957).

[27]  Dodge, M. C. and Arnts, R.  R., A New
      Mechanism for the Reaction of  Ozone with
      Olefins, Int. J. Chem.  Kinetics 11, 399
      (1979).

[28]  Niki, H.,  Maker,  P. D., Savage, C.  M., and
      Breitenbach,  L. P., Chem. Phys. Lett. 415,
      327  (1977).

[29]  Criegee, R.,  Rec.  Chem.  Progs. 18,  113
      (1975).

[30]  O'Neal,  H.  E.  and  Blumstein, C., Int. J.
      Chem. Kinetics  5,  397 (1973).
               Summary  of  Session

    The discussion  centered on the reactivity of
 OH  and 03  and  the  mechanisms of reaction of OH and
 03  with olefins.   The mechanisms of OH-olefin
 reactions  were treated by Cvetanovic.   In his
 work  hydroxyl  radicals were  produced by the photo-
 lysis of N20  (to produce  O'D) in the presence of
 H2  or H20.   In the absence of 02, the  major
 products of  the reaction  can be accounted for on
 the basis  of an additions reaction followed by
 radical-radical and radical  substrate  reactions.
 There was  no evidence that abstraction was
 important  for  ethylene or propylene,  e.g., in the
 case  of the  propylene reaction only minor amounts
 of  products  characteristic of vinyl  radicals, its
 major abstraction  product, were found.   Heicklen
 discussed  the  earlier work which had suggested the
 possible importance of an abstraction  path for
 OH  +  C2H.,.  As pointed out by Niki in  his review
 paper, larger  olefins, particularly those with
 weak  allylic bonds, all could react partially
 through an abstraction mechanism.  Kinetic data
 support this argument.

    Of greater  interest is the question of the role
 of  02 in the reaction. If the initial  reaction  is
 addition,  then in  the presence of 02,  a peroxy
 radical  would  be formed.   The subsequent fate
 of  this peroxy radical is one of the key problems
 in  laboratory  studies of  atmospheric chemical
 reactions.   Cvetanovic reported that,  in the
 presence of  02, the major products of  the
 OH  +  CzH,,  reaction were CH3CHO and C2H5OH, and
 from  OH +  C3H6, the major product by far was
 CH3CHO.  These products are  not readily accounted
 for on the basis of simple radical-radical
 interactions 2R02  •*• R'CHO +  R"OH + 02,  or
 2R02  -»- 2RO + 02.   In the  atmosphere  where NO is
                                                     14

-------
available it is generally assumed that the peroxy
radicals are converted to alkoxy radicals.  The
subsequent fate of the alkoxy radicals is not
understood.  In the mechanism used by Niki (his
fig. 2) based on modeling of the Riverside data,
the alkoxy radical decomposes (a revised version
of the mechanism and figure were supplied by
Carter).  It was pointed out by Batt and by
Golden, however, that the simpler alkoxy radicals
were more likely to react with 02 (see review
paper of Golden and susequent discussion).  The
specific fate of the radical C3H6-OH-02 produced
in the reaction of OH with C3H6 in the presence
of 02 was not resolved.

   The rate constants for OH reactions are in
relatively good shape as discussed by Atkinson.
Rate constants for 03 reactions are still subject
to some uncertainty because of problems in
measurement methodology.  These problems would be
resolved if the mechanisms were better understood.
The major lack of kinetic data is for cyclic
olefins which as Niki pointed out are the most
important class with respect to consumption of
ozone by olefins.

   The mechanisms of ozone-olefin reactions have
not been settled.  Recent work, as reviewed by
Niki, indicates that free radical yields are
smaller than had been thought previously.  Whitten
emphasized that in most cases the radical yields
were the important information needed by modelers
rather than specific reaction channels.  Dodge
presented a general model for the reactions based
on both high and low pressure experimental observa-
tions.  This model appears to rule out the O'Neal-
Blumstein mechanism.   O'Neal agreed to this
interpretation, pointing out the reasons for the
discrepancy.  Golden pointed out that the high
and low pressure results used by Dodge might not
be compatible because of different quenching rates
etc., and the general mechanism should be used with
caution.  A critical  question in this discussion
is the chemistry of the Criegee intermediate,
RCHOO.  In particular what are the rates of
reaction with NO , SO , and aldehydes relative
to isomerization and scission.   Heicklen commented
on the reactions of ozone with chlorinated
ethylenes where a ir-complex may be involved.

   There was considerable interest expressed in
extending observations to more complex systems.
The importance of cyclozlkenes was noted above.
Stedman urged consideration of natural  substances
such as a-pinene and  isoprene.   O'Brien pointed
out the difference between high and low molecular
weight compounds particularly with respect to
acid and aerosol  formation.   This raise unanswered
questions as to the stabilization of the Criegee
intermediate and their role in the formation of
acids.   The origin of organic acids  is  also a
puzzling problem in photooxidation studies of
aldehydes.   Acids are also found in  low pressure
ozonoalysis studies.
                     Comments

 Roger Atkinson, Statewide Air Pollution Research
 Center, University of California, Riverside,
 California  92521

    Rate Constant Data for the Reaction of OH
 Radicals with Alkenes:  The available rate constant
 data and Arrhenius parameters are given in tables
 1  (alkenes),  2 (monoterpenes) and 3 (dialkenes)
 which are taken from the review article of
 Atkinson, Darnall, Lloyd, Winer and Pitts, to be
 published in Advances in Photochemistry.  The
 relative rate constant data of Cox [1] for
 ethane, propene and trans-2-butene, obtained from
 the photolysis of  HCHO-alkene mixtures in air at
 760 Torr total pressure,  have not been included in
 table 1 as the stoichimetry factor was not known.
 The data from relative rate studies have been
 reevaluated on the basis  of what is felt (however
 biased) by the authors to be the "best" rate
 constant for the reference reaction at the
 temperatures  employed in  the respective relative
 rate studies.   The abbreviations used for the
 techniques are:  DF-RF: discharge flow-resonance
 fluorescence;  DR-RA:  discharge flow-resonance
 absorption;   DF-ESR:  discharge flow-esr detection;
 DF-LMR: discharge  flow with laser magnetic
 resonance detection of OH;  DF-MS:  discharge  flow-
 mass spectrometry; FP-KS  flash photolysis-kinetic
 spectroscopy;  FP-RA:  flash  photolysis-resonance
 absorption;  FP-RF: flash  photolysis-resonance
 fluorescence;  MPS: modulation-phase shift; PR:
 pulsed  radiolysis.  The  tables list the literature
 data expressed in  the form of k =  A e~E/RT with k
 being given  at room temperature,  and  A and E  are
 the Arrhenius  preexponential  factor and Arrhenius
 activation energy, respectively.

    For  ethane  the  rate constant at room tempera-
 ture (and up  to  425 K) is  in  the  falloff  region
 between second order  and  third order  kinetics
 [3,6,7,27,31]  below ^ 225 Torr of argon [6]  and
 below ^ 300 Torr of helium [3].   The  rate in table
 1  for ethene are limited  to the high  pressure re-
 sults,  although the rate  constants of Greiner [2]
 (apparently obtained at 100 Torr total  pressure of
 helium) have been  included.

    For  ethane and  propene the limiting high
 pressure room temperature rate constants  appear to
 be  reasonably  well  defined  (table  1)  at about 8 x
 1Q-12 cm3 molec'1  s-1  and  2.5  x 1Q-11  cm3  molec-1
 s~l,  respectively.   For the higher alkenes the
 data  do not  appear to be  as consistent.   The  rate
 constants obtained by Ravishankara et  al.  [18]  for
 cis-2-butene  and,  especially,  for  2,3-dimethyl-2-
 butene  appear  to be low;  that  for  2,3-dimethyl-2-
 butene  [18]  being  a factor  of  2 lower  than the
 room temperature rate constants determined by
 Morris  and Niki  [19],  Perry [25],  and  Atkinson,
 Darnell  and  Pitts  [23].   It is probable that wall
 adsorption problem (as observed for propene  in a
 metal reaction cell  [18])  in  the  static system
 used  by Ravishankara  et al. [18]  is the cause of
 these apparently low rate  constants.

   It would then appear that the flash photolysis-
resonance fluorescence rate constant data of
Atkinson and Pitts   [14,22]  and Atkinson, Perry,
and Pitts [6,20] (which are generally  in good
                                                   15

-------
Table 1.  Rate constant data and Arrhenius  parameters  for the reaction of OH radicals with alkenes.
1012 x A
Alkenes cm'molec 's l E cal mol 1
Ethene 1.26 -903 ± 136
(limiting high
pressure data)
	

2.18 -770 ± 300

„

Ethene-di,


Propene
.-
—
__ __


	
._
_.
4.1 -1080 ± 300

—


—


—


--


—

Propene-de
— — _ —
1-Butene
—
7.6 -903 ± 300

—

1012 x_k _
cm3mo1ec"1s~1
5.33b
5.33 ± 0.65°
6.23 ± 0.33
7.31 ± 0.33
8.1 ± 1.6

7.85 ± 0.79

10.0 ± 1.7

8.24 ± 0.48


17 ± 4
5.0 ± 1.7
14.5 ± 2.2
13.4 ± 3.4


14.3 ± 0.7
20.0 ± 1.0
5 ± 1
25.1 ± 2.5

27.4 ± 5.5


20.8


23.5 ± 3.5


23.5 ± 4.7


25.6 ± 1.2h

18.7
16.8
40.8
15 ± 1
35.3 ± 3.6

27.0

At T K
298-301
300
381 |
416)
305 ± 2

299

296

298 ± 2


300
^ 300
298
298


381
416
300
298

305 ± 2


303


305 ± 2


305 ± 2


298

298
298
298
300
298

303

Technique
FP-KS
FP-RF
PR
Relative
rate
FP-RF

FP-RA

Relative
rate

DF-MS
DF-ESR
FP-RF
Relative
rate


PR
DF-RA
FP-RF

Relative
rate

Relative
rate

Relative
rate

Relative
rate

FP-RF

DF-MS
FP-RF
DF-MS
DF-RA
FP-RF

Relative
rate
Temperature
range
Reference covere
-------
Table 1.  Rate constant data and Arrhenius parameters for the reaction of OH radicals with alkenes.  (continued)
1012 x A
Alkenes cm3molec"Is"
Isobutene
9.2

__


cis-2-Butene
10.4

„


	


	

trans-2-Butene
—
11.2

__



1-Pentene
__



2-Methyl-
1-butene


3-Methyl- 5.23

2-Methyl-
2-butene 36

19.1

__



cis-2-
Pentene
1012 x k
'ia E cal mol"1 cirtolec'V
64.6
-1000 ± 300 50.7 ± 5.1

47.8


61.2
-970 ± 300 53.7 ± 5.4

61.5 ± 12.3


58.6 ± 8.8


42.6 ± 2.5h

71.4
12 ± 10
-1090 ± 300 69.9 ± 7.0

67.6



42.5
29.1



90.1
57.2


-1060 ± 300 31.0 ± 3.1

119
-450 ± 400 78 ± 8

-895 ± 300 87.3 ± 8.8

87 ± 6



62.4

At T K
298
297

303


298
298

305 ± 2


305 ± 2


298

298
300
298

303



298
303



298
303


299

298
298

299

300 ± 1



303

Technique
DF-MS
FP-RF

Relative
rate

DF-MS
FP-RF

Relative
rate

Relative
rate

FP-RF

DF-MS
DF-RA
FP-RF

Relative
rate


DF-MS
Relative
rate


DF-MS
Relative
rate

FP-RF

DF-MS
FP-RF

FP-RF

Relative
rate


Relative
rate
Temperature
range
Reference covered
Morris & Niki, 1971 [19]
Atkinson & Pitts, 1975 297-424
[14]
Wu, Japar & Niki, 1976 [15]
(relative to OH + cis-2-
butene 5.20 x 10"11)9
Morris & Niki, 1971 [19]
Atkinson & Pitts, 1975 298-425
[14]
Lloyd et al . , 1976 [5]
(relative to OH + n- A
butane 2.82 x I0"12)a
Winer et al., 1976 [16]
(relative to OH +
isobutene = 4.80 x 10"11)
Ravishankara et al . , 1978
[18]
Morris & Niki, 1971 [19]
Pastrana & Carr, 1975 [13]
Atkinson & Pitts, 1975 298-425
[14]
Wu, Japar & Niki, 1976
[15] (relative to OH +
cis-2-butene =
5.20 x 10'11)9
Morris & Niki, 1971 [19]
Wu, Japar & Niki, 1976
[15] (relative to OH +
cis-2-butene =
5.20 x lo"11)9
Morris & Niki, 1971 [19]
Wu, Japar & Niki, 1976 [15]
(relative to OH + cis-2-
butene = 5.20 x lo"11)9
Atkinson, Perry & Pitts, 299-423
1977 [20]
Morris & Niki, 1971 [19]
Atkinson, Perry & Pitts, 298-425
1976 [21]
Atkinson & Pitts, 1978 299-426
[22]
Atkinson, Darnall & Pitts,
1978 [23] (relative to
OH + cis-2-butene =
5.29 x 10"11)9
Wu, Japar & Niki, 1976
[15] (relative to OH +
/-•ie_9_ kn+ana -

K





K











K














K


K

K







                                                                               5.20 x lo"'1'1)9
                                                        17

-------
Table 1.  Rate constant data and Arrhenius parameters for the reaction of OH radicals with alkenes.3 (continued)
Alkenes
   1012 x A                    1012 x k
cm'molec'V13  E cal mol"1  cm'molec'V1  At T K  Technique
                                                                                  Reference
                                                                                                         Temperature
                                                                                                            range
                                                                                                            covered
2-Pentene
  (mix. cis,
  trans)
                                            90.1
                                                             298
                                                                    DF-MS
                                                                 Morris & Niki, 1971 [19]
 1-Hexene
                                            31.2
                                              303    Relative    Wu, Japar & Niki, 1976
                                                      rate       [15] (relative to OH + cis-
                                                                 2-butene = 5.20 x 10 11)M
 Cyclohexene
                             62.4              303     Relative     Wu,  Japar  & Niki,  1976
                                                       rate        [15] (relative  to  OH +
                                                                  cis-2-butene =
                                                                  5.20 x 10"11)9

                             73  4  ±  14.7     305  ±  2   Relative     Darnall  et al., 1976 [24]
                                                       rate        (relative  to OH +  iso-
                                                                  butene = 4.80 x lo"11)9
 3,3-Dimethyl-
  1-butene
                                            27.0
                                              303    Relative    Wu, Oapar & Niki,  1976
                                                      rate       [15]  (relative  to  OH +
                                                                 cis-2-butene  =
                                                                 5.20  x  10"11)9
 2,3-Dimethyl-
  2-butene
                               153

                               110 ± 22

                              56.9 ± 1.3h


                               122 ± 8
298

298
DF-MS

FP-RF
                                                             298    FP-RF


                                                           300 ± 1  Relative
                                                                     rate
Morris & Niki, 1971 [19]

Perry, 1977 [25]

Ravishankara et al., 1978
[18]
Atkinson, Darnall  & Pitts, 1978
[23] (relative to  OH + cis-2'
butene = 5.29 x 10"11)9
 1-Heptene
                              35.0  ±  7.0      305  ±  2   Relative     Darnall  et  al.,  1976  [24]
                                                       rate        (relative to  OH  + isobutene
                                                                  4.80  x  lo"11)9
  1-Methyl-
  cyclohexene
                                            91.7 ± 18.3    305 ± 2
                                                      Relative
                                                       rate
                   Darnall et al.
                   (relative to 01
                   4.80 x 10"11)9
                            1976 [24]
                            + isobutene
 *Mean Arrhenius preexponential  factor.
  Total pressure not stated,  but stated  to  be  the  same as  in previous work  [26],  i.e.,  100 Torr  of  helium.
 jEssentially the high pressure  limit from  a Lindemann plot [27].
  Calculated from the Arrhenius  expression  of  reference  [28] for T   305  K.
 ^Calculated from the Arrhenius  expression  of  reference  [6] for T    298 K.
  Rate constant determined for OH + CO at room temperature and low pressure [29].
 ^Calculated from the Arrhenius  expression  of  reference  [14] for T = 300, 303 or  305  K.
  Rate constants at 20 Torr total pressure  with helium as  the diluent gas.   No  pressure effects  were
  observed over the total  pressure range 3-20  Torr (1-butene and cis-2-butene)  or 20-200 Torr  (propene).

    In addition to the above data, Simonaitis  and  Heicklen [30] obtained  rate constant  data  for  propene
  relative to those for the reaction of  OH  radicals  with CO at 373 and 473  K at total pressures  of  ^  400-
  800 Torr (mainly H20).   They obtained  k(OH + C3H6)/k(OH  + CO) (± 103S) = 75 at 373 K and 55 at  473 K.
  Assuming that k(OH + CO) = 3.0 x 1Q-13 cm3 molec-is"1, independent of temperature under these  conditions
  (which is subject to large uncertainties), then  by extrapolation a value  of k(OH +  C3H6)   3.3 x  10"11
  cm3 molec"1*'1 at 298 K and an Arrhenius  activation energy  of E    -1090 cal moT1 may be  obtained.   It
  is evident that this data (which is obviously subject  to large uncertainties  because  of the  assumptions
  made) is in general agreement  with that obtained by references [5.14-18J, as  quoted in the table  above.
                                                         18

-------
                   Table 2.  Rate constant data for the reaction of OH radicals with monoterpenes.
                   Terpene
                                  1011 x k
                               cm'molec"^"1
                              At T K    Technique
                                                         Reference
                   a-Pinene
                   6-Pinene
                                5.5  ± 0.£
                                6.4  ± 1.0
                   d-Limonene   14.2  ± 2.1
                             305  ± 2    Relative    Winer et  al., 1976 [16]
                                        rate       (relative to OH + isobutene
                                                  = 4.80 x  10"11)

                             305  ± 2    Relative    Winer et  al., 1976 [12]
                                        rate       (relative to OH + isobutene
                                                  = 4.80 x  10"11)3

                             305  ± 2    Relative    Winer et  al., 1976 [16]
                                        rate       (relative to OH^t isobutene
                                                                     = 4.80 x 10
                                                                               -11 \a
                   aCalculated from the Arrhenius expression of reference [14] for T   305 K.
 Table 3.  Rate constant data  and Arrhenius parameters for the reaction of OH radicals with dialkenes.
                 1012 x A                  1012  x  k
 Dialkenes     cm3molec~1s~1   E cal mol"1  cra3molec"1s"1  At T  K  Technique
                                                         Reference
Temperature
   range
  covered
 Propadiene
 1,3-Butadiene
 —           —       4.5  ± 2.5    T 300    DF-ESR      Bradley et al ., 1973
                                                       [10]

5.59       -305 ±300   9.30+0.93     299    FP-RF       Atkinson, Perry, &
                                                       Pitts, 1977 [20]

 —           —       72.8 ± 14.6   305 ± 2  Relative    Lloyd et al .,  1976 [5]
                                             rate       (relative to OH + n-
                                                       butane = 2.82  x lo"12)c
                 14.5
          -930  ± 300   68.5 ± 6.9      299    FP-RF       Atkinson, Perry, &
                                                       Pitts, 1977  [20]
                                                                                                  299-421 K
                                                                                                  299-424 K
 ?Mean Arrhenius  preexponential factor.
 °Hay be in the fall-off region between  second order and third order kinetics (see text and reference [20]).
  Calculated from the Arrhenius expression of reference [30]  at 305 K.
agreement with  the relative rate  data [5,15-17,23]
and with the  discharge flow-mass  spectrometric
data of Morris  and Niki [19]), together with the
rate constant for 2,3-dimethyl-2-butene recently
obtained by Atkinson, Darnall, and  Pitts [23]
from a relative rate study, should  be viewed as
the most consistent set of rate constant data.
This is especially so as this set of absolute rate
constant data [14,20,22] also comprises the only
temperature dependence studies for  the alkenes
other than for  ethene.

   Finally, it  should also be noted  that with the
flash photolysis  systems used to  determine OH
radical rate  constants for the alkenes, problems
have been encountered due to secondary reactions
and due to wall  absorption of the reactants.
Thus, although  the [reactant]/[OH]  ratios are
reasonably similar with FP-RA and FP-RF, because
of the higher flash energies  used with RA
detection  (^  1000 joules per  flash  compared with
1 100 joules  per flash for RF detection), second-
ary reactions of OH radicals with the larger
amounts of photolysis products generated by the
more intense  flash may be encoutered with flash
photolysis-resonance absorption systems.
                                                            References

                                      [1]  Cox, R. A.,  Int.  J.  Chem. Kinet.  Symp.  No. 1,
                                           378 (1975).

                                      [2]  Greiner, N.  R.,  J.  Chem. Phys.  53,  1284 (1970)

                                      [3]  Davis, D.  D.,  Fischer, S., Schiff,  R.,  Watson,
                                           R.  T., and Bellinger, W. , J. Chem.  Phys.  63,
                                           1707 (1975).

                                      [4]  Gordon, S. and Mulac, W. A.,  Int. J.  Chem.
                                           Kinet., Symp.  No. 1,  289 (1975T

                                      [5]  Lloyd, A. C.,  Darnall, K. R., Winer,  A.  M.,
                                           Pitts, J. N., Jr.,  J. Phys.  Chem. 80
                                           and Pitts, J. N., Jr., J. Phys. Chem. 80,
                                           789 (1976).

                                      [6]  Atkinson, R.,  Perry,  R. A., and Pitts,  J.  N.,
                                           Jr., J. Chem.  Phys.  66., 1197  (1977).

                                     [7]   Overend,  R.  and Paraskevopoulos, G., J.  Chem.
                                           Phys.  67.,  674 (1977).

                                     [8]   Niki,  H.,  Maker. P. D., Savage, C. M.,  and
                                           Breitenbach, L. P., J.  Phys.  Chem. 82_,  132
                                           (1978).
                                                     19

-------
 [9]   Morris,  E.  D., Jr., Stedman, D. H., and
      Niki,  H.,  J.  Amer.  Chem. Soc. 93_, 3570 (1971).

"10]   Bradley, J.  N., Hack, W., Hoyermann, K., and
      Wagner,  H.  Gg., J.  Chem. Soc. Faraday Trans. I
      69_,  1889 (1973).
Til]  Stuhl, F., Ber. Bunsenges. Phys. Chem. 77.,
      674 (1973).

[12]  Gorse, R. A. and Volmam, D. H., J. Photochem.
      3.. 115 (1974).

[13]  Pastrana, A. V. and Carr, R. W., Jr., J. Phys.
      Chem.  79., 765  (1975).

[14]  Atkinson, R. and Pitts, J.N., Jr., J. Chem.
      Phys.  63, 3591 (1975).

[15]  Wu, C. H., Japar, S. M., and Niki, H., J^
      Environ. Sci.  Health All, 191 (1976).

[16]  Winer, A. M.,  Lloyd, A. C., Darnall, K. R.,
      and Pitts, J.  N., Jr., J. Phys.. Cham. 80.,
      1635  (1976).

[17]  Winer, A. M.,  Lloyd, A. C., Darnall, K. R.,
      Atkinson, R.,  and Pitts, J. N., Jr., Chem.
      Phys. Lett. 51_, 221  (1977).

[18]  Ravishankara,  A. R., Wagner, S., Fischer,  S.,
      Smith, G.,  Schiff,  R., Watson,  R. T., Test,
      G., and  Davis, D. D.,  Int. J. Chem.  Kinet..
      in press  (1978).

[19]  Morris,  F..  D., Jr.  and Niki, H., J.  Phys.
      Chem. 75..  3640 (1971).

[20]  Atkinson,  R.,  Perry,  R. A., and Pitts, J.  N.,
      Jr.,  J.  Chem.  Phys.  B7_, 3170 (1977).

 [21]  Atkinson, R.,  Perry, R. A.,  and Pitts, J.  N.,
      Jr.,  J.  Chem.  Phys. 38., 607  (1976).

 [22]  Atkinson, R.  and  Pitts, J.  N.,  Jr.,  J. Chem.
      Phys.  68, 2992 (1978).                     '

 [23]  Atkinson, R.,  Darnell,  K.  R., and  Pitts,  J.  N.,
      Jr.,  J.  Phys.  Chem., submitted  for publication
       (19787:

 [24]  Dar-nall, K.  R.,  Winer,  A.  M., Lloyd, A. C.,
      and  Pitts,  J.  N.,  Jr.,  Chem. Phys. Lett.  44,
      415  (1976).                              ~

 [25]  Perry,  R.  A.,  Ph.D. Thesis,  University of
      California,  Riverside,  August 1977.

 [26]  Greiner, N.  R.,  J.  Chem.  Phys.  51_, 5049 (1969).

 [27]  Palmer,  H.  B., J.  Chem. Phys. 64_,  2699  (1976).

 [28]  Perry,  R.  A.  Atkinson,  R., and  Pitts, J. N.,
      Jr.,  J.  Chem.  Phys.  64, 5314 (1976).
                                                     Julian Heicklen, Department of Chemistry, The
                                                     Pennsylvania State University, University Park,
                                                     Pennsylvania  16802

                                                        All workers agree that at room temperature and
                                                     atmospheric pressure the predominant reaction of
                                                     HO with CJ.H,, is addition.  The only evidence for
                                                     abstraction is given by Meagher and Heicklen (J_.
                                                     Phys. Chem. 80_, 1645, 1976) who photolyzed H202 at
                                                     2537 A to produce HO radicals.  They found C2H5OH
                                                     as the major product at high pressures and that
                                                     this product became less important as the pressure
                                                     was reduced, as expected for the pressure-sensitive
                                                     addition reaction of HO with C2H4.

                                                        However they also found that CH20 and HCOOH were
                                                     produced in a pressure-insensitive reaction.  From
                                                     this they concluded that abstraction occurred 26
                                                     percent of the time for  [C2Hi,] ^ 2 to 5 Torr,
                                                     [H202] ^ 2 Torr, and N2 = 40 Torr.  They further
                                                     estimated that in the high pressure limit, this
                                                     fraction would further drop below 22 percent.
                                                     Since then, the high pressure limit rate constant
                                                     has been evaluated, and this fraction becomes 7
                                                     percent at the high pressure limit.

                                                        This value is still higher than indicated by
                                                     some other studies, and may be due to energetic
                                                     HO radicals in Meagher and Heicklen's system,
                                                     since the photolyzing radiation provices an excess
                                                     of '\> 23 kcal/mol over that needed to photo-
                                                     dissociate H202.  It is apparent that if the HO
                                                     radicals become significantly energetic (either
                                                     thermally or by other means), and the pressure is
                                                     low enough, then the abstraction reaction must
                                                     become dominant.  Thus pressure and temperature
                                                     studies should be done to determine the conditions
                                                     when the two competitive paths are important.
[29]


[30]


[31]  Howard,  C.  J.,  J.  Chem.  Phys.  615,  4771  (1976).
       Perry, R.  A.,  Atkinson,  R.,  and Pitts,  J.N.,
       Jr., J.  Chem.  Phys.  67.,  5577 (1977).

       Simonaitis,  R.  and Heicklen, J..  Int.  J.  Chem.
       Kinet. 5,  231  (1973).
                                                     William P. L. Carter,  Statewide  Air Pollution
                                                     Research Center, University  of California,
                                                     Riverside, California   92521

                                                         At the  present time, the major uncertainties we
                                                      have found in developing and validating the
                                                      mechanisms for the OH-olefin system concerns the
                                                      rate constant for the decomposition of B-substitut-
                                                      ed alkoxy  radicals, relative to the rate of their
                                                      reaction with 02.   For example, in the OH-propene
                                                      system, the question concerns the rates of
                                                                   OH 0

                                                                CH3CH-CH2 ->• CH3CHOH'+ HCHO
                                                                                                      0)
                                                                   0  OH

                                                                CH3CH-CH2 + CH3CHO + CH2OH          (2)

                                                      relative to the competing reactions with  atmo-
                                                      spheric 02, forming B-hydroxy carbonyl  products
                                                      (see Niki's figure 21)-  There appears  to be some
                                                        Editor's note.  Figure 2 in Dr. Niki's review
                                                        paper is a revised version provided by Dr  C
                                                        The change does not significantly affect Dr
                                                        Niki's conclusions.
                                                   20

-------
 conflicts concerning this.  The study of Niki et
 al.  [1]  on olefin-HONO-NO systems indicates that
 both reactions are fast, while we found, in
 modeling [2] the more recent U.C. Riverside smog
 chamber  data [3] that the acetaldehyde and
 formaldehyde yields in propene-containing NO -air
 systems  are better fit by models which assume that
 reaction (1) is slow and reaction (2) is fast.
 (The assumption that reaction (2) is much faster
 than reaction (1) is consistent with the
 theoretical estimates of Baldwin et al. [4]).
 Assuming that both reactions are fast results in
 the model overpredicting the most recently
 determined acetaldehyde and formaldehyde yields
 by ^ 25 to 50 percent.

    Our results [2] could be reconciled with those
 of Niki  et al. [1] if there were systematic
 calibration errors in the determination of the
 acetaldehyde and formaldehyde yields in the UCR
 smog chamber runs.  However, the yields of these
 products monitored in n-butane-NO -air UCR chamber
 runs using the same techniques [3j agree well with
 our n-butane model predictions [2].  Since in the
 n-butane system, the photooxidation mechanism is
 less uncertain; this tends to indicate that the
 reported yields of these products in the recent
 UCR chamber experiments are probably not in error.

     It should be noted that our smog chamber
 modeling results are completely  inconsistent with
 both reactions (1) and  (2) being slow.  Not only
 do models assuming this consistently underpredict
 acetaldehyde and formaldehyde, they also over-
 predict overall reactivity because both acetol and
 2-hydroxy propanal, which would  be the major
 products if reactions (1) and (2) were slow, are
 expected to react with  OH to form methylglyoxyl

              OH                     OH

           CH3CHCHO  +  OH  ->- H20 +  CH3C-CHO


              OH                      0

           CH3C-CHO  +  02  + H02 +  CH3-C-CHO


 whose rapid  photolysis would contribute significan-
 tly  to radical initiation  [2].


 References

 [1]  Niki, H., Maker, P.  D., Savage, C. M., and
     Breitenbach, L.  P.,  J. Phys. Chem. 82, 135
     (1978).                            ~~

 [2]  Carter, W. P. L., Lloyd, A.  C., Sprung, J. L.,
     and Pitts, J. N., Jr., Computer modeling of
     smog chamber data:  Progress in validation of
     detailed mechanisms for the  photooxidation of
     propene and n-butane in photochemical smog,
     Int. J. Chem. Kinetics 11, 45  (1979).

 [3]  Darnall,  K.  R.,  Winer, A. M.,  and  Pitts, Jr.,
     J.  N., A  smog  chamber  study  of  the  propene-n-
     butane-NO  systems,  in preparation  (1978).
               A
[4]  Baldwin, A.  C., Barker, J.  R.,  Golden, D.  M.,
     and  Hendry,  D.  G., J. Phys.  Chem.  8]_, 2483
     (1977).
   L.  Batt,  Chemistry Department, University of
   Aberdeen,  Aberdeen,  Scotland  AB9 2UE

      By analogy with our studies on the decomposi-
   tion and  other reactions of alkoxy radicals we
   are able  to make some prediction about the two
   unimolecular steps
               0
OH
I
         CH3  - C - CH2 + CH3CHO + CH2OH
               I
               H


               OH  6
               I   I
          CH  - C - CH2 + HCHO + CH3C-OH
               I
               H


   given by Dr.  Niki  in his figure 2.   The predic-
   tions indicate that the decompositions compete
   with difficulty with their reaction with oxygen,
   if at all.   (See more detailed comments in
   session on free radical  chemistry).
   Marcia C. Dodge, Environmental Protection Agency,
   Research Triangle Park, North Carolina  27711

      We recently developed a mechanism for the
   propylene-03  reaction to use in our modeling
   studies.   The mechanism we formulated is based on
   the results  of the two most recently published
   studies  of olefin-03  reactions.   These are Herron
   and Huie's study of the ethylene-03 reaction and
   the Niki  ert  al_.  study of the reaction of cis-2-
   butene with  03 in the presence of HCHO.   The
   results  of both  of these studies  can be  explained
   in  terms  of  the  Criegee mechanism, which is given
   below for the propylene-03 reaction:
   CHsCH = CH2 + 03 —> CH3CH - CH2
                                50%
                                50%
                                   CHaCH
                                   HCH +
                                             :-oo
      We used the mechanism developed by Herron and
   Huie to explain the fate of the HCHOO radical.
   Our treatment of the CH3CHOO radical is based on
   an analysis of the product yields obtained by Niki
   et_ al_.  in their study.   In our model, a fraction
   of the  "hot" CH3CHOO radical is assumed to be
   collisionally-stabilized at atmospheric pressure.
   The rest of the biradical  undergoes rearrangement
   to form a "hot" acid and ester.  The acid and
   ester subsequently decompose to various free
   radical species.

      This mechanism was used to model data collected
   in our  laboratory on the ozonolysis of propylene.
   Four experiments were conducted in Teflon bags in
   air at  atmospheric pressure.  An example of the
   type of fits obtained when we modeled these data
   is  shown  in  figure 1.   The simulated propylene
   and 03  decay curves are in good agreement with
21

-------
  Fig.  1.  Experimental and  simulated  results
           using  the  new mechanism  for olefin-
           ozone  reactions.

the experimental  profiles.   The  fits obtained for
the other three experiments  were equally  as  good.

   Although the new mechanism adequately  explains
the observed decay of propylene and 03, the
mechanism favored by many modelers  in  the past
does not fit the  data.  In the last few years, many
model ars have used a mechanism based on the  O'Neal
and Blumstein treatment of olefin-Oj reactions.
In this mechanism, the primary ozonide, after ring-
opening, can undergo  a number of rearrangements,
the most likely of which is  a-hydrogen  abstraction
to form unstable  hydroperoxides.  These peroxides
can then fragment to  an aldehyde and two  free
radicals:
       I    I
     CHsCH -CH2
                        OOH

                  -» CHaCHCH

                         O
   O
   II
.CHsCH
        OH
 O
 II
HC
                        OOH       O         O
                         I         II         II
                    . CH3CCH2 —» CH3C  + OH + HCH

                       0
When this mechanism was used to model  the data, we
obtained the type of fit shown in figure 2.   The
simulated rate of propylene disappearance is
significantly faster than the observed rate of
                    TIME, minutes

Fig.  2.   Experimental  and simulated results
         using mechanism based on the  O'Neal-
         Blumstein treatment of olefin-ozine
         reactions.
                                                      loss.   Clearly, the data do not support  this  treat-
                                                      ment  of propylene-03 chemistry.

                                                         Although the mechanism developed  in this  study
                                                      adequately explains the observed decay of
                                                      propylene and 03, the results should not be
                                                      construed as definitive.  Additional work  is
                                                      needed  in order to fully elucidate the mechanism
                                                      of  ozone-olefin reactions.
                                                    W.  Tsang,  Center for Thermodynamics and Molecular
                                                    Science,  National  Bureau of Standards, Washington,
                                                    D.C.   20234

                                                       Many modeling studies appear to have as their
                                                    goal  the matching of some particular set of
                                                    experimental  results (smog chamber data).
                                                    Considering the non-existence of a proper data
                                                    base  for such an effort, it is not clear what such
                                                    fits  demonstrate.   Certainly with the available
                                                    number of  adjustable parameters it does not take
                                                    a  very ingeneous investigator to fit the data.
                                                    To  the unwary it may well  appear that the  entire
                                                    problem has been solved.  We know that this is not
                                                    the case  and  it would be more worthwhile to high-
                                                    light disagreements and inability to fit the data.
                                                    This  will  immediately highlight the important
                                                    questions  that must be settled.
                                                    H.  Edward  O'Neal,  Department of Chemistry,  San
                                                    Diego  State  University,  San  Diego,  California
                                                    92115

                                                       Concerning the  O'Neal-Blumstein  mechanism, it
                                                    should be  noted that  in  the  original  formulation,
                                                    the rate of  reaction  from  the molozonide  to the
                                                    Criegee intermediate  was estimated  using  (for
                                                    an  analogy)  the then  available  t-butoxy radical
                                                    decomposition rate constant.  This  is  now known
                                                    to  have an A-factor about  100 times  higher  than
                                                    that used  in the estimate.   The Criegee reaction
                                                    pathway is therefore  corresponding  more important,
                                                    and relative to the competing intramolecular
                                                    H-abstraction pathways,  is now  expected to  be the
                                                    dominant process under many  reaction  conditions.
                                                      Gary  Z.  Whitten,  Systems  Applications, Inc.,
                                                      San Rafael,  California   94903

                                                        In an ozone-olefin reaction,  it is the yield of
                                                      free  radicals  that  is important  and not necessarily
                                                      the kind of  radicals.   Our  recent modeling work
                                                      indicates  that for  ethylene the  yield of free rad-
                                                      icals should not  exceed about 10 percent.  We were
                                                      very  pleased to see the results  of Herron and Huie
                                                      which confirm  that  estimate.   In the case of  the
                                                      prophlene  reaction  we were  pleased to see Niki's
                                                      recent results which indicated a 30 percent radi-
                                                      cal yield, and Dodge's  estimate  of about 38 per-
                                                      cent.  In  our  air models  we are  using a value of
                                                      30 to 35 percent.   Anything greater leads to  a
                                                      marked decay of propylene.
                                                   22

-------
Julian Heicklen, Department of Chemistry, The
Pennsylvania State University, University Park,
Pennsylvania  16802

   The reaction of 03 with olefins proceeds by
first forming the primary ozonide which then
decomposes to give either the Criegee zwitterion
plus a carbonyl compound or free radicals.
However 03 does react with olefins to form a
reversible ir-complex (E. Sanhueza, I. C. Hisatsune,
and J. Heicklen, Chem. Rev., 7£, 801, 1976).
There is no evidence that this species plays any
role in olefins containing only carbon and hydro-
gen.  However this may not be the case with chlor-
nated ethylenes, where kinetic evidence with
CHC1CHC1 gives a reaction rate law second order in
both CHC1CHC1 and 03 at very low pressures.  This
was interpreted as meaning that the reversible
Tr-complex was the active species.  This view is
supported by the fact that the primary ozonides
were not seen in the reactions of 03 with cis- and
trans-CHClCHCl, CH2C12, and C2CU (I. C. Hisatsune,
L. H. Kolopajilo, and J. Heicklen, J. Am. Chem.
Soc., 3704, 1977).  The possible role of the
ir-complex should be considered in 03 reactions
involving substituted olefins.
                Recommendations

          Reactions of Ozone and Hydroxyl
               Radicals with Olefins
   Reactions of olefins with ozone and hydroxyl
radicals are of fundamental  importance for the
chemistry of photochemical  smog and although sub-
stantial progress is being  made in this field,
much further work remains to be done.   This work
should involve both determinations of rate
constants and mechanistic studies.  The latter
should be based on detailed product analysis,
supplemented by computer modeling.

   It is convenient to discuss ozone-olefin and
OH-olefin reactions separately.


        1.   Reactions  of Ozone with Qlefins


A. Mechanism of ozone-olefin reactions

   While significant progress is  being made in the
investigation of the mechanism of ozone-olefin
reactions,  it is clear that the mechanism is not
yet fully understood.   Its  full  understanding is
of crucial  importance  for an understanding of the
chemistry of photochemical  smog.   Some of the
chemistry involved in  the 03-olefin interaction  in
the gas  phase is probably the same or  is similar
to the chemistry of the oxygenated free radicals
reacting with 02, as for example  the radicals
formed by addition of OH to  olefins in the presence
of 02.   Further progress in  this  very  difficult
filed will  require therefore imaginative studies
not only of ozone-olefin reactions but also of the
reactions of 02 with the oxygen containing free
radicals produced in these  systems.  Photolysis  of
 organic acids and esters and generation of
 selected oxygenated free radicals by other means
 are examples of the techniques which could be
 utilized for this purpose in future work.

 Recommendations:

    1)   Selected ozone-olefin reactions, including
 cycloolefins, should  be  investigated under atmo-
 spheric conditions  over  a wide range of experiment-
 al  parameters and with time  resolved analysis of
 the concentrations  of the reacting  species and as
 many products as  possible.

    2)   Techniques should be  developed  to generate
 and study the chemical behavior of  the Criegee
 intermediates in  the  gas phase.

    3)   New approaches should be explored for study-
 ing the reactions of oxygen containing free
 radicals under atmospheric conditions to obtain
 information needed to understand the mechanism of
 ozone-olefin reactions  in the gas phase.

    4)   The formation of aerosols induced by ozone-
 olefin reactions should be studied, the key
 intermediates isolated,  and the critical chemical
 reactions studied.

 B.  Rates of ozone-olefin reactions

    The phenomenological  "rate constants" of the
 reactions of 03 with a  number of simple terminal
 olefins in the gas  phase, show good mutual agree-
 ment and describe well  the rates of consumption of
 these  olefins.  However, their exact relation to
 the "true" bimolecular  rate constants will only be
 resolved when the mechanism of ozone-olefin
 reactions in the gas phase becomes  fully under-
 stood.  The "rate constants" for internal  olefins
 measured in different laboratories  show greater
 discrepancies.  These discrepancies may be largely
 due to the fact that the range of experimental
 conditions has not  been  sufficiently broad to
 establish potential  trends  in the values.   A better
 understanding of  the  reaction mechanism will  no
 doubt  also help to  resolve  these discrepancies.

    The  difference between the data obtained at
high 03-olefin concentrations in the gas phase and
in  non-polar  solvents those obtained at low 03-
olefin  concentrations in the gas phase and in non-
polar  solvents those obtained at low 03-olefin
concentrations under conditions similar to those
in  the  polluted troposphere is puzzling.  Further
work with the object of resolving this discrepancy,
while not of the  highest priority, could help in
the understanding of the mechanism of the 03-
olefin  reactions  in the  gas phase.

Recommendations:

    1)   Measurements of the rates of 03-olefin
reactions in the gas phase should be extended to
cover a substantially broader range  of experimental
conditions.

    2)   In view of the reportly very  large 03
consuming effect of some olefins (e.g. cyclo-
olefins, terpenes) in the atmosphere, their
reaction rates should be redetermined.
                                                  23

-------
  2-  Reactions of Hydroxyl Radicals with Oleflns

A. Mechanism of OH-olefin reactions

   The mechanism of the OH reactions with ethylene
and propylene in the absence of 02 appears now to
be reasonably well understood.  Hydroxyl radicals
add to these two olefins and there is little or no
H atom abstraction at room temperature.  (A sugges-
tion that there is approximately 8 percent H atom
abstraction from ethylene at atmospheric pressure,
possibly due to "hot" OH radicals, is given in a
separate comment further below).  The mechanism
of OH-olefin reactions in the presence of 02, a
process of crucial importance for the chemistry of
photochemical smog, is unfortunately very in-
completely understood.

Recommendations:

   1)  Study of the mechanism of the OH-olefin
reactions in the absence of 62 should be extended
to olefins other than C2Hi, and propylene,
especially to the olefins known to be present in
the polluted atmosphere.

    2)  Very  high  priority  should be attached to
detailed studies  of the OH-olefin_reac_tjons_ in_the
 presence of  02,  especiaTly'for  oTefTns Tnown to be
 present  in polluted atmosphere.

    3)  Studies  of the  OH-olefin reactions under
 atmospheric  conditions  in  the  presence of varying
 amounts  of NO   (and  possibly  of other  pollutants,
 such as  SOa,  etc.)  should  be  carried out with as
 detailed  analysis of  the  reaction products  as
 possible.

 B.  Rates  of  OH-olefin  reactions

    Good  experimental  techniques for  the  determina-
 tion of  OH-olefin reaction rates  are now available.
 However,  caution has to be exercised to  assure
 accurate determination of the very  small  reactant
concentrations used in some experiments and to
establish the extent of the interfering secondary
reactions, in particular of the OH-free radical
secondary reactions.

    2)  A  set of accurate values of the rate
constants of the OH reactions with selected
olefins  (including ethylene under conditions
similar  to those in the lower atmosphere would be
desirable in order to  establish whether the rates
are affected  by oxygen  in  the air.  The case  of
ethylene  is  of special  interest in this respect
because  of a  possibility of interception  (and
consumption  by reaction) of the "hot"  CH2CH2OH
radicals  by  02.  Such  an interception  could result
in  an  appreciable  increase in  the rate constant
of  the OH-C2Hi, reaction in air  relative to the
value  obtained in  laboratory measurements  in  the
absence  of 02.

    3)  An ongoing  critical  review of the  rate
constants would be very useful.

   Rate constants  for the simple olefins  are  now
probably known to  within ± 20 percent.   The value
of the rate constant for C2H^ at 1  atm  is
probably also accurate within ± 20 percent.   Rate
constants for higher olefins and cycloolefins  are
less satisfactory,  especially the values obtained
by the competitive  technique.   No values are
available for some  important naturally occurring
olefins such as terpenus and isoprene,  although
the latter could  be roughly estimated from the
value of the rate constant for 1,3-butadiene.   The
range of the literature values of the rate constant
for acetylene is  large  (a factor of about 5-6)
and further determinations are required.

Recommendations:

    1)  Further determinations of the rate constants
are required for higher olefins, cycloolefins,
isoprene, terpenes  and acetylene.
                                                   24

-------
Session II
Aldehydes

-------
                                 TROPOSPHERIC CHEMISTRY OF ALDEHYDES
                                             Alan C. Lloyd

                              Environmental Research and Technology, Inc.
                                        2030 Alameda Padre Serra
                                   Santa Barbara, California  93103


         This paper presents a survey of the current published literature on aldehydes, and to
      a lesser extent, the other oxygenated hydrocarbons, as related to their role in modeling
      the troposphere.  Sources, ambient levels, photochemistry, and free radicals, reactions
      of these substances are treated.

      Keywords:  Aldehyde; free radical; photolysis; reactions; review; troposphere
                1.   Introduction

   Aldehydes are major products in the oxidation
of hydrocarbons and play a rather unique role in
the photochemistry of the polluted troposphere.
For example, they  can contribute to photochemical
smog, eye irritation, and odor problems.  Their
importance has been recognized for over a decade
(Leighton, 1961; Altshuller and Cohen, 1963;
Altshuller and Bufalini, 1965).  While significant
progress has been  made in defining the photo-
chemistry, kinetics, and mechanism of aldehyde
photooxidation, much remains to be learned about
their ambient concentrations as a function of
time, season and location.  Since aldehydes,
both aliphatic and aromatic, occur as primary and
secondary pollutants and are direct precursors of
free radicals in the atmosphere, aldehyde chemistry
represents an important subject area.  The
understanding of this topic is necessary to meet
the objective of modeling tropospheric chemical
reactions.  In this context, the major objective
of this paper is to consider the historical
interest in aldehydes; their sources and
atmospheric concentrations; the photochemistry,
kinetics and mechanism of their reactions and
finally to delineate current measurement needs
and recommend research priorities based on
assessment of the  current status of knowledge of
the chemistry of aldehydes in the troposphere.

   In addition, the role of other oxygenated
hydrocarbons in tropospheric chemistry will be
addressed briefly.  Although aldehydes are the
main oxygenated hydrocarbons generally considered,
and will receive major considerations here, other
classes of oxygenated hydrocarbons merit
consideration and  should be assessed in terms of
their involvement  in the chemistry of the
polluted troposphere.  Thus ketones, esters,
ethers and alcohols will be briefly considered
to assess their possible importance in modeling
the troposphere.  The major areas of uncertainty
will  be discussed  and research priorities
suggested.

   This paper is  an attempt to survey the current
published literature on aldehydes (and, to a
lesser extent, other oxygenated hydrocarbons) as
the work relates to modeling the troposphere.  It
is hoped that the discussion periods will  extend
the coverage to include unpublished work,  prelimi-
nary results, and peripheral studies which have
a direct bearing on the overall thrust of this
paper.

  2.  Previous Work and Importance of Aldehydes

   Initial impetus for the interest in the role
of aldehydes in photochemical air pollution
stemmed largely from the possibility that they
were connected with eye irritation which became
a major phenomenon and problem in the Los  Angeles
basin during the 1940's.  However, an early
Stanford Research Institute study (SRI 1950)
concluded that "concentrations of aldehydes have
rarely exceeded 0.2 parts per million by weight
and the high concentrations did not coincide with
periods of eye irritation.  This lack of correla-
tion tends to indicate that aldehydes alone are
not responsible for eye irritation."  Subsequent
work indicated that acrolein was present on highly
polluted days and this compound is known to be a
potent eye irritant (Los Angeles Air Pollution
Control District, 1950; Altshuller and McPherson,
1963; Scott Research Labs, 1969).  Acrolein and
formaldehyde were shown to be produced upon
irradiation of dilute automobile exhaust and
olefin-NOx mixtures (Schuck, 1957; Schuck and
Doyle, 1959).

   Aside from the possible relationship of
aldehydes to eye irritation, it was subsequently
proposed (Leighton and Perkins, 1956; Leighton,
1961) that aldehydes could act as precursors to
radicals which could either directly form oxidant
or oxidize NO to N02.  This possibility received
support from the results of several experimental
studies focused on the photooxidation of aldehydes
under laboratory and simulated atmospheric
conditions and generally employed formaldehyde
and the lower molecular weight aliphatic aldehydes
(Haagen-Smit and Fox, 1956; Altshuller and Cohen,
1963; Altshuller, Cohen et al., 1966; Johnston
                                                   27

-------
 and  Heicklen,  1964;  Altshuller, Cohen et al.,
 1967;  Cohen,  Purcell  et al.,  1967; Purcell  and
 Cohen, 1967;  Bufalini  and Brubaker, 1969).

   Recently  Dimitriades et  al., (1972) and  Pitts
 et al.,  (1976)  carried out  experiments in a  smog
 chamber  illustrating  the effect of initial
 aldehyde concentrations on  oxidant production
 under  simulated atmospheric  conditions.   Figure 1
 shows  the significant impact  of initial  aldehyde
 concentrations  on  ozone formation  in a nine-hour
 irradiation  of  a surrogate  hydrocarbon mixture
 (Pitts et al.,  1976).   Thus  an  approximately  100
 percent  increase in  initial  formaldehyde
 concentration  from 91  to 185  ppb increases  the
 maximum ozone  concentration  by  approximately
 25 percent from about 0.39  to nearly 0.5 ppm  in
 nine hours.   Clearly, the rate  of  formation  of
 03  is  enhanced  but it is possible  that the  03
 maximum value  would  not be  significantly increased
 if the irradiations  were carried out sufficiently
 long.
 Fig. 1.  Effect of added HCHO on ozone formation in
          long-term irradiations of surrogate mix-
          ture (from Pitts et al.,  1976).

   Aldehydes can provide significant sources  of
radicals such as H02, OH and R02 which can  influ-
ence the rate at which photochemical oxidants are
formed under ambient conditions.  With the  advent
of appropriate computer calculation facilities
to handle complex kinetic mechanisms, a number of
workers demonstrated this effect by carrying  out
computer simulations of atmospheric chemistry
both with and without initial  aldehydes (Niki,
Daby and Weinstock, 1972; Calvert et al.,  1972;
Demerjian, Kerr and Calvert, 1974;  Dodge and
Hecht, 1975; Levy, 1974; Whitten and Dodge, 1976;
Graedel, 1976; Carter et al.,  1978).  Many  of
these calculations have focused on  formaldehyde
which photodissociates to produce significant
amounts of H02 radicals under ambient conditions.
Thus Demerjian et al., (1974)  have  shown that
this route is the most important source of  H02
radicals in the atmosphere.

   Although there is some uncertainty attached to
the quantum yields for photodissociation into
radicals of HCHO as a function of wavelength
(vide infra), aldehydes are  well established  as
important ingredients in photochemical  smog
formation.
    The role of aldehydes as eye irritants and
 radical precursors has been given above.  An
 additional role for aldehydes is as precursors to
 the formation of peroxyacyl nitrates.  These can
 be formed by the reaction mechanism
      RCHO + OH + RCO + H20
RCO
               02
            N02 ->• RC03N02
                  peroxyacyl nitrate
    Peroxyacyl nitrate type compounds have been
 found in many parts of the world e.g., Penkett
 et al .  (1975) in England, van Ham and Nieboer
 (1972)  in Netherlands, in Japan (Akimoto and
 Kondo,  1975) and in the U.S.A. (Stephens, 1969;
 Lonneman et al .  (1976)).

      3.   Sources  and  Ambient  Concentrations

    Sources.   There are primary and secondary
 sources  of aldehydes  in the atmosphere.   The
 primary  sources  are related to combustion and
 result from  incomplete combustion  in,  for example,
 internal  combustion engines,  diesel  engines  and
 stationary sources, such  as  incinerators,  etc.
 (Altshuller  et al., 1961;  Linnell  and  Scott,  1962;
 Elliot et al., 1955).   Automobiles are a  signifi-
 cant  source  of aldehydes  and  the latter  account for
 up  to one-tenth of  the hydrocarbon emissions  (Black,
 1977).   Oberdorfer  (1964)  and Seizinger  and
 Dimitriades  (1972)  have analyzed the individual
 aldehydes emanating from  pre-controlled  automobiles.
 Table 1  shows the  percentage  of aldehydes from
 automobile exhaust  as  determined by  several
 workers  (Oberdorfer,  1964;  Fracchio  et al . ,  1967;
 Wodkowski  and Weaver,  1970; Wigg et  al . ,  1972).
 It  is evident from  these  emission  sources  that
 formaldehyde  is the largest aldehyde component.
 Similar  but more extensive  results are shown  in
 table 2 which were  obtained by  Seizinger  and
 Dimitriades  (1972).

    It can  be seen that  in addition to the  saturated
 aliphatic  aldehydes, acrolein--a potent eye
 irritant—is also present.  In  addition,  benz-
 aldehyde  and formaldehyde are produced, along  with
 alcohols,  ethers  and ketones.   One would of course
 expect variations in the relative amounts  of
 these compounds depending on the fuel used, e  q ,
 see table  1 .

   With the advent of hydrocarbon control measures
 for automobiles,  the aldehyde concentrations have
 been  reduced along with the hydrocarbons.  However,
 different control  techniques apparently have
 varying effects  upon the percentage reduction  of

T 2s yBlLk°f?Q7^ W,!th the remaini"9 hydrocarbons.
 ihus, Black (1977) shows interesting  data for
emissions from automobiles using thermal  reactors,
Jean burn technology and catalysts  of various
kinds to  reduce  hydrocarbons.   Table  3 shows a
cssreducnf               ve            ,
class reductions for various automobiles employing
different hydrocarbon control  systems.  S1™^oying

^1^7 ha atVT "'h"9 the "talyst s^em
rather than the lean burn system, effect greater
reductions of aldehydes.                 greater
                                                   28

-------
                Table 1.  Exhaust ialdehyde analyses (adapted from National Academy of Sciences, 1976).

                                        Fraction of total exhaust aldehydes, volume percent
         Aldehyde	

         Formaldehyde
         Acetaldehyde

         Propionaldehydeb
         Acrolein

         Butyraldehydes
         Crotonaldehyde

         Valeraldehydes

         Benzaldehyde

         Tolualdehydes
         Other

             TOTAL
               Nigg et al. (1972)a  Oberdorfer (1964)  Wodkowski and Heaver  (1970)a  Fracchio et al.  (1967)
                  66.7
                   9.3
                  15.7

                  I.3:2
                  3.2

                  1.9
                           100
 72.5

  8.7
  4.3
  7.0

  7.2

  0.3

100
 70.2

  7.2

  0.4

  9.8

  0.4

  0.4

  0.4

  8.5


  2.5

100
 59.9

 14.3

 {7,


  3.0

  1.4


  3.3

  5.9

  5.2

100
 69.3

  7.5

  0.7

  2.6

  1.0

  0.4


  5.4

  3.1

 10.0

100
 72.9

  8.5
  1.7

  0.4


  4.3


  5.8

100
         aExhausts from two different gasolines.
          Also includes acetone of unknown proportion.
Table 2.
Oxygenates  in  exhaust from simple hydro-
carbon fuels  (from Seizinger and Dimi-
triades,  1972.).
             Oxygenate
Acetaldehyde               .
Propional dehyde  (+ acetone)
Acrolein
Crotonal dehyde  (+ toluene)
Tiglaldehyde
Benzal dehyde
Tolual dehyde
Ethyl benzal dehyde
o-Hydroxybenzal dehyde  (+ Cm
 aromatic)''                u
Acetone (+ propional dehyde)
Methyl ethyl ketone
Methyl vinyl ketone (+ benzene)
Methylpropyl  (or isopropyl)
 ketone
3-Methyl -3-buten-2-one
4-Methyl -3-penten-2-one
Acetophenone
Methanol
Ethanol                  f
Cs alcohol (+ C3 aromatic)
2-Buten-l-ol  (+  C3HeO)
Benzyl alcohol
Phenol  + cresol(s)
2,2,4,4-Tetramethyltetrahydro-
 furan
Benzofuran
Methyl phenyl ether
Methyl  formate
Nitromethane
C5H50
C5H100
                        Concentration
                         range, ppma
                          0.8- 4.9
                          2.3-14.0
                          0.2- 5.3
                          0.1- 7.0
                         <0.1- 0.7
                         <0.1-13.5
                         <0.1- 2.6
                         <0.1- 0.2
                         <0.1- 3.5

                          2.3-14.0
                         <0.1- 1.0
                          0.1-42.6
                         <0.1- 0.8

                         <0.1- 0.8
                         <0.1- 1.5
                         <0.1- 0.4
                          0.1- 0.6
                         <0.1- 0.6
                         <0.1- 1.1
                         <0.1- 3.6
                         <0.1- 0.6
                         <0.1- 6.7
                         <0.1- 6.4

                         <0.1- 2.8
                           <0.1
                         <0.1- 0.7
                         <0.8- 5.0
                           <0.1
                         <0.1- 0.2
                         <0.1- 0.3
 Values represent  concentration levels in exhaust
ufrom all  test fuels.
 Data represent unresolved mixture of propion-
 alydehyde +  acetone.  Chromatographic peak shape
 suggests  acetone  to be the predominant component.
^Tuolene is the predominant component.
 The Cio aromatic  hydrocarbon is the predominant
 component.
 Benzene is the predominant component.
The  aromatic hydrocarbon is the predominant com-
ponent.
                       Aldehydes are  also emitted from some stationary
                    sources (EPA-AP42).   These include power plants
                    burning oil and coal  and sources  such as incinera-
                    tors,  animal rendering facilities,  or gasoline
                    and diesel engines  operated at stationary source
                    facilities.  Typical  levels from  stationary
                    combustion sources  are given in table 4 (National
                    Academy of Sciences,  1976).

                       Secondary.  Sources of aldehydes  include the
                    oxidation of hydrocarbons in the  presence of NO
                    under  ambient atmospheric conditions.   Figure 2
                    shows  a concentration time plot for  the formation
                    of  secondary pollutants including  aldehydes from
                    an  irradiated hydrocarbon-NOx mixture under
                    simulated atmospheric conditions.  Major sources
                    in  such systems are the reactions  of ozone and OH
                    radicals with olefins, and radical decomposition
                    products, e.g.,

                        03 + C3H6 -»•  HCHO + CH30 + radical products
                        RCHOH + 02 *  RCHO + H02

                    In  addition, aromatic aldehydes can  be formed by
                    the reaction of OH with aromatics, i.e., benz-
                    al dehyde.

                       Ketones and hydroxyaldehydes or hydroxy-ketones
                    can obviously also be formed in such oxidation
                    systems.

                       Ambient Concentrations.   The major problem in
                    measuring atmospheric concentrations of aldehydes
                    is  the lack of an appropriate monitoring technique
                    applicable to this low concentration regime.   For
                    example, wet chemical  techniques, such as the
                    chromatropic acid method for measuring for-
                    maldehyde, are subject to interferences and
                    uncertainties in  the  results can be  large.
                    Consequently, results  of studies in  the Los Angeles
                    area in the 1940's and 50's may be subject to
                    large  error.   For example,  levels of 1.87 ppm in
                    1956 reported for total  aldehydes by the Los
                   Angeles County Air Pollution Control  District are
                   suspiciously high.  More typical  values of around
                   0.4 ppm were reported for lower molecular weight
                   aldehydes on days of  significant air pollution
                    in  Los Angeles in the late 1940's  and early 50's
                    (Katz, 1961).
                                                   29

-------
               Table 3.  Automobile exhaust hydrocarbon (and aldehyde) emission patterns (from Black, 1977).
                 Total exhaust                     Percentage of total hydrocarbon, wt.%
         Control    Hydrocarbon,                                                                       u
                             Methane  Paraffin  Acetylene  Benzene  Aromatic  Ethylene  niefin Formaldehyde  aldehydes
1972
350 CID
1974
Mazda
rotary
40 CID
1975
Chev.
Impala,
350 CID
Honda
CVCC
(proto. )
91 CID
Volvo
(proto)
130

non- 1.15 9.29 40.68 11.04 2.43 22.30 12.97 26.00 3.28
catalyst
thermal
reactor K4? 8^9g 4K62 6_2Q 2.04 18.04 17.45 33.70 6.83



oxidation .25 15.44 57.09 2.72 1.44 19.68 8.76 20.44 1.60
catalyst

chargefled .28 4.64 37.82 7.29 1.79 22.24 15.11 32.46 6.00
(lean)
three-way
oxidation" .16 19.91 56.59 4.73 1.47 21.60 5.17 17.10 .94
catalyst
7.00


11.00



3.64


11.29


3.25

  Table 4.  Typical  emission of several classes of
           compounds  including aldehydes from sta-
           tionary combustion sources  (from Na-
           tional  Academy of Sciences, 1976).
Compounds
Hydrocarbons
Aldehydes
Formaldehyde
Organic acids
Emission,
Coal
0.3
unknown
0.003
10
Ib/ton of
Oil
1.0
0.5
0.006
%5
fuel
Gas
1.0
0.5
0.008
2
    0.54
             60
                   120   180   240
                   ELAPSED TIME (min)
                                     300
                                            360
Fig. 2.
Results of a typical smog chamber  experi-
ment (SAPRC evacuable chamber).  Irradia-
tion of a propylene-NO-N02 mixture in  air.
Initial experimental conditions--0.5 ppm
propylene, 0.45 ppm NO, and  0.05 ppm N02
in 760 Torr of highly purified  air.
   Aldehydes have been  measured in various parts
of the world, but the most extensive body of data
exists for Los Angeles  (Stanford Research Institute.
1950; Renzetti and  Bryan,  1961; Altshuller and
McPherson, 1963; Scott  Research Laboratories,
1969; California Air Resources  Board, 1972).  The
studies of Altshuller and  McPherson showed typical
formaldehyde and acrolein  levels of ^ 0.04 ppm
and ^ 0.07 ppm respectively during September to
November 1961.  More recently,  the California Air
Resources Board (1972)  measured formaldehyde at
levels up to 0.1 ppm, while daily average levels
were around .035 ppm.   Acetaldehyde exhibited an
average concentration of .02 ppm, while no other
aldehydes or ketones were  detected above their
threshold of 0.1 ppm.   The Air  Resources Board
found that aldehyde levels in the eastern part of
the Los Angeles basin were significantly lower
than those in downtown  Los Angeles: specifically,
formaldehyde was found  to  average less than 0.2
ppm and acetaldehyde less  than  .015 ppm at Azusa.
Figure 3 shows hourly concentrations measured at
two locations in the Los Angeles area in 1968
by Scott Research Laboratories.  Both locations
show sharp decreases in afternoon levels of total
aliphatic aldehydes.

   The advent of Fourier transform infrared
spectroscopy  (FTIR) has added a significant new
dimension to the measurement of trace pollutants,
including aldehydes, in ambient air.  The technique
is specific and sensitive. Hanst and coworkers
(1975) first applied the method in Pasadena in
1972.  They measured HCOOH at surprisingly high
levels but detected no  significant amounts of
HCHO.  More recently, low levels of formaldehyde
(<_ 10 ppb) have been observed in the eastern
part of the Los Angeles basin by Tuazon et al.
(1978a) also using  FT-IR spectroscopy.  These
workers detected levels of formic acid up to 10
ppb and there was no obvious correlation between
the HCHO and HCOOH  ambient concentrations.
However, this study has been largely superceded
by more recent results  using improved absorptivi-
                                                    30

-------
   a I
   t- a.
   uj o
   O

0.16

0.14
0.12
0.10
0.08
0.06
0.04
0.02

0
0.14
0.12
0.10

0.08
0.06
0.04
0.02
n
I I I I I
— ^X\ HUNTINGTON _
-,x^ \ PARK
, 	 ^r 	
/ X'^Y.
/ » /\ v -
— -///'*%» -
- / ''
> /,''' 	 ALIPHATIC ~
-,? ALDEHYDES -
L— ' 	 FORMALDEHYDE _
	 ACROLEIN
. EL MONTE
A -
/ \
— 1 \ —

- .,^*** /'/--- -
^ *Jf'
~*t
**' "
i i i i i
\J.\J IO
0.016

0.014
0.012
0.010
0.008 I
Q.
0.006 z"
O
0.004 H
0.002 =
Z
0 uj
0.014 g
0
0.012 z
0.010 J
O
0.008 g
0.006 **
0.004
0.002
n
                          12
                      LOCAL TIME
Fig. 3.
   601—
                                    HCHO
             1200
                     1400      1600
                       TIME (hoursl
Fig.  4.
         Ambient concentrations  of  HCHO  and  HCOOH
         as  a  function of  time measured  in River-
         side, October 14,  1977  (from  Tuason et  al..
         1978).

 ties for several  species including  HCHO and  HCOOH
 (Tuazon et al., 1978b).  Some of the results  from
 this 1977 study are shown in figure 4.  The
 ambient levels of HCHO for this October 14,  1977
 day (^ 36 ppb)  are significantly higher than
 measured in the previous study.  This  is partly
 due to the improved absorptivities  used in the
 1977 study.  There is some evidence that the  new
 absorptivities  would also reduce the levels  of
 HCOOH reported by Hanst el al.  (1973)  in the
 Pasadena study (Winer, 1978).

   In the last  few years  continuous measurements
 for formaldehyde  have been undertaken  in certain
 areas of New Jersey (Cleveland et al., 1977).
 This continuous monitoring showed a correlation
with vehicle traffic and a seasonal  variation
with higher levels in summer than in winter.   Peak
formaldehyde concentrations were in the range of
14 to 20 ppb at four sites monitored.  For example,
figure 5 shows  formaldehyde levels  reported  for
          Hourly aldehyde concentrations at two Los
          Angeles sites,  October 22,  1968 (from Air   Fig. 5.
          Quality Criteria Document for Hydrocarbons,
          1970).
                                                                                 12

                                                                             HOUR OF DAY
                                                              Diurnal concentrations of formaldehyde at
                                                              Newark, New Jersey for different days of
                                                              the week, from June l,to August 31 for the
                                                              years  1972, 1973, and 1974  (from Cleveland
                                                              et al., 1977).
Newark as a function of the day of the week.  In
Japan, Katou, (1972), observed high levels of the
unsaturated aldehyde, acrolein.  The average
concentration measured was 7.2 ppb, but maximum
levels as high as 273 ppb were reported.

   With the advent of FT-IR spectroscopy employed
by several groups of workers (e.g., Calvert et
al.; Hanst et al.; Niki et al. and Pitts, Winer
et al. more reliable data for aldehydes should
become available for both ambient and simulated
atmospheric conditions.  Of necessity, the
geographical area covered will be limited in the
near future by the complexity and expense of the
instrumentation.

         4.   Kinetics and Mechanism

   This section is divided into two parts -- the
first discusses the primary attack of radicals on
aldehydes and the second part discusses the fate
in the atmosphere of the radicals produced.

   The aldehydic hydrogen in aldehydes is relatively
weak (C-H bond strength is 86 kcal mol"1, Trotman-
Dickinson and Kerr, 1975).  Consequently, this
hydrogen atom will be susceptible to attack by
radical species present under atmospheric conditions.
Possible species are 0(3P), 0(1D), OH, H02, N03
and HSOj,.  Of these OH is likely to be by far the
most dominant.

A. Radical and Atom Reactions with Aldehydes

   OH Radical Reactions.  OH is generally thought
to abstract an H atom from aldehydes — chiefly
the aldehydric H atoms, i.e., the reaction
                                                          OH +  RCHO •* H20 +  RCO
                                          (1)
                                                   31

-------
Niki and coworkers  have carried out rate studies
for the largest  number of aldehydes ranging  from
the Ci-C3 aliphatic aldehydes to benzaldehyde
(Morris and Niki,  1971; Morris et al., 1971; Niki
et al., 1978).   The two former studies were  carried
out at low pressure using a discharge flow-mass
spectrometer  technique for the generation of
reactants and analysis of products respectively.
In their latest  study, Niki et al. used the
photolysis of HONO  to generate OH radicals
in the presence  of  the aldehyde and C2Hi, or  C2Di,
at close to atmospheric pressure and monitored
the decay of  reactants by Fourier transform  infra-
red spectroscopy.   In this way, rates of reaction
of aldehydes  relative to ethylene were determined.
These  values  were  placed on an absolute basis
using  the appropriate rate constant for the  OH
reaction with C2Hi,  at atmospheric pressure  (Niki
et  al.,  1978).   These values are shown in table  5
with the modification that the rate constant for
the OH + C2H1, reaction obtained by Atkinson  et
al.,  (1977) was  used to reduce the relative  values
of  Niki  et  al.  (1978).

   Also  included are the recent results for  HCHO
and CH3CHO obtained by Atkinson and Pitts (1978)
using  a technique yielding absolute rate constants.
These  latter  workers used a flash photolysis-
resonance fluorescence technqiue and carried out
the first study  of aldehydes over a range of
temperature  (299-426 K).  Arrhenius activation
energies for  the two aldehydes studied are small
with  acetaldehyde  exhibiting a negative value.

    Table 5  allows  a comparison among the results
obtained by  the  various workers employing three
different  techniques.  The agreement between the
earlier  work  of Niki and coworkers (1971) and the
most  recent study of Niki et al. (1978) is
excellent  for HCHO and CH3CHO but only fair  for
C2H5CHO.   These results agree well with those of
             Atkinson and Pitts  (1978)  for  CH3CHO but are about
             50 percent higher for  HCHO.

                If one assumes an atmospheric OH concentrations
             of 106 radicals cm"3,  the  rates of decay of HCHO
             and CH3CHO by  reaction with  OH are •*- 4.2 percent
             and 5.8 percent h J respectively.

                0 Atom Reactions.   Attention here will be
             focused on ground state atomic oxygen,  (03P)
             since this more abundant than  0(1D) in  the lower
             troposphere.   0(3P) reacts with aldehydes in the
             same manner  as OH,  by  abstracting the aldehydic
             H atom.  However, the  reaction is a chain branching
             one compared with a chain transfer reaction in the
             case of OH
                  0 +  RCHO  + RCO + OH
                                            (2)
                Results  of several  studies of 0(3P) reacting
             with a variety of aldehydes are shown in table 6.
             No results  are shown from purely high temperature
             studies  such  as shock  tubes.

                The most extensive  data are those obtained by
             Singleton et  al.  (1977)  for four aldehydes-acetalde-
             hyde, propionaldehyde,  n- and iso-butyraldehyde.
             These workers  used a phase shift technique and
             covered  a temperature  range of 298-472 K.  They re-
             ported that at the high  end of their temperature
             range, abstraction of  the alkyl group H atoms be-
             came significant  for the higher molecular weight
             aldehydes.  However, under atmospheric conditions,
             abstraction of the aldehydric H atom is likely to
             dominate.   The room temperature rate constants in
             table 6  show  that there  is generally good agreement
             among the different workers for acetaldehyde but
             significant differences  for propionaldehyde and
             butyraldehyde.  The technique used by Singleton et
             al., would  suggest that  their results should be more
             reliable and  should be  used in any modeling studies.
                Table 5.  Rate constant data and Arrhenius parameters  for the reaction of OH radicals with aldehydes.
Reactant
HCHO



1012 x A
cm'molec'H"13
—
—
12.5
"
1012 x k
E cal mol"1 cm3molec~'s~1
>6.6
14 ± 3.5
175 ± 300 9.4 + 1.0
14.4 i 0.8
at 1 K
300
298
299
298 + 2
Reference
Herron and Penzhorn, 1969
Morris and Niki, 1971
Atkinson and Pitts, 1978
Niki et al.,1978 (relative
to_OH + C2H» 8.00 x
Temperature
range
covered


299-426 K

        CH3CHO
                     6.87
15 + 3.E

  <20
                             -510 + 300   16.0 ± 1.6

                                        15.2 ± 1.6
        C2H5CHO
        C6HSCHO
                                        12.8 ± 0.8
                 (Atkinson et al.,
            1977))

  300      Morris, Stedman and Niki,  1971

295 + 2     Cox et al., 1976 (relative to
            OH + HONO = 6.6 x 10~12
            (Cox, et al., 1976))

  299      Atkinson and Pitts, 1978

298 ± 2     Niki et al., 1978 (relative to
            OH + C2Ht = 8.00 x 10~12
            (Atkinson et al., 1977))

  298      Morris and Niki, 1971

298 + 2     Niki et al., 1978 (relative to
            OH + C2Hi, = 8.00 x 10"12
            (Atkinson et al., 1977))

298 ± 2     Niki et al, 1978 (relative to
            OH + C2H,, = 8.00 x 10~12
            (Atkinson et al., 1977))
                                                     299-426 K
                                                     32

-------
              Table  6.  Arrhenius parameters and  rate constants for the reaction of oxygen atoms  (03P) with aldehydes.
                       A              E               k
      Reactant    (cm3 molec'V)  (kcal mol'1)   (cm3 molec"'s ')   TK        Technique
                                               Reference
      HCHO
      CH3CHO       1.20x10""    1.460 ± 0.153
                   2.3 x 10""         2.36
      C2H5CHO      1.30x10""     1.727 + 0.066
                  1.41 x 10""         3.8
      C3H,CHO      1.66 x 10""     1.702 + 0.040
      i-C3H7CHO    1.32x10""     1.445 ±0.091
      CH2   CHCHO  7.8 X 10"12         2.0
      (acrolein)
1 5 x 10"13     300    Discharge flow-mass
                     spectrometry  (DF-MS)

1.6 X 10"13     300    DF-MS
1.5 x  10"'3     300    Discharge flow-product
                     analysis, epr and
                     chemiluminescence

            298-472   Hg sensitization-
                     chemiluminescence

4.3 x  10"13     298

4.8 x  10"13     298    Discharge flow-product
                     analysis, epr and
                     chemiluminescence

            300-480   Discharge flow-
                     chemiluminescence

4.5 x  10"13     298

5.0 x  10"13     298    Hg sensitization-
                     product analysis

            298-472   Hg sensitization
                     chemiluminescence

7.0 x  10"13     298

            300-480   Discharge flow-
                     chemi1umi nescence

2.3 x  10"13     298

            298-472   Hg sensitization-
                     chemiluminescence

9.5 x  10"13     298

2.5 x  10"'3     298    N02 photooxidation-
                     quantum loss of N02

            298-472   Hg sensitization-
                     chemiluminescence

1.2 x  10"12     298

            300-480   Discharge flow-
                     chemiluminescence

2.7 x  10"13     298
Herron and Penzhorn
(1969)

Niki, Daby, and
Weinstock (1969)

Mack and Thrush
(1973)
Singleton et al.
(1977)
                                                                                            Mack and Thrush
                                                                                            (1974)
                                                                                            Cadle and Powers
                                                                                            (1967)
                                                                                            Cvetanovic (1956)
                                                                                            Singleton et al.
                                                                                            (1977)
Cadle  et al. (1972)




Singleton et al. (1977)




Jaffee and Wan  (1974)


Singleton et al. (1977)




Cadle  et al. (1972)
6.0 x 10""

Croton- 3.3 x 10""
dehyde

6.4 x 10""
2.84 296-423
4.9 X 10"13 296
2.3 ?
8.3 X 10"13 298
2.43 296-423
1.09 x 10~12 296
Relative rate
relative to 0 +
propylene

Discharge flow-
chemi luminescence

Relative rate
relative to 0 +
propylene
Gaffney, Atkinson,
and Pitts (1975)

Cadle et al. (1974)

Gaffney, Atkinson,
and Pitts (1975)
   The three  major studies of formaldehyde  at room
temperature show excellent agreement.   The  rate
constant appears to  be about  one third that of
acetaldehyde  which reflects the weaker aldehydric
H bond in acetaldehyde.

   If  one assumes an  0(3P) atom concentration of
105 atoms cm"3,  for  the lower troposphere,  the
rates  of reaction of HCHO and CH3CHO  with 0(3P)
are 5.4 x 10"3  and 1.5 x 10"2 percent h"1
            respectively.   Thus this route  will  be unimportant
            for the  atmospheric chemistry of aldehydes.

               Measurements have  been reported for the  un-
            saturated aldehydes acrolein, CH2 =  CHCHO and
            crotonaldehyde, CH3CH   CHCHO.   Cadle  et al.  (1972,
            1974)  used a discharge  flow technique  over  the
            approximate temperature range of 300-480 K.   The
            values shown in table 6 are in  reasonable agreement
            with those of  Gaffney et al.  (1975).   The latter
                                                          33

-------
workers used a relative rate technique  using  the
mercury photosensitized decomposition of  N20  at
2537 A to generate 0(3P) atoms.   Product  analysis
was by gas chromatography.

   H02 Radicals.  No room temperature rate  constants
have been reported for the  reaction  of  H02  radicals
with aldehydes, although Baldwin  and coworkers
(1972) have_obtained a value of 1.6  x 10~15 cm3
        "
           _
molecule"^  J at 773  K  for  formaldehyde.   In
addition, Hendry and_Mabey  (1973)  reported a  value
for HCHO of 7.1 x  10"18
                        cm"
                           molecule  V1  at  373  K.
In order to obtain  an  estimate  for  the  rate  of
reaction of
      H02 +  HCHO + H202  +  CHO
                                         (3)
 one may  employ  the  expression  suggested  by  Lloyd
 (1974) of  1.7 x 10"12  exp(-4700/T)  cm3
molecule
cm3 molecule
 to produce M
:s   at 298 K.
                               2.8  x  10
    Combining  this  value  with  a  typical  H02
 concentration of 109  radicals cm3  for  the_polluted
 troposphere one  obtains  a  value of 1 x 10  3
 percent  h"1 for  the  rate of disappearance  of HCHO
 by  reaction with H02.  This rate would be  about
 an  order of magnitude  smaller in the unpolluted
 troposphere.

    Although measurements of the rate constant for
 H02 reacting  with  HCHO would  be desirable  from a
 scientific viewpoint,  unless  current measurements
 at  higher temperatures are grossly in  error, it
 does not appear  from an  atmospheric modeling view-
 point that this  reaction plays  a significant role
 in  the chemistry of  the  polluted troposphere.

    Alkoxy Radicals.   Kelly and  Heicklen (1978)
 have recently measured the rate constant for the
 lowest molecular weight  compound in the series,
 methoxy radicals reacting  with  acetaldehyde
      CH30 + CH3CHO ->• CH3OH + CH3CO
                                         (4)
 This is a radical  transfer reaction.   The authors
 photolyzed azomethane in the presence of acetal-
 dehyde and oxygen  at 298 K and from a product
 analysis obtained  U/ks   1.4 ± 2.8 where k5
      CH30 + 02 -»• CH20 + H02
                                         (5)
Using the value of 6 x 10"16 cm^ molecule"1
s :  obtained by Barker et al.  (1977)_for ks,
 authors guoted k% = (8.3 ± 1.7)  x 10"
                                              the
 molecule
               at 298 K.
                                        cm3
    The steady state  concentration  of CH30  in  the
 lower polluted troposphere  is  around 5  x 106
 radicals cm"3.   Thus the rate  of disappearance  of
 acetaldehyde by reaction with_CH30 in thejower
 troposphere is about 1.5 x  10  2  percent h  i.
    NO 3 Radicals.
 the reaction
                  Morris and Niki  (1974)  studied
      N03 + CH3CHO •+ HN03 + CH3CO
                                         (6)
 using a long path length IR cell  operated at 300 K
 and near atmospheric pressure.   The results from
 studying the reaction of mixtures of N205 and
                                                      CH3CHO were interpreted using numerical  integration
                                                      for the  participating reactions.  A value  of

                                                        k6 =  1.2 x  10"15 cm3 molecule"1 s"1  at  300 K

                                                      was obtained by  varying k6  until  a good match of
                                                      the calculated and observed N205  decay  was
                                                      obtained.

                                                        _Assuming an N03 concentration  of  109 radicals
                                                      cm 3  for the polluted troposphere, one  can estimate
                                                      that  the rate  of disappearance  of CH3CHO by
                                                      reaction 6  is  0.4 percent h l.

                                                        HSOi,  Radicals.  This  radical  is given some
                                                      consideration  here although no  experimental  data
                                                      are available  for the  relevant  reaction:
                                                          HSO,, + RCHO * H2SOi, + RCO
                                                                                               (7)
                                                     Benson (1978) has suggested that the HSOi, radical
                                                     could be more reactive than RO radicals  in either
                                                     adding to the double bond of olefins or  in H-
                                                     abstraction reactions.  Since aldehydes  have a
                                                     relatively weak C-H bond, reaction 7 is  a logical
                                                     candidate to consider in HC-NO -SO  photooxidation
                                                     systems .

                                                        HSOi, can be formed in the sequence of reactions

                                                               OH + S02 + M + HS03 + M
                                                                   HS02
                                                                   HS05 + NO
                                                                              HS05
                                                                                   +  N02
                                                         Using AHf(HSOO    -125 kcal mol"1 (Benson,
                                                      1978)  one can estimate that reaction 7 for HCHO
                                                      is 17.1  kcal  mol  *  exothermic.  (The similar
                                                      reaction for OH radicals reacting instead of HSOi,
                                                      is 32.30 kcaljnol"1 exothermic and for CH30 is
                                                      17.5 kcal mol * exothermic).   Thus, by comparison
                                                      with the sole experimental _measurement of k.,, one
                                                      would  expect k7 >_ 8.3 x 10 15 cm3 molecule"1 s"1
                                                      at 300 K.  However, this reaction should be studied
                                                      experimentally since thermochemistry is not always
                                                      a reliable guide for estimating kinetic data.

                                                      B.  Atmospheric  Reactions of Radicals Produced
                                                         from  Attack  of  Radicals  on Aldehydes

                                                         We  have  seen above that  radicals  of the form
                                                      RCO are  produced from the reaction of atoms  and
                                                      free radicals with  aldehydes.   In  this  section,
                                                      the subsequent  fate of these  radicals under  atmo-
                                                      spheric  conditions  will  be  discussed.   Differentia-
                                                      tion is  made  between  the acyl  radicals  and their
                                                      aromatic equivalents  since  there  is  evidence
                                                      (Niki  et al., 1978)  that the  radical  produced from
                                                      benzaldehyde  react  differently from their aliphatic
                                                      equivalents.

                                                         The simplest acyl  radical  is formyl  produced
                                                      from formaldehyde.   Three reaction paths  are
                                                      possible for  its reaction under atmospheric
                                                      conditions.   These  are:
                                                          HCO + 02 -* H02 +  CO

                                                          HCO + 02 -+• OH + C02

                                                      HCO + 02 + M -*- HC03 + M
                                                                                              (8)

                                                                                              (9)

                                                                                             (10)
                                                    34

-------
   The following results support the conclusion
that reaction (8) is the main route for this
reaction:

     the reaction obeys second order kinetics at
     low pressures (Washida et al., 1974),

     the rate constant for HCO + 02 is independent
     of pressure over the range 20-500 Torr
     (Shibuya et al., 1977),

     the direct identification of H02 formation
     from HCO by laser magnetic resonance (Radford
     et al., 1974),

     the observation of H02 formation from HCO +
     On at 1 atmosphere (Hunziker,  1975).

   However, results  from recent studies by Osif
and Heicklen (1976)  and Niki et al., (1977)
suggested that reaction (10) was the dominant
pathway.  Both of these studies used the Cl atom
sensitized decomposition of formaldehyde in the
presence of 02.  They assumed that formic acid
formation was a good indicator of reaction (10),
since HC03 radicals  would be converted to HC02
radicals and subsequently HCOOH by H-abstraction
from the aldehyde, thus

     HC03 + HC03 + 2HC02 + 02           (11)

     HC02 + HCHO -»• HC02H + HCO.         (12)

Osif and Heicklen measured k10/ks ^5 ± 1  indepen-
dent of pressure over the range studied (70-700
Torr) while Niki1 et al., obtained a value > 9 for
the same ratio at atmospheric pressure.  Hanst
and Gay (1977) also  used the Cl atom catalyzed
oxidation of HCHO.  They irradiated low concentra-
tions of C12/HCHO/N02 mixtures in 1  atmosphere of
air and the analysis performed using FT-IR spectro-
scopy.  From the small yield of HCOOH and the
observations of peroxynitric acid H02N02, they
concluded that reaction (10) was unimportant in
their system.

   Horowitz et al. (1978) have subsequently
suggested that formic acid can be formed by routes
other than those involving reaction (10), speci-
fically

     OH + HCHO + (HOCH20) + HC02H + H   (13)

    H02 + HCHO •* (H02CH20) •*• HC02H + OH (14)


they state that reactions (13) and (14) are 22 and
60 kcal mol"1 exothermic respectively.  These
workers photolyzed mixtures of HCHO at 3130 A at
low pressures in the presence of 02 and added C02,
and measured the quantum yields of formation of
H2, CO and C02 and the loss of 02.   A lower limit
for kio of 1.21 x 10"28 cm6 molecule"2 s"1 was
obtained from the measured ratio ki0M/(ki0M + k8)
>. 0.049 ± 0.017 (obtained from experiments using
 Niki  (1978) suggests that  formation  of  HCOOH  in
 his system may be explained  by  Reaction (14)  as
 suggested by Horowitz et al.  (1978).  This would
 reduce the importance of Reaction  (10)  in the
 system used by Niki et  al.  (1977).
PHCHO   8 Torr and Pn    1-4 Torr) and Washida
et al 's (1974) valu£2of ke.  Contrary to Osid
and Heicklen (1976) and Niki et al. (1977),
Horowitz et al. conclude that reactions  (8) and
(10) assume about equal importance under atmo-
spheric conditions.  Clearly, further work is
needed to clarify this discrepancy.

  Table  7.  Rate constants for HCO +  02 •* H02  + CO
           at 298 K.
      Rate constant,  k8    Pressure
   (cm3molec"1s-1)  x  1012    Torr
                      Reference
5.7

6.0

5.3

3.8

± 1.2

± 0.9

± 0.7

± 0.6

4

20

530

7

Washida, et al.
(1974)
Shibuya, et al.
(1977)
Shibuya, et al.
(1977)
More, quoted in
Shibuya (1977)
   Table  7 shows the generally  good  agreement  for
results of rate constant determinations  of  ks.
Washida et al. (1974)  used  a  photoionization mass
spectrometer  coupled to a flow  system  to obtain
k8 =  (5.7 ± 1.2) x  10~12 cm3  molecule"1  s"1 at
room  temperature.   Recently,  Shibuya et  al  (1977)
generated HCO radicals in the absence  and
presence  of 02 by the  flash photolysis of CH3CHO
     CH3CHO + hv(> 2000 A) -> CH3 + HCO.

From an analysis of the behavior of HCO  radical
decay, a  value of ke =  (5.6 ± 0.9) x  10"12  cm3
molecule
at 298 K was obtained.  This value
 is essentially  independent of pressure  (see
 table  7) and  is  in excellent agreement  with  that
 of Washida et al.  (1974) but larger than the value
 obtained by Moore  quoted in Shibuya et  al.

   Under atmospheric  conditions, acyl radicals
 other  than formyl  are generally assumed to react
 with 02 by addition:
     RCO + 02
 RCO-0,
Subsequent reactions  in the polluted atmosphere
with NO and N02 occur

  RCO-02 + NO -> N02 + RC02 •+  R + C02     (15)

 RCO-02 + N02 •+ RCO'02  • N02

              peroxyacyl nitrate         '   '

   The most commonly  studied  member of  the  peroxy-
acyl nitrate family is  peroxyacetyl nitrate (PAN)
(Stephens, 1969;  Pate,  Atkinson and Pitts,  1976;
Hendry and Kenley, 1977; Cox  and Roffey,  1977).
The kinetics and  mechanism of PAN  formation and
thermal decomposition have been discussed recently
(Pate et al., 1976; Hendry and Kenley,  1977;
Baldwin et al., 1977; Carter  et al. 1978) and is
beyond the scope  of this paper.

   Niki et al.  (1978) note that, based  on OH
reactivity, aliphatic and aromatic aldehydes,
                                                    35

-------
should be equally efficient at converting NO to
N02.  However, they note that smog chamber studies
show that benzaldehyde is far less reactive than
the aliphatic aldehydes in producing ozone
(Dimitriades, 1974).  They suggest that radicals
formed by H abstraction from benzaldehyde are
efficient NO  scavengers.  Support for these
suggestions would significantly aid the understand-
ing and computer modeling of aromatic hydrocarbon
photooxidation.

         5.   Photochemistry of Aldehydes

   The photodissociation of aldehydes is an import-
ant radical generation mechanism in the formation
of photochemical air pollution (Leighton, 1961;
Altshuller and Bufalini, 1965, 1971; Pitts, 1969;
Calvert et al., 1972; Demerjian et al., 1974;
Levy, 1974; Hecht, Seinfeld and Dodge, 1974;
Dodge and Hecht, 1975; Whitten and Dodge, 1976).
    20


    16
     2000 2200 2400 2600 2800 3000 3200 3400 3600 3800
                    WAVELENGTH, A

Fig. 6.  Absorption spectra of formaldehyde (1),
         acetaldehyde (2), and propionaldehyde (3)
         (from Calvert and Pitts, 1966).

 Figure 6 shows  the  absorption spectra  for  some
 common aldehydes  which  illustrates  that they will
 absorb well  beyond  3000 A.  The  two reactions of
 most significance can be generalized in terms of
 a radical  and molecular route:
         RCHO + hv
               R + HCO

               RH + CO
The  radical route is the more important one for
modeling tropospheric chemistry.  The rate
constant for a particular primary process is an
important quantity in assessing the importance of
the  process in the atmosphere.  It is given by
k,(0,h)  -
 1
                 J(X,0,h)
                                     .
where 0 is the solar zenith angle, h is the height
above ground, X is the wavelength, J is the actinic
flux, a is the cross section and $ is the quantum
yield for species i.  Of these parameters, the
quantum yield $ as a function of wavelength has
been subject to major uncertainty; for example,
singificant differences among experimentally
determined values have been reported for formalde-
hyde (McQuigg and Calvert, 1969; Sperling and
Toby, 1973).

   Quantum Yields for HCHO photolysis as a
Function of wavelength.  Considerable attention
 has been given to formaldehyde photolysis  in
 recent years, partly because of its importance in
 photochemical air pollution.  There appears to be
 general agreement that:

    a.  the major final products are H2 and CO (in
        the absence of 02),

    b.  the primary reaction paths are

      HCHO + hv -* H + HCO                 (17)

                + H2 + CO                 (18)


    However, differences have arisen in the determi-
 nations of the variation of the ratio of the
 quantum yields of reactions (17) and (18), $i7/$1B,
 as a function of wavelength.  The earlier studies
 (pre 1975) showed no consistent trend with wave-
 length variation (Gorin, 1939; Klein and Schoen,
 1956;  DeGraff and Calvert, 1967; McQuigg and
 Calvert,  1969; Sperling and Toby, 1973).   The
 results of these studies have been superceded  by
 results from more recent and definitive studies
 and can be discounted (Horowitz and Calvert,  1978;
 Lewis  and Lee, 1978).   Consequently, only the
 recent studies (post 1975) will be discussed here
 and emphasis  will  be on  the radical  production
 route,  reaction  (17).

    As  pointed out by Horowitz and Calvert (1978),
 all  the recent studies  show that the quantum
 yield  for the radical  process,  $17,  increases more
 steeply with  decreasing  wavelength below  about
 3400 A.   This  is  illustrated in figure  7.   This
 figure  is  similar  to  that  given by Horowitz and
 Calvert (1978b)  but  it  has the  additional  results
 of Moortgat et al.  (1978).   These  results  give
 slightly  higher  values of  $i7  and  show  §  lesser
 tendency  to level  off below about  3200  A.

    The  results for $1? reported by Lewis  et al.
 (1976)  should  be corrected by  a factor  of  1.89
 according to  Lewis and Lee  (1978).   Lewis  et al.
 irradiated a mixture of HCHO  and NO  with mono-
 chromatic radiation  (= 1 A  bandwidth) from a
 tunable laser  and measured  the  intensity of
 chemiluminescence produced  from excited HNO*
 formed  in the  three-body recombination  H + NO + M.
 H  atoms were produced in reaction  (17).   Relative
 values of $17  were obtained and were converted to
 absolute values using $J7   0.36 ± 0.04 at 3035
 A  determined in a separate experiment (Lewis and
 Lee, 1976).  It is this value which  has been
 redetermined by Lewis and Lee (1978) and increased
 to 0.68 ± 0.05.  This change brings  the earlier
 results of Lewis et al. (1976)  into  line with the
 most recent determinations discussed below.

   Horowtiz and Calvert (1978a and b) have  recently
 determined *17 at 3130 A and over the wavelength
 range 2890 to  3380 & at 398 K.  They obtained »16
 values by measuring the quantum yields for  H,
 production, $H2, in the photolysis of HCHO-
 isobutene mixtures assuming that the high  concen-
 trations of isobutene scavenged the H atom
production by reaction (17).  Values of «17 and

Sure^Hn  'TIT'1 fT $H2  in the Photolysis of
pure HCHO.  These authors found $17 and $,D to be
e^ntia ly l  until the longest wavelength8

0 a!163370 f? *)-n T^/I^T" t0 1nc^ase f™
0 at 3370 A to = 0.7 at 3175 A in general  agreement
                                                   36

-------
with other recent studies.  Results of experiments
at 3380 A with added oxygen lead the authors to
conclude that little if any dissociation of HCHO
into radicals occurred at 3380 A and longer wave-
lengths, in contrast to earlier results.

   Marling (1977) photolyzed 4 Torr of HDCO using
either a high pressure mercury arc coupled with a
monochromater of monochromatic laser jrradiation
in the wavelength range 3040 to 3530 A.  He
measured the relative yields of H2, D2 and HD as
a function of wavelength using a mass spectrometer
and noted that the (H2 + D2) yield is a measure of
radical production via reaction (17), since H2 and
02 can only be formed following reaction (17).
Marling found that radical  production reached
55 percent at wavelengths less than a 3200 A, but
no measurements of the absolute decomposition
yield were reported.   Horowitz and Calvert (1978)
have placed Marling's results on an absolute basis
using their value of $17 = 0.76 at 3030 K and
applied the ratio of 0.55 measured at 3040 A by
Marling.  They find that, within the uncertainties
introduced by isotopic differences in the HCHO and
HDCO molecules, Marling's reduced results agree
reasonably well with their own (see fig. 7 and
table 8).

   Clark (1976) photolyzed HCHO in the presence of
NO using a tunable laser.  The NO was employed as
a radical trap for HCO radicals and H atoms and
was used in both small  and large quantities and
the effect on $y2 and $QQ measured.  He observed
the same sharp increase in $17 in the 3200 to
3300 A range as other recent studies, and concluded
that $17 + Oje   1.0.  Different results were
obtained in experiments with low NO and high NO.
Clark assigned more credibility to results obtained
with low NO and suggested that high NO enhanced
HCHO decomposition by reaction (17).  However,
Horowitz and Calvert (1978)  have questioned this
interpretation as a consequence of the results of
their studies with added NO which showed little
effect on $H2.  They conclude that Clark's results
for $H2 at high NO are more appropriate estimates
of $i8, and find that when these results for $ia
are used, Clark's results for $17 are in excellent
agreement with their own studies.   In view of the
results of other recent studies, the suggested
reworking of Clark's results appears valid.

   Moortgat et al. (1978a) have studied the photo-
lysis of HCHO at several wavelengths in the range
2700 to 3600 A.  They employed two types of
experiment .-- HCHO in a mixture of N2 with a small
amount pf C3H6 added, and HCHO in a 20:80 mixture
of 02 and N2 — and both systems were operated at
atmospheric pressure.  Consequently, results from
the latter should be directly applicable to
modeling the lower troposphere.  A Xenon arc mono-
chromater was used to isolate the desired wave-
lengths.  Measurements of the ratio of H2 and CO
production yielded values for the ratio
*ie/$i7 + $ie-  Subsequent work (Moortgat et al.,
1978b) reported values for $17 + $is> and
consequently, $i7.  Although initially the values
for $17 obtained from the C3H5 added experiment
were lower than those for the N2:02 system,
subsequent allowance for a small contribution to
the H2 formation by H abstraction from C3H6
(Moortgat, 1978) gave excellent agreement between
the two sets of results as shown in table 8.
   Examination of figure  7  shows  that  results  of
recent studies show a consistent  trend although
there remains some scatter  in the  results  of
different studies.  However, it is  possible  to
draw a line or band which incorporates most  of the
results within their experimental  error.   This
significant improvement in  our knowledge of  $17
is very helpful in improving our capability  to
model atmospheric chemistry in the  polluted
troposphere.
0.2
0.1
0
27
Fie
1
c
£
•
c
A




LEWIS ETAL 1978
HOROWITZ AND CALVERT 1978
MARLING 1977
CLARK 1976
MOORTGAT ET AL 1978



•
[
• D
.DC x
•
1 1 -4
00 2800 2900 3000 3100 3200 3300 3400
WAVELENGTH, (A*)
. 7. Primary Quantum Yield, $17, for process
          HCHO + hv -* H + HCO,  as a function of
          wavelength.   Data  of  Marling and Clark
          plotted using interpretation of Horowitz
          and Calvert.

    In  order to compare the  rate of photolysis
 with the depletion of formaldehyde by radical
 processes one can calculate a  photolysis rate  of
 ^ 13 percent h"1 for a solar zenith angle of 20°
 and using the value of kJ7  given by Horowitz and
 Calvert  (1978b).

    Photolysis of Acetaldehyde.   Acetaldehyde is
 commonly used as a surrogate for aldehydes  of
 higher molecular weight  than formaldehyde.   Its
 absorption  spectrum is shown"in figure  6, which is
 taken  from  Calvert and Pitts (1966).   As with
 formaldehyde,  major uncertainty is  concentrated
 in  the quantum yields  of the primary  processes
 (Calvert  and  Pitts,  1966).

    Parmenter  and  Noyes (1963) carried out emission
 studies  and Archer et  al. (1973)  used the triplet
 state  induced  cis-trans  isomerization of butene-2
 to  study  the  primary processes  in  acetaldehyde
 photolysis.  These  studies  have been  summarized
 by  Weaver et al.  (1976,  77).  In  a  comprehensive
 study, these  latter workers  obtained  results which
 are consistent with the  previous  studies.   Weaver
et  al photolysed  CH3CHO  vapor at  3130 A  in  the
presence of 02 or  02   N2 mixtures  at 298 K.  The
products formed at a function of  pressure and
added 02 were measured over  the  pressure  range  20
to 640 Torr.  Weaver et  al.   postulated the  following
                                                   37

-------
                Wavelength
Table 8.  Wavelength dependence for the quantum yield t,
        of HCHO photodecomposition into H and HCO.

                 Quantum Yield $'

2767
2754
2840
2841
2882
2890
2894
2930
2934
2950
2982
2991

3030
3035
3039
3040
3050
3088
3130
3140
3163
3166
3172

3175
3180
3195
3210
3230
3250
3260
3264

3267
3270
3296
3298

3300
3310
3324
3335
3340
3360
3378
3380
3392

3402
3550
^Original
"Results
Lewis Lewis
et al.a and Lee

0.68 + 0.10
0.51

0.64 0.64 ± 0.10



0.49

0.62



0.68 0.68 ± 0.05


0.49
0.60

0.70

0.62





0.51


0.42




0.28










< .19



results multiplied by
quoted by Horowitz and
Horowitz Marling6 Clark0 Moortgat
and Calvert
0.48


0.65

0.701
0.711
0.740

0.80

0.70
(0.48)
0.760

0.84
0.760


0.735
0.760
0.80

0.64
(0.42)
0.692
0.636
0.635
0.554
0.519
0.460 0.456
0.442
0.48
(0.42)
0.48
0.438

0.38
(0.31)
0.330
0.097
0.097
0.212
0.113
0.048
0.111
0.00
0.01
(-0.03)
0.13
0.04
1.89 as recommended by Lewis and Lee (1978).
Calvert from Marl ings experimental data.
et al.
0.44


0.66





0.81





0.81





0.80












0.49














0.09
0.03


                 Values quoted are those from high NO data reinterpreted by Horowitz and Calvert.  Numbers in
                 parentheses are those given by Clark for low NO pressures.
reactions to account for their results

     CH3CHO +  hv  -* 1[CH3CHO]n            (19)

     (quenchable  part of excited singlet
      state)

                  + :[CH3CHO]'            (20)

     (non-quenchable part of excited
      singlet  state)

      '[CHjCHO]1  -* CH3 + HCO             (21)


                  -»- CH3CHO                (22)

   1[CH3CHO]n  + M -> 1[CH3CHO]0  +  M        (23)

     (vibrationally equilibrated electronically-
      excited  single state)
                               1[CH3CHO]0


                           3[CH3CHO] + 02

                              [complex]
3[CH3CHO]


[complex]

CH3 + CO +  H02
(24)


(25)

(26)
                            Weaver et al. found no ChU produced  at 3130 A
                         in  the presence of  NO  and so concluded  that the
                         other primary process, reaction  (27), was
                         negligible at 3130  ft.
                              CH3CHO + hv +  Chk  + CO
                      (27)
                         The quantum yields  obtained are shown  in table 9,
                         which is taken from Weaver et al.  (1976, 77).

                            Reactions  (24) through (26) which predict a
                         pressure dependent  reaction between the excited
                                                      38

-------
   Table 9.  Quantum yields of the  primary processes
            in acetaldehyde photo-oxidation as  a
            function  of excitation wavelength
            (Weaver et al., 1976).

   X(R)     0{CO + CM*}    0{CH3 + HCO}   ^{triplet}
3340
3130
2967
2804
2654
2537
^Cv^m Pa'
0
0
	 	
0.1 5a
0.28a
0.64a
1 wovt anH Pitt';
0
0.05b
	
<0.30C
0.36a
0.36a
Mqfifil.
1.0d
0.84e
0.59d
0.48d
—
—

     ims wor* and Archer,  et al.   (1973).
   cCalculated from the total quantum yield of the
     free radical process (0 {CH3  + HCO}   0.39,
     Calvert and Pitts (1966); the fraction of that
     from the triplet, 0.18; and ^{triplet} at 2804
    1   0.48.
     From Parmenter and Noyes  (1963).
   eFrom Parmenter and Noyes  (1963) and  Archer,
     etal.   (1973).

 triplet state and 02,  form the novel  part of
 Weaver et al.'s results.  Although they were able
 to fit their results with their suggested
 mechanism,  the quantum yields  obtained  are  subject
 to uncertainty since  the mechanism for  secondary
 reactions may affect  the results.   For  example,  it
 was  assumed that  HCO  reacted with 02  by  addition
 (contrary to the  conclusion of our discussions
 above) and  the reaction  of  CH302  with H02  had  to
 be omitted  in order to fit  the data.

   Weaver et al calculate the  rate of CH3CHO
 photolysis  in the atmosphere for  a solar zenith
 angle of zero (i.e. the  maximum photolysis  rate)
 and  find the rates for reactions  (21) and  (26) to
 be 2.8 x 10~6 and 8.7  x  10"6 s"1  respectively.
 They report the overall  rate constant for all  free
 radical processes to  be  2.3 x  10"5 s  = 8.3  percent
 h"1.  This  may be compared with the role of deple-
 tion of CH3CHO by reaction with OH radicals
 (assumed [OH]   105 radicals cm"3) of 5.8 percent
 h"1.

                  6.   Other Oxygenates

   In addition to aldehydes, there are several
 other classes of oxygenated hydrocarbons known or
 suspected to be present  in  urban  areas of the
 troposphere.  These include ketones,  alcohols,
 esters and  ethers.  Their possible role in
 atmospheric chemistry will be  discussed briefly.

   Discussion of radical reactions with these
 oxygenates-will  be limited to  the  OH  radical.
 This was shown to be the major species attacking
 aldehydes and is likely  to be  the  most important
 intermediate for the remaining oxygenates.  Of
 the four classes of oxygenated hydrocarbons
 mentioned above, only  ketones  undergo photolysis
 under ambient conditions and this  process is
 discussed later.

   Sources  and Ambient Concentrations of Ketones,
Alcohols, Esters and Ethers.  All  of  these classes
of compounds are used  in commercial solvents
 (Burnelle et  al.  1966;  Wilson  and Doyle, 1970;
 Levy and Miller  1970;  Laity et al.  1973).   These
 compounds occur  in  paints,  degreasing solvents,
 etc.  Additionally,  automobile exhaust contains
 small quantities  of all  the above mentioned classes
 as illustrated by table  2.   Ketones and alcohols
 are also formed  via  secondary  reactions in the
 atmosphere  in a  manner  similar to that for the
 aldehydes.

   Data on  ambient  concentrations of these
 compounds are scarce and the only available quanti-
 tative measurements  appear  to  be  for selected
 ketones.  Acetone measurements of 0.3 to 0.9 ppb
 have been reported  by  Robinson et al.  (.1973) for
 remote areas of  California, Idaho,  Vermont and
 Washington.  Methyl  ethyle  ketone (MEK) has been
 observed in Riverside,  California at  concentrations
 of 1 to 6 ppb (Stephens  and Burleson,  1976).
 Smoyer et al. (1971) report the detection  of
 94 ppm of MEK in  ambient air near a chemical  re-
 clamation plant  in Maryland.   Grob  and Grob (1971)
 detected acetophenone in ambient  air  in Zurich,
 Switzerland.

   Reactions  of  Oxygenated  Hydrocarbons with the
 Hydroxyl Radicals.   The results of studies of the
 reactions of  OH  radicals with  individuals  oxygenat-
 ed hydrocarbons  are  summarized in table 10, which
 is a modified version of that  given by Atkinson
 et al. (1978).   Most of these  determinations are
 for one temperature  (around 300 K)  and were obtained
 using a relative  rate technique.   Absolute values
 were obtained as  indicated  in  table 10.

   The rates of  disappearance  of  these oxygenates
 due to reaction with OH  radicals  are  given in
 table 11, assuming an [OH]  of  106  radicals cm"3.
 The rates given  for  the  ketones will  underestimate
 their disappearance  rates in the  atmosphere since
 they can photodissociate under solar  radiation
 indicdent at the  earth's surface.   Of  the  remaining
 classes, the ethers  appear  the most reactive and
 the acetates the  least  reactive.

   Photodissociation of  Ketones Under  Ambient
 Conditions.  Ketones play a similar role to alde-
 hydes in that they photolyse to produce radicals
 which promote the oxidation of NO to  N02 with the
 concomitant formation of photochemical  smog.
 Ketones were suggested to be precursors  of
 peroxyalkyl radicals by  Purcell and Cohen  (1967)
who examined the  photoxidation  of 2-methyl-l-
 butane in the presence of acetone.  They found
 the rate of oxidation of the olefin  increased as
 the ratio of ketone  to olefin  increased.   The
 reactivity of ketones and other oxygenates under
 simulated atmospheric conditions  has been  studied
 by several  workers (Burnelle et al. 1966;  Wilson
 and Doyle,  1970;  Levy and Miller, 1970;  Laity
et al.  1973).

   Ketone photolysis has  been  described by Calvert
and Pitts (1966)  and the photolysis and photo-
oxidation of ketones have been  summarized  recently
by Lande et al.  (1976).  The absorption  spectra
of some common ketones are  shown  in  figure 8
 (Calvert and Pitts 1966).   Due  to uncertainty in
the behavior of ketones  under  atmospheric
conditions, radical  production  is often  assumed
to be 100 percent efficient.   For example  for MEK,
the reactions are
                                                   39

-------
Table 10.  Rate constant data for the reaction of OH radicals with other oxygen-containing organics.
Reactant
Ketones
Methyl ethyl ketone
Methyl isobutyl ketone
Diisobutyl ketone
Ketene
Ethers
CH3OCH3
Diethyl ether
Di-n-propyl ether
Tetrahydrofuran
CH2 CHOCH3
Alcohols
CH3OHb

C2H5OH

n-Propanol

Isopropanol

n-Butanol
CH2 CH CH2OH
Acetates
Methyl acetate
Ethyl acetate
n-Propyl acetate
sec-Butyl acetate
Methyl propionate
Ethyl propionate
1012 x_k _
cm3molec"1s"1
3.4 + 1.0
14 i 4
24 ± 7
>1.7

3.5 ± 0.35
8.9 + 1.8
16.3 ± 3.3
13.9 ± 2.8
33.5 + 3.4
1.06 ± 0.11
1.06 ± 0.11
3.3 + 0.4
3.74 ± 0.37
4.3 ± 0.4
5.33 ± 0.53
6.7 + 1.3
5.48 ± 0.55
7.6 + 1.1
25.9 + 3.3
0.18 ± 0.05
1 .94 + 0.22
4.1 ± 0.8
5.3 + 1.1
0.29 ± 0.10
1.77 ± 0.25
at T K
305 + 2
305 + 2
305 ± 2
% 295

299
305 x 2
305 ± 2
305 ± 2
299
292
296
292
296
292
296
305 + 2
296
292
440
292
292
305 ± 2
305 ± 2
292
292
Technique
Relative
rate
Relative
rate
Relative
rate
Relative
rate

FP-RF
Relative
rate
Relative
rate
Relative
rate
FP-RF
Relative
rate
FP-RA
Relative
rate
FP-RA
Relative
rate
FP-RA
Relative
rate
FP-RA
Relative
rate
Pulse
radiolysis
Relative
rate
Relative
rate
Relative
rate
Relative
rate
Relative
rate
Relative
rate
Reference
Winer et al . (1976) (relative to
OH + isobutene = 4.80 x 10"11)a
Winer et al. (1976) (relative to
OH + isobutene = 4.80 x 10"11)3
Winer et al. (1976) (relative to
OH + Isobutene = 4.80 x 10'11)3
Faubel, Wagner, and Hack (1977)
(relative to OH + C302 = 1.4 x 10"12
Faubel et al . (1977)

Perry, Atkinson, and Pitts (1977)
Covered T range 299-424 K obtained
A = 1.29 x 10 n cm3molec~ V and
E = 770 ± 300 cal mol"1
Lloyd et al . (1976) (relative to
OH + Isobutene = 4.80 x 10~")a
Lloyd et al . (1976) (relative to
OH + Isobutene = 4.80 x 10'11)3
Winer et al . (1977) (relative to
OH + isobutene = 4.80 x 10"11)3
Perry, Atkinson, and Pitts (1977)
Covered T range 299-427 K. Obtained
A = 6.10 x 10 12 cm'molec'V and
E - -1015 ± 300 cal mol'1
Campbell, McLaughlin, and Handy (1976)
(relative to OH + n-butane =
2.60 x 10~12)c
Overend and Paraskevpopulos (1978)
Campbell, McLaughlin, and Handy (1976)
(relative to OH + n-butane
2.60 x 10~12)c
Overend and Paraskevpopulos (1978)
Campbell, McLaughlin, and Handy (1976)
(relative to OH + n-butane =
2.60 x 10~12)c
Overend and Paraskevpopulos (1978)
Lloyd et al . (1976) (relative to
OH + isobutene = 4.80 x 10 ]1}a
Overend and Paraskevopoulos (1978)
Campbell, McLaughlin, and Handy (1976)
(relative to OH + n-butane =
2.60 x 10~'2)c
Gordon and Mulac (1975)
Campbell and Parkinson (1977) (relative
to OH + n-butane = 2.60 x 10~12)c
Campbell and Parkinson (1977) (re-
lative to OH + n-butane = 2.60 x 10~12)
Winer et al . (1977) (relative to
OH + isobutene 4.80 x 10 ")a
Winer et al . (1977) (relative to
OH + isobutene = 4.80 x 10"")a
Campbell and Parkinson (1977) (relative
to OH + n-butane = 2.60 x 10~'i)c
Campbell and Parkinson (1977) (relative
to OH + n-butane * 2.60 x 10~12)c
Calculated from the Arrhenius expression of Atkinson  and  Pitts  (1975).
bOsif, Simonaitis, and Heicklen (1975 ) determined rate constants relative to that for OH + CO of
  k(OH + CH3OH)/k(OH + CO) = 0.63 ± 0.10 at 298 K and 0.98  ± 0.20 at 345  K.  However, total  pressures
(CH3OH + N20 + CO) of 28-203 Torr were used.  Since no data are available for the pressure dependence of
  the OH + CO rate constant with CH3OH or N20 as the diluent gas, no guantitative estimate of k(OH + CH3OH)
  can be made, apart from setting k(OH + CO) >_1.5 x 10"13cm3molec 's '  at """   '"         "
  hence  k(OH + CH3OH) >.(0.95 ± 0.15) x ID"12 cm'molec
at 298 K.
            at 298 K (Perry et al. (1978)) and
GCalculated from the Arrhenius expression of Perry et al.  (1976) for T   292 K, which is also in excellent
 agreement with the value obtained by Campbell, Handy, and Kirby (1975).
                                                     40

-------
   Table 11.  Rates of reaction of selected ox-
             genated hydrocarbons with the OH rad-
             ical (assumed [OH] = 106 radicals
             cm 3) at around 300 K.
Compound

Ketones

  MEK

  Methyl isobutyl  ketone

  Diisobutyl ketone

Ethers

  Dimethyl ether

  Diethyl ether

  Di-n-propyl  ether

  Tetrahydrofuran

  CH2   CHOCH3

Alcohols

  CH3OH

  C2H5OH

  n-C3H7OH

  i-C3H7OH
                          Reaction  Rate (% h"1)
                                  1.2

                                  5.0

                                  8.6
                                 1.3

                                 3.2

                                 5.9

                                 5.0

                                12.1
                                 0.38

                                 1.3

                                 1.7

                                 2.2

                                 2.7
CH2CHCH2OH
Acetates
Methyl acetate
Ethyl acetate
n-Propyl acetate
s-Butyl acetate
Methyl propionate
Ethyl propionate
9.3

0.06
0.70
1.5
1.9
0.10
0.64
Fig.  8.   UV  absorption spectra for acetone (1),
         diethyl  ketone (2),  MEK (3),  and methyl
         n-butyl  ketone (4)  (from Calvert and
         Pitts,  1966).


     CH3COC2H5 + hv + CH3CO  + C2H5      (27)

                    ->• CH3  +  COC2H5      (28)
    Carter et al.  (1978) suggest  that  triplet
 formation is the  dominant  initial  process  under
 ambient conditions and this species reacts with
 atmospheric 02  to give ethyl radicals  and  acetyl
 peroxy radicals,  i.e., reaction  (27).   They state  >
 that the alternative route, reaction  (28)  is less
 favored thermodynamically.

    However, major uncertainty remains  concerning
 the photolysis of ketones  under  ambient conditions
 including the quantun efficiency of radical
 production in the presence of 02.

    If one compared the realtive  rates  of aldehydes
 and ketone photolysis under simulated  atmospheric
 conditions given by Carter et al.  (1978),  and
 uses  the atmospheric aldehyde photolysis rates
 given earlier,  one may estimate that the photolysis
 rate  of MEK in  the atmosphere is  ^ 10  percent h l.

 1.   Importance  of Aldehydes and Other Oxygenates
     in Modeling Atmospheric Chemistry

    The smog  chamber studies carried out under
 simulated atmospheric conditions  have  been
 mentioned earlier and adequately  demonstrate the
 importance of aldehydes  and other oxygenates in
 promoting photochemical  smog formation.  Computer
 modeling studies have further emphasized the
 importance of these  compounds.  Computer results
 have  been shown  to be significantly impacted by
 uncertainties in initial  concentrations and photo-
 chemical  parameters  such  as quantum yields  (Niki
 et  al.,  1972; Demerjian et  al., 1974;  Dodge and
 Hecht 1975;  MacCracken and  Sauter,  1975; Graedel
 et  al.  1976;  Whitten  et al.  1976; Dodge and
 Whitten  1976; Baldwin et  al.  1977;  Carter  et al.
 1978).   Thus  Dodge and Hecht (1975) state  that
 aldehyde  photolysis  is among the  most  critical
 reactions  for quantitative  photochemical smog
 modeling  when one  combines  the  sensitivity  with
 the uncertainty  in the rates and  mechanism  of the
 reactions.

    The situation  is more complex  for modeling
 atmospheric  conditions.  Uncertainties  in the
 photochemistry and ambient  concentrations are
 compounded by ill-defined emission  rates for
 aldehydes  and other oxygenates.   These  combined
 uncertainties can  have a significant impact upon
 model calculations.   For example, figure 9  shows
 the results of two runs carried out with an
 atmospheric trajectory model being  developed by
 Environmental Research and  Technology  (ERT),  under
 Coordinating Research Council,(CRC) funding.   The
 photodissociation  rates used in the model are those
 given by Peterson  (1976).   This model  is being
 developed using data  taken  during the  Los Angeles
 Reactive Pollutant Program  (LARPP)  carried  out  in
 1973  (Martinez and Parker,  1976).  The model
partitions the hydrocarbons  into  five classes
including separate classes  for HCHO and other
aldehydes, RCHO.   It  is evident from the results
shown in figure 9, that the two parameters  which
we varied in this calculation can have  a
significant impact on the results.  Of  course,
each parameter must be can'ed separately to
isolate individual effects.  The  choice of  initial
 aldehyde concentrations is  important  in controlling
 the radical concentration available to promote
 N02 formation, but the absolute effect is modified
                                                   41

-------
                                   kHCHO~2'0kRCHO

                                 [ALDEHYDES]g = 3.6 pphm

                              [FORMAI_DEHYDE]O = 2.5 pphm
                                    kHCHO~kRCHO
                                 [ALDEHYDES]O =

                              [FORMALDEHYDE] o = 2.5 pphm
                        12 noon

                  LOCAL TIME (minutes)
                                        5:20pm
 Fig. 9.  Effect of aldehyde photolysis rates and
          initial concentrations on a trajectory
          model run for November 5, 1973 in the
          Los Angeles basin.

 by the source input fluxes for aldehydes.

                     8.   Summary

    The above discussions have covered the  sources,
 ambient concentrations, radical reactions  and
 photochemistry of aldehydes,  and to a lesser
 extent, ketones, alcohols, ethers  and esters.   In
 general, the photolysis and reactions with the OH
 radical appear to be the major sinks  for oxygenat-
 ed hydrocarbons in the  lower  atmosphere.   The
 state  of knowledge of OH radical  reactions with
 these  oxygenates is currently adequate for modeling
 purposes given the uncertainties  in other  areas.
 Consequently, a major thrust  towards  further
 refinement  in these rate constants specifically
 for tropospheric modeling purposes has  little  merit
 and attention should be focused on other areas.
 These  areas  may be summarized under the  general
 categories  below.

    Ambient  Concentrations.  Much greater emphasis
 should  be placed upon obtaining concentration-time
 profiles  for aldehydes  and ketones in  the
 atmosphere  as  a function  of location,  season  and
 time of day.   The  increase in  chemical sophistica-
 tion of atmospheric  models, which  have been
 validated using well characterized smog chamber
 data, places  increased  demands  on  the quality and
 extent  of ambient  air quality measurements.  This
 is  typified  in  the  case of aldehydes  and ketones
 for which chemical  reactions may exist in  the
 computer model  mechanism,  but for  which no ambient
 air quality  data are available  either for  initial
 conditions or to test the  predicted concentration-
 time behavior of these  pollutants.

    A greater knowledge  of  aldehyde and other
 oxygenate emissions  is  also needed to form the
 basis of a good emission inventory for modeling
 purposes.

    It is realized, of course,  that the above
 requirements have not been met  in many locations
 for the common hydrocarbon classes of alkenes,
alkenes and arenes and in some cases,  not even for
non-methane hydrocarbons.
    Photooxidation.   The photodissociation of
 aldehydes and ketones appears to be the major
 depletion mechanism for these compounds in the
 lower atmosphere based on the calculations present-
 ed earlier.   Although there has been significant
 advances in  our knowledge of formaldehyde quantum
 yields,  the  ambient photolysis rates of other
 aldehydes, and particularly ketones, are poorly
 known.   Consequently, further studies of the
 photolysis of aldehydes and ketones as a function
 of pressure  up to atmospheric, and in the presence
 of 02 should be carried out.  The uncertainty in
 quantum  yields need to be reduced substantially
 to about ± 25 percent, since the oxygenate photo-
 lysis steps  are important in radical production.

    Kinetics  and Mechanism.   As indicated above,
 the general  status  of knowledge for the most
 important free radical, OH, is satisfactory for
 modeling purposes.   The mechanism of reaction
 should receive further attention in the areas of:
 HCO oxidation  under ambient conditions; OH and
 H02  addition  to formaldehyde as suggested  by
 Horowitz,  Su  and Calvert (1978); the oxidation of
 aromatic aldehydes  under ambient conditions,  and
 the  photooxidation  of ketones and other oxygenates
 under ambient  conditions.

    Finally, the kinetics  of the possible HSOi,
 radical  reactions with  aldehydes and ketones
 should be  studied to  test the suggestion of  Benson
 (1978).

                    Acknowledgments

   The trajectory modeling  work is  being performed
 under Coordinating  Research Council  funding.   The
 author wishes  to thank  Doctors  Roger Atkinson,
 Karen Darnall  and Arthur  Winer  of the Statewide Air
 Pollution  Research  Center,  University of California,
 Riverside  for  helpful comments  on this  manuscript
 and for  access  to results prior to  publication;
 Professor  J.  G. Calvert of  the  Ohio  State University
 and Dr.  G. K.  Moortgat of the Max-Planck Institute,
 Mainz, for receipt  of results prior  to  publication;
and the  Chemical Kinetics Data  Center of the Nation-
al Bureau  of Standards for  providing  a  bibliography
on aldehyde photooxidation.

                    References

 Akimoto,  H. and Kondo,  J.,  Advan.  Env.  Sci.  and
    Tech..  5., 1  (1975).      	

 Altshuller, A.  P. and Bufalini, J.  J.,  J.  Photo-
    chem.  and  Photobiol.  4.,  97  (1965).

 Altshuller, A.  P. and Bufalini, J.  J.,  Environ.
    Sci.  and Technol. 5.,  39  (1971).

 Altshuller, A.  P. and Cohen,  I.  R.,  Presented  at
    the  145th  National Meeting  of the American
    Chemical  Society, New York,  (1963).

Altshuller, A.  P.,  Cohen, I.  R.,  and  Purcell
    T. C. , Can. J.  Chem.  44,  2973  (1966).

Altshuller, A.  P.,  Cohen, I.  R.,  and  Purcell
    T. C., Science  156, 937  (1967).

Altshuller, A.  P. and McPherson,  S.  P   j  Air
    Poll.  Control Assoc.  13., 109  (1963).  	
                                                   42

-------
Altshuller, A. P., Miller, D. L., and Sleva, S. E.,
    Anal. Chem. Q, 621 (1961).

Archer, A. S., Cundall, R. B., and Palmer. T. F.,
    Proc. Roy Soc. A334. 411  (1973).

Atkinson, R., Darnall, K. R., Lloyd, A. C., Winer,
    A. M., and Pitts,  J. N.,  Jr., Advances in
    Photochemistry, in press  (19787!

Atkinson, R., Perry, R. A., and Pitts, J. N., Jr.,
    J. Chem. Phys. 66, 1197 (1977).

Atkinson, R. and Pitts, J. N., Jr., J. Chem. Phys.
    63_, 3591 (1975).

Atkinson, R. and Pitts, J. N., Jr., J. Chem. Phys.,
    accepted for publication  (1978).

Baldwin, A. C., Barker, J. R., Golden, D. M., and
    Hendry, D. G., J.  Phys. Chem. 81_, 2483 (1977).

Baldwin, R. R., Fuller, A. R., Longthorn, D., and
    Walker, R. W., J.  Chem. Soc., Faraday Trans.  I
    68, 1362 (1972).

Barker, J. R., Benson, S. W., and Golden, D. M.,
    Int. J. Chem. Kinetics 9. 31  (1977).

Benson, S. W., Chemical Reviews 78. 23 (1978).

Black, F., Proceedings of the International
    Conference on Phtochemical Oxidant Pollution
    and  Its Control, Volume  II, 1053  (1977).

Brunelle, M. F., Dickinson, J. E. and Hamming,
    W. J., Effectiveness of  Organic Solvents in
    Photochemical Smog Formation, Solvent
    Project Final Report, County  of Los Angeles
    Air Pollution Control District, July 1966.

Bufalini, J. J. and Brubaker, K.  L., The Photo-
    oxidation of Formaldehyde at  Low Partial
    Pressures, in Chemical Reactions in Urban
    Atmospheres, C. S. Tuesday, ed., p. 225
    (American Elsevier Publishing Co., New York,
    1971).

Cadle, R. D.  and  Powers,  J.  W.,  J. Phys.  Chem.
    71, 1702 (1967).

Cadle, R. D., Lin, S.  S.,  and Hausman,  R.  F., Jr.,
    Chemosphere j. 15  (1972).

California Air Resources Board,  Staff Report
    (January 1972).

Calvert, J. G., Kerr, J.  A., Demerjian, K. L.,  and
    McQuigg, R. D.,  Science 175. 751 (1972).

Calvert, J. G. and Pitts, J. N., Jr., Photo-
    chemistry, (John Wiley and Sons, Inc., New
    York, 1966).

Campbell, I. M.,  Handy, B. J., and Kirby, R. M.,
    J. Chem. Soc. Faraday Trans.  I 71, 867 (1975).

Campbell, I. M.,  McLaughlin,  D.  F., and Handy,
    B. J., Chem.  Phys. Lett.  38.,  362 (1976).

Campbell, I. M. and Parkinson, P. E., Chem. Phys.
    Lett.. in press (1977).
 Carter,  W.  P., Lloyd, A. C. , Sprung, J. L., and
     Pitts,  J.  N., Jr., Accepted for publication
     in Int.  J. of Chem.  Kinetics, (1978).

 Clark, J.  H.,  Ph.D.  Thesis,  University of
     California, Berkeley (1976).

 Cleveland,  W.  S., Graedel ,  T.  E., and Kleiner, B.,
     Atmos.  Environment 11 .  (1977).

 Cohen, I.  R.  and Purcell,  T.  C. , Atmos. Environ-
     ment 1 ,  689 (1967);  Environmental  Science
     and  Technology 1, 845  (1967).

 Cox, R.  A.,  Derwent,  R.  G. ,  Holt,  P.  M. , and Kerr,
     J. A.,  J.  Chem.  Soc. Faraday Trans.  I  72,
     2061
 Cox,  R.  A.  and  Roffey,  M.  J.,  Environ.  Sci .  and
    Tech. U_, 900  (1977).

 Cvetanovic,  R.  J.,  Can.  J.  Chem.  3^,  775  (1956).

 DeGraff,  B.  A.  and  Calvert  J.  G. ,  J.  Amer.  Chem.
    Soc.  89_, 2247  (1967).

 Demerjian,  K. L. ,  Kerr,  J.  A.,  and Calvert,  J.  G.,
    Advan.  Environ.  Sci. Techno!.  4^ (1974).

 Dimitriades, B.  and  Wesson, T.  C.,  J. Air Pollu-
    tion  Control Assoc.  22_, 33  (1972T

 Dodge, M. C. and Hecht,  T. , Int. J. Chem. Kin.
    Symp. No. 1, 155  (1975).

 Elliot,  M. A.,  Nebel, G. V., and Rounds, F.  G.,
    J. Air Pollution  Control Assoc. 5_,  103  (1955).

 Faubel ,  C., Wagner,  H. Gg.  and  Hack,  W. , Ber.
    Bunsenqes.  Phys.  Chem.  81_,  689  (1977).

 Fracchio, M. F.  , Schuetts,  F. J., and Mueller,
    P. K. , Environ.  Sci. Techno!.  1:915-922
    (1967).

Gaffney,  J.  S. ,  Atkinson, R. ,  and Pitts, J.  N.,
    Jr.,  J.  Amer. Chem.  Soc. 9_7, 5049 (1975).

Gordon, S. and Mulac, W. A., Int. J. Chem.  Kinet.
    Symp. No. 1, 289  (1975).

Gorin, E. , J. Chem. Phys. 7_, 256 (1939).

Graedel,  T.  E.,  Farrow,  L.  A., and Weber, T. A.,
    Atmos. Env.  10., 1095 (1976).

Grob,  K.  and Grob,  G. , J. Chromatogr. 6£, (1),
    1-13  (1971).

Haagen-Smith, A. J. and Fox, M. M. , Ind. Enq.
    Chem. 48, 1484 (1956).

Hanst, P. L. and Gay, B. W. , Env.  Sci.  and Tech.
    11, 1105 (1977).

Hanst, P. L., Wilson, W. E., Patterson, R.  K. ,
    Gay,  B. W.,  Chaney,  L.  W.,  and  Burton, C.  S.,
    A Spectroscopic  Study of California Smog,
    U. S. Environmental   Protection  Agency,  EPA
    Publication No.  650/4-75-0006  (February  1975).

Hecht, T. , Seinfeld,  J., and Dodge, M.  C.,  Environ.
                                                   43

-------
    Sci.  and Tech. 8, 327 (1974).

Hendry, D. G. and Mabey, Private Communication
    (1973).

Hendry, D. G. and Kenley, R. A., J. Amer. Chem.
    Soc.  99., 3198 (1977).

Herron, J. T. and Huie, R. E., J. Phys. Chem.
    Ref.  Data 2, 467 (1974).

Herron, J. T. and Penzhorn, R. D., J. Phys. Chem.
    73_, 191  (1969).

Horowitz, A. and Calvert, J. G., Int. J. Chem.
    Kin., in press (1978a).

Horowitz, A., Su, F., and Calvert,  J. G., in press
    (1978).

Hunziker,  H. E.  and  Wendt, H.  R., Near  Infrared
    Electronic  Absorption of Organic  Peroxyl
    Radicals, Paper  No. 48, 12th International
    Symposium on Free Radicals, Laguna Beach,
    California,  (January 4-9,  1976).

Jaffe, S. and Wan, E.,  Environ. Sci. Techno!. 8_,
    1024  (1974).

Johnston, H. S.  and  Heicklen,  J., J. Amer. Chem.
    Soc. 86, 4254  (1964).

Katou, T.,  Proceedings  of the  Int. Symp. on Air
    Pollution,  p. 419,  Tokyo (1972).

Katz, M.,  in Air Pollution (Columbia Univ. Press,
    New York, 1961).

Kelly, N.  and Heicklen, J., J. of Photochemistry
    8, 83  (1978).

Klein, R.  and Schoen, L. J., J. Chem. Phys. 24_,
    1094  (1956).

Laity, J. L., Burstein, I. 6., and Appel, B. R.,
    in Solvents:  Theory and Practice, Advances
    in Chemistry Series No. 124, American
    Chemical Society, Washington, D.C., (1973).

Lande, S. S., Durkin, P. R., Christopher, D. H.,
    Howard,  P.  H., and  Saxena, J., Investigation
    of Selected  Environmental  Contaminants:
    Ketonic  Solvents, prepared for EPA, Office
    of Toxic Substances, PB 252 970  (May 1976).

Leighton, P. A., Photochemistry of Air Pollution
    (Academic Press, New York, 1961).

Leighton, P. A.  and  Perkins, W. A., Conference on
    Chemical Reactions  in Urban Atmospheres, Air
    Pollution Foundation, Report No. 15 (1956).

Levy, H., Advances in Photochemistry 9, 364 (1974).

Levy, A. and Miller, S. W., Final Technical
    Report  on the Role  of Solvents in Photo-
    chemical Smog Formation, National Paint,
    Varnish, and Lacquer Association,
    Washington,  D.C. (1970).

Lewis, R. S. and Lee, E. K. C., J.  Phys. Chem.
    82, 249  (1978).
 Lewis,  R. S., Tang,  K.  Y., and  Lee,  E.  K.  C,,
     J.  Chem. Phvs. 65_,  2910  (1976).

 Linell,  R.  H. and Scott, W.  E.,  Arch.  Environ.
     Health  5_. 616 (l9620.

 Lloyd,  A. C., Int. J. Chem.  Kinetics 6,  169
     (1974).              ~

 Lloyd,  A. C., Darnall,  K.  R.,  Winer, A.  M.,  and
     Pitts,  J. N., Jr.,  Chem.  Phys.  Lett.  42,
     205  (1976).                           ~

 Lonneman, W. A., Bufalini, J.  J., and  Seila, R. L.
     Environ. Sci. and Tech.  lp_,  374  (1976).

 Los  Angeles  Control  District,  Second Technical
     and Administrative  Report,  (1950-51).

 MacCagken,  M. C. and Sauter,  G.  D.,  ed.,
     Development of an Air  Pollution  Model  for
     the San  Francisco Bay Area,  Final  Report
     to  the  NSF, Vol. II, (October 1975).

Mack, G. P.  R.  and Thrush, B. A., J. Chem. Soc.
     Faraday Trans.  I 69, 1  (1973).

Mack, G. P.  R.  and Thrush, B. A., J. Chem. Soc.
     Faraday Trans.  I 69, 208  (1973T

Mack, G. P. R.  and Thrush, B. A., J. Chem. Soc.
     Faraday Trans.  I 70, 178  (1974~T

Marling, J., J.  Chem. Physics 66. 4200 (1977).

Martinez, J. R.  and Parker,  R. 0., Los Angeles
     Reactive Pollutant Program  (LARPP) Data
    Archiving and Retrieval, ERT Doc, No.
     P-1464W prepared for Coordinating Research
    Council   (1975).
McQuigg, R.  D.  and  Calvert, J. G., J. Amer. Chem.
    Soc. 91, 1590 (1969).

Moortgat, G., Private Communication, (March,  1978)

Moortgat, G. K., Slemr,  F., Seiler,  W., and
    Warneck, P., Chem.  Physics Lett., in press
     (1978a).             	

Moortgat, G. K., Slemr,  F., Seiler, W., and
    Warneck, P., Presented at the 13th Informal
     Photochemistry  Conference, Tampa, FL,
     (January 19785).

Morris, E.  D. ,  Jr.  and Niki, H., J.  Chem.  Phys.
    55_, 1991 (1971).

Morris, E.  D.,  Jr.  and Niki, H., J.  Phys.  Chem.
    Zi, 3640 (1971).

Morris, E.  D.,  Jr.  and Niki, H., J.  Phys.  Chem.
    78, 1337 (1974).

Morris, E.  D.,  Jr.,  Stedman, D. H.,  and Niki, H.,
    J.  Amer. Chem.  Soc.  93_, 3570 (1971).

National Academy of Sciences Vapor Phase  Organic
     Pollutants,  National Research Council  (1976).

Niki, H., Private Communication, (1978).
                                                   44

-------
Niki, H.,  Maker, P., Savage C., and Breitenbach, L.,  Schuck,  E.  A.,  Eye Irritation from Irradiated
    Presentation at the 173rd ACS National  Meeting,
    New Orleans, Louisiana, (March 20-25, 1977).

Niki, H.,  Daby, E.  E., and Weinstock, B., Advances
    in Chemistry Series No. 113, 16 (1972).

Niki, H.,  Daby, E.  E., and Weinstock, B., 12th Symp.
    Combustion p.  277 Comb. Institute (1969).

Niki, H.,  Maker, P. D., Savage, C. M.,  and
    Breitenbach, L. P., J. Phys. Chem,  in press
    (1978).

Oberdorfer,  P. E.,  Automotive Engineering Congress,
    Detroit, January 9-13, 1967.

Osif, T. L.  and Heicklen, J., J. Phys.  Chem.  80_,
    1526 (1976).

Osif, T. L., Simonaitis, R., and Heicklen,  J.,
    J. Photochem 4, 233 (1975).

Overend, R.  P. and Paraskevopoulos, G., Private
    Communication (1978).

Parmenter, C. S. and Noyes, W.  A., Jr., J.  Amer.
    Chem.  Soc. 85_,  416 (1963).

Pate, C. T,  Atkinson, R., and Pitts, J. N.,  Jr.,
    J. Environ. Sci. Health, Environ. Sci.  Eng.
    All. 19  (]976T-

Penkett, S.  A., Sandalls, F. J., and Lovelock,
    J. E., Atoms.  Env. 9_, 139 (1975).

Perry, R.  A., Atkinson, R., and Pitts,  J. N., Jr.,
    J. Chem. Phys.  64, 5314 (1976).

Perry, R.  A., Atkinson, R., and Pitts,  J. N.,  Jr.,
    J. Chem.  Phys.  67, 611 (1977).

Peterson,  J. T., Demerjian, K.  L., and  Schere,
    K. L., Proceedings of the International  Conf.
    on Photochemical Oxidant and Its Control,
    Raleigh^ N.C.  (1977).

Pitts, J.  N., Jr.,  J. Air Poll. Control Assoc.
    19., 658  (1969).

Pitts, J.  N., Jr.,  Winer, A. M., Darnall, K.  R.,
    Doyle, G.  J.,  and McAfee, J. M., Chemical
    Consequences of Air Quality Standards and of
    Control  Implementation Programs:  Roles  of
    Hydrocarbons,  Oxides of Nitrogen and  Aged
    Smog in  the Production of Photochemical
    Oxidant, University of California,  Riverside:
    California Air  Resources Board, Con-tract No.
    4-214, Final Report (May 1976).

Purcell, T.  C. and Cohen, I. R., Atmospheric
    Environment 1.  689 (1967).

Radford, H.  E., Evenson, K. M., and Howard,  C. J.,
    J. Chem. Phys.  60, 3178 (1974).

Renzetti,  N. A. and Bryan, R. J., J. Air Pollution
    Control  Assoc.  11_, 421 (1961).

Robinson,  E., Rasmussen, R. A., Westberg, H.  H.,
    and Holdren, M. W., J. Geophys. Res.  78,
    5345 (1973).
    Auto  Exhaust.   Air  Pollution  Foundation,
    San Mario,  CA,  Technical  Report No.  18 (1957).

Schuck, E. A. and Doyle,  G. J., "Photooxidation
    of Hydrocarbons in  Mixtures Containing
    Oxides of Nitrogen  and Sulfur Dioxide".   Air
    Pollution Foundation, San Marino,  CA.,  Report
    No. 29 (October 1959).

Scott, W. E. and Reckner, L.  R.,  "Atmospheric
    Reaction Studies  in the Los Angeles  Basin".
    Progress Report.  (November  15,  1968  to
    January 15, 1969).  Scott Research Laboratory.
    APRAC Project CAPA  7-68 (1969).

Seizinger, D. E. and  Dimitriades,  B.,  Journal of
    the Air Pollution Control Assoc. 22_, 47
    (1972).

Shibuya,  K., Ebata, T., Obi, K., and Tanaka,  I.,
    J. Phys. Chem. 81_, 2292 (1977).

Singleton, D. L., Irwin,  R. S., and Cvetanovic,
    R. J., Can. J. Chem.  55_, 3321   (1977).

Smoyer, J. C., Shaffer, D. E., and  DeWitt,  I.  L.,
    Inst.  Environ. Sci., Tech. Meet., Proc.,  17,
    339-45 (1971).

Sperling, H. P. and Toby, S., Can.  J. Chem. 51
    471 (1973).

Stanford Research Institute, Third  Interim Report
    on Studies to Determine the Nature and
    Sources of Smog to the Western  Oil  and Gas
    Association, p.  12, (November 1950).

Stephens, E. R., Adv.  in  Environ.  Sci.  and Tech.
    1, 119 (1960)'

Stephens, E. R. and Burneson, F.,  Private
    communication quoted  in Winer et al.  (1976).

Tadasa, K., Imai, N.,  and Inaba, T., Bull. Chem.
    Soc.  Japan 4£, 1758 (1976).

Trotman-Dickenson, A.  F. and Kerr,  J. A., in
    Handbook of Chemistry and Physics,  57th ed.
    (The Chemical Rubber Company,  Cleveland,
    1976/77).

Tucazon,  E. C., Graham, R. A., Winer, A.  M.,
    Easton, R.  R., Pitts, J. N., Jr., and Hanst,
    P. L., Atmos, Envir.. (1978a).

Tucazon,  E. C., Winer. A. M., Graham, R.  A.
    and Pitts,  J. N.,  Jr., Advances in Envir. Sci.
    and Tech. 10 in press (1978b).

Van Ham and Nieboer, H., T.  NO Nieuws.  5, 170
    (1972).

Washida,  N., Martinez, R. I., and Bayes, K. D.,
    Z. Naturforsch.  29A, 251 (1974).

Weaver, J., Meagher, J., and Heicklen,  J., J.
    Photochem.  6_, 111  (1977).

Whitten,  G. and Dodge, M. C., 197th American
    Chemical  Society Meeting, New York,  (1976).
                                                   45

-------
Wigg, E. E., Campion, R. J., and Petersen, W. L.,
    The Effect of Fuel Hydrocarbon Composition
    on Exhaust and Oxygenate Emissions.  SAE
    Paper No. 720251 Presented at the Society of
    Automotive Engineers Automotive Engineering
    Congress, Detroit, Michigan, January 10-14,
    1972.

Wilson, K. W. and Doyle, G. J., "Investigation of
    Photochemical Reactivities of Organic
    Solvents", Stanford Research Institute
    Report,  EPA  Contract HCPA 22-61-125
    (September 1970).

Winer, A. M., Private Communication. (1978).

Winer, A. M., Lloyd, A. C., Darnall, K. R., and
    Pitts, J. N., Jr., J.  Phys. Chem. 80, 1635
    (1976).

Winer, A. M., Lloyd, A. C., Darnall, K. R.,
    Atkinson, R., and Pitts, J. N., Jr., Chem.
    Phys. Lett.  51_, 221 (1977).

Wodkowski, C. S. and Weaver, E. E., The Effects
    of Engine Parameters,  Fuel Composition, and
    Control  Devices on Aldehyde Exhaust Emissions.
    Paper presented at West Coast Section, Air
    Pollution Control Association Meeting, San
    Francisco, CA, October 8, 1970, 32 pp.
              Summary  of  Session

   The discussion centered on  the  photolysis  of
formaldehyde   with and without added  oxygen.
Heicklen began the discussion  by characterizing
the photooxidation of aldehydes as the most
important unresolved problem in air pollution
chemistry.   He pointed out that there  appears to
be general  agreement on the  mechanism  and  quantum
yield for the photolysis of  formaldehyde  in  the
absence of oxygen, but that  when oxygen is added,
results from various studies suggest that  the
system is actually poorly understood.   Of  particu-
lar concern is that the results appear to  differ
from laboratory to laboratory, and sometimes  within
the same laboratory, as emphasized in  additional
remarks by Calvert.

   Warneck reported on results obtained in his
laboratory (with Moortgot, Glens,  and  Seiler) on
the uv photolysis of CH20.   These  experiments were
unique in that low formaldehyde concentrations
were used,  and both synthetic  air  and  pure
nitrogen buffer gasses were  added  at pressures up
to atmospheric.  These results (see fig.  7 of
Lloyd's review paper) appear to be in  general
agreement with other recent  results, within  the
uncomfortably large scatter  of the data.   At  355
nm, a pressure effect was observed, and there no
longer was 1 to 1 relation between H2  and  CO
production.  Warneck suggested this may be due to
an interaction of 02 with excited  formaldehyde,
possibly by hydrogen abstraction.

   Parkes noted that, during his study of the
photooxidation of methyl radicals  using a  Phillips
dark lamp (peak emission at  360 nm), the photolysis
of formaldehyde was a complication. The addition  of
an atmosphere of  isopentane  quenched the photolysis.
   At  shorter wavelengths, Calvert  noted  that  they
 observed  no quenching with up to  50 Torr  of  added
 isobutene.  At longer wavelength  (^ 390 nm)  the
 addition  of 360 Torr of C02 resulted  in quenching,
 as measured by the decrease in  H2 yield.   The
 addition  of isobutene resulted  in no  further
 change.

   There  was some discussion of the state from
 which  formaldehyde dissociated.   Ravishankara
 asked  if  a triple state could be  involved, in
 light  of  the effect of 02.

   Calvert replied that Ed Lee  had  tried  to
 distinguishe between triplet and  singlet  formalde-
 hyde by making the triplet via  energy transfer
 from mercury.  Apparently both  dissociation
 processes still occur.  Calvert felt  that the
 evidence  does not support triplet involvement, but
 that the  excited singlet crosses  to high  vibra-
 tional levels of the ground state which then
 dissociate.

   Heicklen questioned this explanation since  the
 addition  of an atmosphere of nitrogen resulted in
 no quenching of the formaldehyde  dissociation.

   Ravishankara mentioned some  work by George
 Atkinson  on dye laser photolysis  of H2CO,  monitor-
 ing CO production by resonance  emission.

   The question of HCO oxidation, which is directly
 related to the observation of HCOOH in the photo-
 oxidation of CH20, was discussed  by Heicklen and
 Niki.  In his contributed remarks,  Heicklen,
 discussed the problems associated with the reaction
 of HCO with 02.  The direct evidence  suggests  that
 the reaction products are CO and  H02.

   The photolysis of C12-02-CH20  system leads  to
 HCOOH, however, which has been  interpreted as
 evidence  for the production of  HC03.   Niki revealed
 that they have found peroxynitric acid by  adding
 N02 to the C12-02-CH20 system,  further supporting
 the production of H02.  In addition,  they  have
 investigated the possibility that HCOOH is produced
 in the reaction OH + CH20.  They  photolyzed  nitrous
 acid and  formaldehyde and found very  little  formic
 acid,  suggesting that OH + CH20 is  not the source
 of HCOOH.  The reaction H02 + CH20  is  still
 considered a possibility.

   The discussion ended with some remarks  about
 some sources of aldehydes which could  become more
 important in the future.  Calvert noted that the
 use of CH3OH as a fuel in internal  combustion
 engines leads to large CH20 emission.  Demerjian
 added  that diesels produce large  amounts  of  heaver
 aldehyde.
                    Comments
Julian Heicklen,  Department of Chemistry, The
Pennsylvania State University, University Park,
Pennsylvania  16802

   Moortgat et al.  (presented at 13th Informal
Conference on Photochemistry, Clearwater Beach,
Florida, 1978) have determined the primary photo-
chemical processes  in CH20 photooxidation in an
excellent study.   At 3130 A they find the free
                                                   46

-------
radical process (HCO + H) to occur with    0.80
and the molecular process (H2 + CO) to occur with
ij> = 0.20.   However there are still some naging
problems.

   The ratio of the two paths of 4.0 is consistent
with the value of 3.2 found by Osif (Ph.D. thesis,
Pennsylvania State University, 1976), but is
sufficiently greater than the value of 2.1 found
in the absence of 02 (A. Horowitz and J. G. Calvert,
Intern. J. Chem. Kinetics, in press) to suggest
that the primary process must be altered in the
presence of 02.  This is not unexpected since the
photochemistry does not occur from the initially
formed excited singlet state but proceeds with an
induction  period (P.  Avouris, W.  M. Gelbart, and
M. A. El-Sayed, Chem. Rev..  77., 793, 1977).
Presumably the photochemistry proceeds through a
triplet (as in other aldehydes) or another
intermediate (HCOH) which can be attacked by 02-
If so the  formation of molecular H2 is surprising.

   A further baffling point is the results
reported by Osif; and Horowitz and Calvert; and
Su, Horowitz, and Calvert (13th Informal Conference
on Photochemistry, Clearwater Beach, Florida,
January, 1978) that cj){CO} could exceed 6 in the
presence of 02.  Clearly these high yields indicate
some chain process which may be due to surface
affects.

   Finally Morrison in our laboratory has studied
CH20 photooxidation at 3130 ft in a Teflon-lined
11-liter cell and finds that 02 quenches all CO
and H2 formation with a half-quenching pressure of
4 Torr.  However HCOOH is produced.  This result
appears to be completely at  variance with all
other studies.
Julian Heicklen,  Department  of Chemistry,  The
Pennsylvania  State  University,  University  Park,
Pennsylvania   16802

   There are  two  problems in HCO oxidation to be
considered here:   the rate coefficient and products
of the reaction.   There are  three measurements of
the rate coefficient.   Washida, Martinez,  and Bayes
(1. Naturforsch 29A,  1974 and Shibuya, Ebata, Obi,
and Tanaka (J. Phys.  Chem. 81_, 2292,  1977) reported
5.6 x lo~12 cm3/s at  room temperature while Clark,
Moore, and Reilly (Int.  J. Chem.  Kinetics  10, 427,
1978) obtain  4.0  ± 0.8 x lo"12 cm3/s.Thus there
is about a 30 percent uncertainty in  this  number.

   Of more significance  is the fact that Shibuya
et al.,  obtain no pressure dependence  for  this
rate coefficient, and conclude that there  is  no
addition reaction of  02  to HCO and that H  abstrac-
tion is  the sole  reaction path.   This  conforms to
Hunziker's observation reported at the 12th
Informal  Conference on Photochemistry  (June,  1976)
that he  could not find an absorption  due to HC03
in the HCO-02 reaction.

   This  leaves a  puzzling phenomenon  since both
Osif and Heicklen (J.  Phys.  Chem. 80.,  1526, 1976)
and Niki,  Maker,  Savage, and Breitenbach  (reported
at the 173rd American Chemistry Society Meeting,
New Orleans, March, 1977) found HCOOH in the Cl-
CH20-02 system, which they considered as evidence
for HC03 as a precursor.  HCOOH has also been  seen
in other systems, as well as  in polluted atmospheres
If HCOOH does not come from HCO oxidation,  then
where does  it come from?
 A.  R.  Ravishankara,  Applied Science Laboratories,
 EES,  Georgia Institute of Technology,  Atlanta,
 Georgia   30332

   George Atkinson has measured CO  production in
 H2CO flash photolysis  and  it seems  to  agree with
 most other data.

   We  have measured  OH +  CH3OH  (and also OH +
 C2H5OH) using  flash  photolysis    resonance
 fluorescence.   It agrees well with  the indirect
 measurements of Campbell.   So the rate constants
 for the other  alcohols  (measured  by Campbell)
 are also  probably correct.
                Recommendations
1.       Role of Aldehydes

1.1     Kinetics.   We have classified the status of
        the rate data into the following caegories:

        Satisfactory (S)   no need for further work

        Less Satisfactory (L.S.)   further work
        is desirable but not of high priority

        Deficient (D)   more work is needed

        Insufficient Data (I.D.)

1.1. (a) Hydroxyl Radical Reactions  HO + RCHO -»•
        H20 + RCO
        HCHO
        CH3CHO
        C2H5CHO
        CH2 = CHCHO
        C6H5CHO
L.S.
S
S
I.D.
S
        In the case of C2H5CHO and C6HsCHO the rate
        constants are from single studies which are
        probably reliable, since the same authors
        have also studied HCHO and CH3CHO with
        consistent results.  It would be useful to
        have a second, independent determination
        (say be flash photolysis experiments) to
        confirm the data for these aldehydes.

1.1.(b) Oxygen Atom Reactions  0(3P) + RCHO *
        HO + RCO

        HCHO
        CH3CHO
        C2H5CHO
        n-C3H7CHO
        i-C3H7CHO
        CH2 - CHCHO
        CH3CH   CHCHO
S
D
D
I.D.
I.D.
L.S.
S
                                                   47

-------
1.1.(c)  H02 Reactions
        There are no experimental data available,
        and the present estimated kinetic paramters
        need to be substantiated by experiment.
1.1.(d) CH30 Reactions
        It would be useful to confirm the relative
        rate data relating to the reaction CH30 +
        CH3CHO with absolute rate data on CH30 +
        aldehydes.   Such reactions are probably
        more of kinetic interest than of importance
        to atmospheric modeling.

1.1.(e) N03 Reactions

        More data are needed on this class of
        reactions,  which might be of importance
        under special conditions.

1.1.(f) HSCK Reactions

        No kinetic data exist for this species,
        which could  be important in atmospheric
        chemistry.,

1.1.(g) Formyl  Radical Reactions

        The major  reaction appears to be

             HCO + 02 + CO + H02

        for which  there are reasonably consistent
        room temperature data.  More work is needed
        on this reaction, particularly in relation
        to the  formation of HCOOH.

1.2     Photolysis

1.2.(a) HCHO

        The photooxidation of HCHO is tied in with
        the above  reaction:  HCO + 02 -* CO + H02-
        The mechanism of HCOOH formation appears
        to be dependent upon the experimental
        conditions,  e.g., under conditions of high
        [HCHO]  and low total pressure, the yields
        of HCOOH are relatively high.  Further
        work is needed to ascertain if HCOOH is an
        important product under atmospheric
        conditions.

        The quantum yield data for the free radical
        fragmentation process are now in satis-
        factory agreement.

        It might be useful  to have confirmation of
        the single set of cross-section data for
        HCHO.
        It would be useful to carry out further
        work to confirm the existing photochemical
        data on CH3CHO.

1.2.(c) Higher Aldehydes

        Photochemical  studies of the higher alde-
        hydes would be useful, but are not of
        prime importance to atmospheric chemistry.
2.

2.1
Role of Ketones
Kinetics
2.1.(a) Hydroxyl Radical Reactions

        It would be useful to confirm the existing
        data by further study and and at the same
        time to obtain information on the mechan-
        isms of the photooxidation of the lower
        ketones.

2.2     Photochemistry

        Confirmation is needed both for the exist-
        ing quantum yield measurements and for the
        photoabsorption cross-sections, preferably
        under atmospheric conditions.

3.      Role of other Oxygenated Species

        Alcohols, ethers and esters appear to be
        of little importance in the lower
        atmosphere.
                                                   48

-------
      Session III
Organic Free Radicals

-------
                                         ORGANIC  FREE  RADICALS
                                            David  M.  Golden

                                           SRI  International
                                     Menlo  Park, California   94025
        The role of  free  radicals  in  the  chemistry  of the lower troposphere is reviewed.   Methods
     of predicting and estimating  kinetic parameters  are  discussed with  particular reference to
     alkoxyl radical decomposition,  isomerization,  and reaction with oxygen.   Data needs, accuracy
     and priorities  are considered.

     Keywords:  Alkoxyl;  kinetics;  radicals;  review; troposphere.
              Introduction

   This  paper,  specifically prepared for the
Workshop on  Chemical  Kinetic Data Needs  for
Modeling the Lower Troposphere,  is built around
several  key  questions  proposed to the speakers by
the organizers.

1.  Why  is this  topic  important  with respect
    to the chemistry  of the lower troposphere?

   A simplified general  scheme for understanding
the chemistry of the  lower  troposphere is given
in figure 1.   We see  the role of organic free
radical  chemsitry in  those  mechanisms, and we
are quickly  led to understand that modeling of
this complex chemistry will  require knowledge of
many rate constants involving organic radical
species, both aliphatic and aromatic.   In fact,
we can readily see that the numbers of individual
rate constants  which will be needed is enormous.
Thus, we need to be able to make reliable predic-
tions and estimations  based on a carefully
selected data base.
   NO
  , + hi/


0 + NO

  U
                                      R'
                              N0
                   R0 + N02 -«-
                   Aldehyde (Ketone) + H02
                                     • RO,
                           N0
 Fig. 1.  A simplified scheme  for the chemistry  of
         the lower troposphere.
2.  What is the current status of our knowledge?

   This question is best addressed in terms of a
framework for codification and extrapolation of
kinetic data which allows us to choose the pursuit
of specific data from which we can infer the
maximum amount of new information.

   A slight digression to remind us of the method
involved seems in order here.

   Thermochemistry.  It is impossible to begin a
discussion of the theoretical basis for critical
evaluation and extrapolation of thermal rate data
without first discussing methods for estimating
thermochemical quantities, such as AH£ T, ASS,
and C° T for molecules.
     P.T
   Group Additivity.   When a sufficient data base
exists, we have found [I]1 the method of group
additivity to best fit the need for accuracy and
ease of operation.   The basic concept and
assumptions involved in the group additivity
method are as follows:

   For the disproportionation reaction

     RNN'R + SNN'S J RNN'S + SNN'R

any additivity approximation assumes that A$ =
A$ , where $ is any molecular property, and A*
is the contribution to that property due to   °
symmetry changes and optical  isomerism.  For the
molecular properties of interest here, AHT ->• 0,
AC  T  ->- 0, and AST -* S0   R In K , where K

o-(RRNN'R)a(SNN'S)/a(RNN'S)a,SNN'R), a(X) being the
symmetry number including both internal and
external  symmetry.   An additional term for entropy
of mixing, due to the existence of optical isomers,
must also be included.
  Figures  in brackets indicate literature
  references at the end of this paper.
                                                   51

-------
   If the molecular framework NN'  is two atoms or
greater, these relationships imply the additivity
of group properties, which include all nearest-
neighbor interactions, since a group is defined
as an atom together with its ligands (e.g.,  in
the group C-(H)3(C), the central C atom is bonded
to three H atoms and one C atom).  Thus, the
equation

     CH3OH + CH3CH2OCH3 + CH3CH2OH + CH3OCH3

implies the additivity of the properties of  the
groups C-(H)3(C), C-(H)3(0), 0-(C)(H), C-(H)2(C)(0),
and 0-(C)2, if the appropriate A$  = Ao.

   We have developed group additivity methods that
permit the estimation, for many  organic chemicals
in the gas phase, of heats of formation + 1
kcal/mol, and of entropies and heat capacities to
± 1 cal/(mol-K), from which free energies of
formation can be derived to better than ± 2  kcal/
mol.

   It should be noted that entropy and heat
capacity are molecular properties that can be
accurately estimated under much less stringent
conditions than energy (or enthalpy).   Thus,  the
method of bond additivity seems to work quite
well (± 1  cal/(mol-K))  for estimating the former
properties, but not at all  well  (± 4 kcal/mol)
for the latter.

   Structural  Considerations and Model  Compounds.
If sufficient thermochemical data is lacking for
the estimation of group properties, entropy,  and
heat capacity can often be adequately estimated
from structural  parameters of the molecule.
(Enthalpy estimates are more difficult, requiring
a better knowledge of potential functions than
are usually available).  The methods of statistical
thermodynamics may be used to calculate C° and
S° directly for those molecules where a  p
complete vibrational assignment can be made or
estimated.

   Also,  "reasonable'1 structural  and vibrational
frequency "corrections" to the corresponding
established thermodynamic properties  of "reference"
compounds  may  be made.   A suitable choice  of
reference  compound,  i.e., one  similar in  mass
size and  structure  to the unknown,  assures  that
the external  rotational  and translational  entropies
and heat  capacities  of the reference  and  unknown
compounds  will  be the same and that many  of the
vibrational  frequencies will  be similar.   The
basic assumption is  that S° and  C°  difference can
be closely estimated by considering only  low-
frequency  motions thought to be  significantly
changed in the  unknown.   Fortunately,  entropies
and heat  capacities  are not excessively sensitive
to the exact choice  of these vibrational
frequencies, and estimates  of  moderate  accuracy
may be made with relative ease.

   Kinetics.   The extension of thermochemical
estimation techniques to  the evaluation of  kinetic
data rests largely  on the validity  of transition
state theory.

   The transition state theory expression  for a
thermal  rate constant is:
          (kT/h)  expC-AGft/RT]
 (The units are s"1  for first order and atm l s ]
 for second order).   And
            +  [(T-300)    Tln(T/300)].

 (In  the ideal  gas approximation we can drop the
 standard state notation of AH? and ACJjj).   If the
 empirical  temperature dependence is represented
 by

              k   ATB exp(-C/T)

 A=   k/h(300)  exp[(As|00-)/R]

 B =  (<&$> + R)/R

 c    (AH!OO   (300))/R

 k    Boltzmann's  constant

 h    Planck's  constant
ASSj^Q     entropy  of activation  at  300 K,  standard
          state  of 1  atm.

AH|OO     enthalpy of activation at 300 K.


     average  value  of the heat capacity  at
          constant pressure of activation  over the
          temperature range 300  T K.

    If we  wish to  express  second-order rate
constants  in concentration units instead  of
pressure  units, we  must multiply by  RT in the
appropriate units.   This  has the effect of
writing:
     k   A'T   exp(-C/T)

where A'    AR and B'    B + 1  =
                                ( + 2R)/R
   Thus, simple "Arrhenius behavior" which will
be sufficient for lower tropospheric temperature
is characterized for first-order reaction by
AC+ = -R; (AC+   AC+ = ACJ), and for second-order
reactions using concentration units by AC? = -2R
(or AC|   -R).                           p

   In the case of simple Arrhenius behavior:

     k = A exp(-B/T)

log  A = log(ek/h) + AST-/R;

     B = (AHl- + R)/R

   Thus, the quantities AH-t, ASt, and AC $ are  of
interest.  We apply similar methods to those
already discussed with respect to thermochemistry
to view rate data in a reactional framework.
These rechniques are discussed in some detail  by
Benson [2], but certain points are worthy of re-
emphasis here.
                                                   52

-------
   We begin by classifying reactions as unimolecular  state which becomes tighter as the temperature
or bimolecular.  (The only termolecular processes     rises.
of interest to us will be energy transfer
controlled bimolecular processes).
   Unimolecular Processes.

   Simple Fission:  AB -»• A + B

   Complex Fission:  Molecule -*• Molecule +
                     Molecule (or radical)

   Isomerization:  Intramolecular atom rearrange-
                   ment

   Bimolecular Processes.

   Direct Metathesis:  A + BX * AX + B

   Addition:  A + Molecule + Stable Adduct
              (reverse of complex fission)

   Association:  A + B •* A   B (reverse of
                 simple fission)

   The first thing to notice is that of all these
 reactions, only direct metathesis reactions are
 not subject to becoming energy transfer limited
 at high temperatures and low pressures (i.e., in
 the "fall-off" region!).  This means that not
 only does the so-called high pressure rate
 constant need to be estimated or known, but the
 extent of fall-off, as well.  Methods are
 available for making fall-off correction [3].

   In hydrocarbon reactions in the troposphere,
 including those of aromatic compounds, we may
 expect that most direct metathesis reactions
 will involve the exchange of a hydrogen atom
 between larger groups.  A simple, semi-empirical
 prescription exists for estimating the value of
 AS  for these types of reactions.  First, one
 realizes that these values are limited between
 the "loosest" possible model (A-factor equals gas
 kinetic collision frequency) and the "tightest"
 possible model in which R"«H"«R'  is represented
 by the molecular R-R'.  Experience using data in
 the 300 < T/K < 700 has taught us that generally
 the AS  value corresponds to a transition state
 only slightly looser than the tightest possible
 value.

   Since the other two classes of bimolecular
 processes are the reverse of unimolecular reac-
 tions,  we may consider them in that direction.
 (The equilibrium constant is either known or
 estimable).   Once again using experimental  results
 as our guide, we note that model  transition state
 which correspond to the values of AS  are genera-
 lly "tight".  That is, we may visualize them as
 minor modifications of the reactant molecule,
 usually involving some increase in rotational
 entropy due to slight enlargement of certain
 bonds.   The dominant entropic feature is usually
 the stiffening of internal rotations as a result
 of multiple bond formation or ring formation [2].

   Bond  scission  reactions  present  a particular
problem,  since it  is  particularly difficult to
locate a  transition  state.   Recent  work [2,4],
both experimental  and  theoretical,  indicates that
these reactions  can  be modeled with a transition
    An  example  of the  use  of these methods to
 evaluate  rate  constants for modeling the lower
 troposphere  is  taken  from Barker et al.  [5]:

    Alkoxyl Radical  Decomposition Reactions.   The
 decomposition  reactions of alkoxyl  radicals
 provide a  good  example of a family of reactions
 for which  an adequate number of accurate studies
 have been  made.   Most of  the studies have been
 made on t-butoxyl  radicals, but several  other
 radicals  have  been  studied as well.   All  the
 studies were determinations of relative  rate
 constants  and  so we returned to the original data
 and recomputed  it on  the  basis of current values
 for the reference reaction rate constants.

    Three  different  reference reactions have  been
 used:
 RI
  ONO

^iCR2R3
RiCR2R3 + (CH3)3CH
                                 (CH3
                 0
RiCHR3 + NO
                     + HNO
                                             o:
                             (2)
                                              (3)
   Values chosen for ki were those obtained by
Batt et al. [6], and are in good agreement with
those obtained by Golden et al. [7], the value
of k2 chosen was that determined by Berces and
Trotman-Dickenson [8]; the values chosen for k3
were derived from disproportionation/combination
ratios and values of ki [6,9].  Other reported
data were not used because their reference reaction
rates are not sufficiently well known.

   The recalculated data are presented in figures
2 through 5, and the corresponding Arrhenius
parameters are presented in table 1.  The data for
t-BuO are the most extensive (fig. 2), covering
nearly four orders of magnitude.  The individual
sets of experimental data taken independently show
a rather wide range of Arrhenius parameters and
appear to be inconsistent, but taken together, the
actual data give a reasonably good straight line
with parameters, log k/s"1 = 15.1   16.2/0.
Given the entropy change of the reaction, AH2 =
41.2 Gibbs/mole, the A-factor for the reverse
reaction is A  = 107'9 M"1 s"1, a value very close
to that for the reaction of methyl radicals with
isobutene (log A = 8.0) [10].  This suggests a
self-consistent method for evaluating and codifying
the limited data available for the other alkoxyl
radical  reaction: choose an A-factor for the
reverse reaction and find the corresponding activa-
tion energy.  If this unified scheme is used, the
alkoxyl  decompositions can be considered together
as a class, rather than individually.

   The decomposition of an alkoxyl radical is the
reverse of the addition of an alkyl radical to
the carbon atom of a carbonyl group, which is
analogous to alkyl  radicals adding to the
2-position of a primary olefin.  Since data are
only available for alkyl radicals adding to the
1-position of primary olefins, the assumption was
                                                   53

-------
                                                           3.
Fig.  2.   t-BuO'  + Me + acetone,  y. ref. [11]; o,
         ref. [12]; A, ref. [13]; o, ref. [14];
         dashed line, ref. [6(c)j; solid line given
         by log k(s-!)   15.2   15.9/6.
       Table 1   Experimental values for RO' decomposition
              rates.9
                                                                                                      3.5
                                    EtO' " Me + CHO.  o, ref.  [15],  line  given
                                    by log k(s-1)   13.7   21.6/9.
                                                                    M
                           This equation predicts activation  energies with
                           an uncertainty of about  ±  0.5  kcal/mol.   It
                           predicts that the reverse  reaction has an activa-
                           tion energy given by
= E
                                         AH°
                                                                         + RT = 13.6
     Radical  log A  log E  AS2  log Ar log Aest
                0.29 AH°
                       K
(5)
                                          Ref.
EtO'
i-PrO'
s-BuO'
t-BuO'
13.7
16.1
16.4
14.9
15.1
22.1
20.6
18.0
15.3
16.2
33.4
37.8
37.7
41.2
8.2
8.2
8.0
8.0
13.7
14.6
14.4
15.2
22.1
17.4
13.9
14.2
16.3
[67]
[68,69]
[70]
[65d]
[71-74]
     Units:  E in kcal/mol; Ar in M-IS-I; A in s-l.

made that the A-factors for addition  to  both  ends
of an olefin double bond are the same  and  only
the activation energies differ.  Thus, A-factors
for analogous alkyl radical plus olefin  reactions
were chosen from the  tables of Kerr and  Parsonage
[10], corrected for any difference in  reaction
path degeneracy, and  applied to the alkoxyl
reactions.

   Assumed A-factors  for the reverse  reaction,
A , are summarized along with  ASn, log A  ,,  and
corresponding activation energies E in table  1.
A plot of E vs AHg is presented in figure  6 and
gives a good straTght line:
                           Although these equations  apply  to ^ 400 K where
                           most of the experiments were  carried out, the
                           estimated activation energies will  be negligibly
                           different at ^ 300 K.

                              Estimated decomposition  rate constants for a
                           number of alkoxyl radicals  at 300 K and atmo-
                           spheric pressure are presented  in table 2.   Fall-
                           off corrections were obtained bv use of the
                           Emanual RRK Integral  tables [3,19].   For the
                           experimental data available, the estimated  rates
                           are accurate to about a factor,  of two,  as' demo-
                           strated by comparing estimated  and  observed rate
                           constants (figs.  2-5).  The differences apparent
                           between the estimated and experimental  rate
                           constants are due to the ±0.5  kcal/mol
                           uncertainty in estimating the reaction  activation
                           energy and round-off errors on  the  A-factors.

                              Estimates made when no experimental  data are
                           available can be appraised by using the propaga-
                           tion of errors equation.  Since  A-factor and
                           activation energy are usually estimated independen-
                           tly, the uncertainty in log k can  be written
     E   12.8 + 0.71 AH°   AH°  >  0
                   (4)
         12.8
AH° £ 0  .
                                                            log k
log A
                                                    o|/e2
                                                                         (6)
                                                    54

-------
Fig.  4.
                          1000/T
i-PrO'  ^ Me + MeCHO.  o, ref. [16]; o,
ref.  [17]; line given by log Ks-1) =
14.6   17.8/6.
                                                                                1000/T
                                                      Fig. 5.
s-BuO'  5 Et + MeCHO.  o, ref. [18]; ref.
[6(d)]; line given by log k(s- ) = 14.4
14.6/6.
  25
  20
 ' 15
  10
                           EI400KI = 12.8 + 0.71  (AH|,)
    0   1   2   3  4   5   6   7   8,  9   10  11   12  13
                     AH" / kcal mol'1
                        R

 Fig.  6.   Correlation  between activation energy E
          and  enthalpy  of reaction AH^.

where a,    k,  a-,    „,  and ov are standard devia-
tions in  Tog k, Tog A, and activation energy
respectively,  and 9 =  2.303 RT.  Since log A is
probably  uncertain  by  ± 0.5, and the activation
energy is  uncertain by about ± 1 kcal/mol, the
value for  a,    k%  0.88 at  300 K, and k is
uncertain  by about a factor of eight.  This
represents a relatively favorable case for estima-
tions, since an adequate amount of kinetic data is
                                             available,  and  it  is  fairly  consistent.   For the
                                             reactions discussed below, very  little data is
                                             available,  and  the realiability  of the estimates
                                             is much  lower.

                                                Alkoxyl  Radical Reactions with  Oxygen.   The
                                             only  reliable Arrhenius  parameters [21] known for
                                             this  class  of reactions  are:
                                                           CH30 + 02 ->• CH30 + H02

                                                           log k/M'V1" 8.5   4.0/0  .
                                                                                           (7)
                                              Rates  for  other members  of this class can only
                                              be  estimated  after making assumptions regarding
                                              variations in A-factors  and activation energies.
                                              The concomitant uncertainties in estimating
                                              rate constants will  be relatively large since
                                              little is  known about such variations.

                                                 The A-factors for this group of reactions are
                                              expected to be similar to that for methoxy
                                              radicals,  aside from the reaction path degeneracy
                                              (n) factor; therefore, log A can be estimated as
                                              follows:
                                                   log  Aest/M"
               8.0 +  log  n
(8)
                                                 Estimates  for activation energy variations are
                                              rather problematical, especially when E  is low.
                                              Two alternative methods can be used:
                                                    55

-------
                           Radical"
                                      Table 2.  Estimated RO' decomposition rates.

                                        AHR    ASR  109 A/  log A(s-')  Eestc   k/kffi
                                                                              Mmin-1)
c-co-
cc-co-
6
CH!:C
ccc-co-
fl
Jc
c
c-co-
c
HOC-CC
HOCCC-CO-
6
HOCCC-COH
6
(HO)2CCC-COH
6
(HO)2CCC-C(OH)2
0
(HO)3CCC-C(OH}2
C
HOC-CO-
n
cc-cc
12.4
9.4

7.1
8.9
2.6

4.3

6.8
(3.5)f
8.7
-9.7

-9.6
-30.9
-24.5
11.4
(8.2)f
-5.1
33.4
35.0

37.8
36.3
37.7

41.2

38.0
36.6
39.2

39.2
37.7
37.7
37.7
40.3
8.2
8.0

8.2
7.5
8.0

8.0

8.0
7.1
7.1

6.8
6.8
6.5
7.5
7.5
13.7
13.8

14.6
13.6
14.4

15.2

14.5
13.3
13.8

13.5
13.2
12.9
13.9
14.5
21.6
19.5

17.8
19.1
14.6

15.9

17.6
(15.3)f
19.0
12.8

12.8
12.8
12.8
20.9 ,
(18.6)f
12.8
0.003
0.6

0.5
0.8
0.7

0.5

0.8
1
1

1
1
1
0.9
0.8
2.1 x
1 .7 ^

1 .6 x
2.9 x
2.9 x

1.5 x

2.8 x
(2.2 x
2.1 -
2.1 "

1.0 x
5.2 x
2.6 x
3.6 x
(2.2 x
8.2 x
10-3
101

102
10'
105

105

103
105)f
102
106

106
105
105
10° f
106
                                    IC-CC re
                          ^Notation:  HOC-CC represents HOCH2CHCH3 •» HOCH + HCCH3, etc.
                           A-factor for analogous alkyl radical + alkene association reaction.
                          y...t  12.8 + 0.71  HS (kcal/mol).
                          °FaTT-off estimated from RRK tables for 1 atm, 300 k.
                          j:Rate constants for 300 K and atm air.
                           Based on Group Additivity [2], not on experimental AHS for propane-1,
                           2-diol [20].
                                                                     Table 3.  Estimates:  RO- + 02 reactions
                                                                                                   III

                                                                                       - 10.6 + 0.25  E = 11.5 + 0.29
    (1)   Assume E  is constant for the entire
         homologous series.

    (2)   Assume that an  empirical relationship
         that holds for  other radical reactions
         applies to this series as well.

A simple empirical relationship [22] that  gives
Ea with  an  uncertainty  of  about ± 3 kcal mol  :
for exothermic H,  OH, and  CH3  reactions is  given
by equation  (9):

     E    11.5 + 0.25 (AHR),
      a                   K
where AHR is the enthalpy  of reaction.  For
reaction (7), equation  (9)  predicts E  = 5  kcal
mol"1, about 1 kcal  mol  '  too high. Equation  (9)
has two  different ways  to  give the proper  E
for reaction (1);

     E    10.5 + 0.25 (AHJ
      a                   K
     E   = 11.5 + 0.29 (AHR)   .
      a                                                     The  overall uncertainties for  this family of
    Rate  constants for a number of alkoxyl  radical     reactions may be estimated as before.  Log A is
lons Reaction log (A)est tf
Jl
(1
yes
given

(9)
kcal
»n (9)
'a

(10)
(11)

CH30 + 02 8.
EtO + Oz 8.
n-PrO + Oz 8.
i-PrO + 02 8
n-BuO + 02 8
s-BuO + 02 8

5
,3
.3
.0
.3
.0
Effective first-order


k
2.
1.
1.
4
(min
0 x
3 x
3 x
6.7 i
1.3 x
6.
rate

.0
"')
105
10s
10s
10"
10s
.7 X 10*
constants


X
K>
k (min'1)
2.0
8.2
8.2
1.5
3.5
1.1
at 300

x 10s
x 10s
x 10s
x 106
x 10s
x 106
K in air (2.1

x (AH°)
k (min"1)
2.0 x 10s
1.3 X 106
1.3 x 10s
3.7 x 10s
5.8 X 105
2.2 x 10'
x 10s pplti Oz).

reactions  were estimated  by  the three methods
and are  presented in table 3.   Considering  the
large uncertainty associated with equation  (2)
                                                         probably uncertain  by ± 0.5 and  the activation
                                                         energy  is probably  uncertain by  an  average of
                                                         ±1.5  kcal/mol.  Thus,  a,    ,  %  1.2,  and the
and  the  low activation  energies, the  estimates in     estimated rate constant isyuncertain  by about a
                                                         factor of 16.  If  the activation  energy is
                                                         uncertain by an average of % ± 2.5  kcal/mol, the
column  III  of the table are  highly uncertain  and
may well  be upper limits  to  the correct  rate
constants.   Similarly, rate  constants in  column I     estimated  rate constant  is  uncertain  by a factor
of the  table may be near  the lower limits.
                                                         of -v 80.
                                                       56

-------
   Alkoxyl Radical  Isomerization Reactions.   The
importance of alkoxyl  radical isomerization
reactions has been  inferred from smog chamber
data [23], as well  as  from more qualitative
consideration [24].   The estimation of  the  iso-
merization rates  is  relatively straightforward but
the estimates are somewhat uncertain, as  discussed
below.

   A-factors for  5-membered ring (5R) and
6-membered ring  (6R)  isomerizations were  estimated
to be lO11'^"* and  lO10'^"1 (per H-atom),
respectively.  The  estimate for the 5R  transition
state was made by noting that in tying  up the
methyl and ethyl  internal rotations, the  change
in entropy is about  -6.6 Gibbs/mole; subtracting
another 0.3 Gibbs/mole for the reaction coordinate,
we obtain log A™   11.7 for three abstractable
H-atoms.  Thus,  for each abstractable H-atom,
log A5R  (per-H)  = 11.2.

   For the 6R transition state, a model transition
state was used.   For the decomposition  of ethyl
vinyl ether  (EVE),  log A   11.4 at 700  K.  If log
A  is  about the  same at 300 K and SU(EVE)  =  82.6
Gibbs/mole [2],  then the entropy of the transition
state is  74.3 Gibbs/mole.  In comparing the EVE
transition state  and that of n-butoxyl  radical,
EVE has  some  doube bond character, and  that of
n-butoxyl will  be looser by  about 0.6 Gibbs/mole.
n-Butoxyl has one more hydrogen atom, worth about
0.2 Gibbs/mole,  and has spin, contributing 1.4
Gibbs/mole.   Adding all of these corrections
gives S°+ =  76.5 and log A   11.4 for three
abstractable  H-atoms; thus log AgR  (per H)    10.9.
The uncertainties in these estimates  are  probably
±  4 Gibbs/mole  and log A is  uncertain by  ± 1.

    Activation energies may be estimated from the
activation energies for H-abstraction by  alkoxyl
radicals  [25]  by adding a "strain" energy of 0.5
kcal/mol  for 6R reactions and 5.9 kcal/mol  for
5R reactions  [2].  These activation energies are
rather  uncertain, probably ± 2 kcal/mol.   The
combination  of  the two sources of error by the
propagation  of  errors formula gives  an  estimated
uncertainty  in  log k of ± 1.8 at room temperature.
Thus, the rates  are estimated to be  uncertain by
about a  factor  of 60.

    The method for estimating these  rates  is
summarized  in  table 4 and estimated  rates for
several  alkoxyl  radicals are presented  in table
5.  All  reactions are assumed to be  at  the high-
pressure limit.

3.  What  do  we  need to know?

    First  of  all,  we need to  test some of the
preceding ideas  with experiments conceived for
just  that purpose.   Since the ideas are based on
the transition  state theory  formalism,  it is
important to  address the question of  limits of
validity  of  transition state theory.   In  general,
these testing  reactions should be measured under
conditions where isolated reactions can be
observed, as  the extraction  of individual rate
constants from  complex reacting systems is fraught
with  difficulty.

    Preliminary  results from  our own  laboratory
indicate  that the isomerization reaction  of
        Table 4. RO-  isomerization reactions—estimation
               procedure.

        E = E (abstraction) + E (strain)


        Hydrogen Abstracted    E (abstraction), kcal/mol

         RCHz-H                    7.2

         RCH(OH)-H                  6.0

         RiR2CH-H                  4,1

         RiR2RsC-H                  4.1

         RC(OH)2-H                  4.1a
        Strain Energy

         5-membered ring

         6-membered ring
                          5.9 kcal/mol

                          C.5 kcal/mol
        A-Factor (per abstractable H)

         5-membered ring           A = 1011'2 s"1

         6-membered ring           A = lO10'9 s"1


        ""Estimated.




       Table 5.  Estimated RO- isomerization reaction rates.
        Reaction
                         log A(s"')  E(kca1/mo1)   k(min'')
occcc
6
cccc
HOCCCCO
* HOCCCC
OH
,cccc-
» HOCCCCOH
11.4

11.7
11.2
7.7

13.1
6.5
3.7

S.6
1.9
x 10'

x 10'
X 10'
 0

HOCCCCOH * (HO)2CCCCOH     11.2
         0-

   (HOhCCCCOH * (HO)2CCCC(OH)2   10.9
 (HO)2CCCC(OH)2 - (HO)3CCCC(OH)2   10.9
        OH     OH

       CCCCO- » CCCCOH
                           11.4
                                     6.5
                                     4.6
                                     4.6
                                     7.7
                                             1.9 X
                                       2.2 X 10'
                                             2.2 x 109
                                             3.0 x 10"
           OH     OH             OH          OH
 "Notation:  CCCCO- -» CCCCOH represents CH3CHCH2CH20- * CHZCHCH2CHZOH

primary alkoxyl  radicals predicted  by Barker
et al. [5],  does indeed take place  at the rate
expected.   In  this experiment, nbuO' radicals
were generated from the VLPP t>f  nbnONO  in the
presence of  DI.   The mass spectrum  of products
indicated  the  production of both nbuOD  and
DCH2CH2CH2OH.

   There are many  examples of reactions  for which
rate constants  have been studied, but product
studies are  lacking.  Thus, in reactions  of OH
with olefins current models must arbitrarily
decide on  branching ratios.  This is equally true
in aromatic  systems.

   In all  of the above discussion of estimation
of rate data,  the  importance of  thermochemiqal
values for all  species has been  emphasized.   It
is particularly important to have a good  set of
values for the  entropy and heat  of  formation
of organic free radicals.
                                                     57

-------
    Very few spectroscopic assignments  exist for
modest-to-large size organic free  radicals and
entropies  (and heat capacities) have generally
been  estimated by methods discussed earlier.
Uncertainties arise from changes in hindered
rotation barriers and changes  in skeletal  bending
frequencies.

    Heats of formation of organic free  radicals
have  been  measured by a variety of techniques.
All have been extensively presented in  the
literature.   Common techniques and problems
associated with them are:

1.  Bond scission activation energy assigned  as
    bond strength:
          R-R1
           R + R'
     (b)

     (c)
     (d)
EI is a function of T.

k_  is a function of T.

Small fractional errors  in  slope  of
Arrhenius line lead to large  absolute
errors in EI.  (This problem  is overcome
to a large extent by use of relative rate
techniques).

ki and k   are functions of P.
2.  Mass  spectrometric techniques requiring cycles
involving ion thermochemistry:
R ->• R+ + e
S * R+ + e
AP-IP AH° Q(R) AH°

IP
AP
,0(S>
     (a)   Difficulties  measuring IP and AP (a
          whole  literature)
                    R
                           EA


                           PK
                                           AH
          R + e"


              RH ^ R" +  H"1


          AHa   EA   AH°(H+)    AH°(RH)  + AH°(R)

    (a)  Difficulties measuring EA and  pK .
                                          a

3.   Halogenation kinetics  (principally  iodination):

    Measurement of the rate constant  k!  for
          RH +  I
                         -> R + HI
and the assumption that  E    =  1  ± 1  kcal  mol"1

    (a)  Validity of  the assumption

   Recently several workers  have suggested that
the value of AHf(R) determined by iodination
techniques are two low.  This  says,  in effect,
that if ki is correctly  measured E   must be lower
than 1 ± 1 kcal/mol"1.             ''
                                                T 9.50
                                                 " 9.30
                                                !_ 9.10
                                                ^ 8.90
                                                £8.70
                                                 S 8.50
                                                                      n	1	1	r
                                                                              CH2D
                                                                             Oj + i'
                                                                        CH2-CH2
                                                      log(k3/M-' s-'l  (9.93 ± 0.22)- 4°303* RT—-T = 965 K
      8.50 9.00  9.50 10.00  10.50 11.00  11.50 12.00  12.50
                         10" K/T

Fig. 7.  Arrhenius plot,   "bibenzyl  as precursor;
         D benzylvinylether  as  precursor for
         benzyl radicals.
                                                             9.20

                                                             9.10

                                                          7" 9.00

                                                          J" 8.90
                                                          5 8.80

                                                          21 8.70

                                                          f 8.60

                                                             8.50

                                                             8.40
                                                                 log(k3/M-1 s-1) = (9.73 ±0.211 -
                                                                              4000 ± 1000
                                                                               2.303 RT
                                                                                                 T = 1000K
                                                       log(k3/M-' s-M = (9.58 ±0.35) - 3°°° 3* R1°°° T = 635 K
        I    10   11    12   13   14   15   16
                     104 K/T


         Arrhenius plot.   --- methathesis reaction
         involving DI; — metathesis reaction
         involving HI.
                                                 We have  recently tested this possibility with
                                              several experiments where R is allyl or benzyl
                                              radical.  Figs.  7  and 8 show the results.  These
                                              Arrhenius parameters are shown to be compatible
                                              when a suitable  transition state model is chosen
                                              with values  of AHf 298 (ally!) = 39.1 ± 1.0  (corre-
                                              sponding to  E.i  -2.3 kcal mol'1) and  AHf)298
                                              (benzyl)   46.6  ±  1.5 (corresponding to E_ ' =
                                              2.5 kcal mol"1)-   Thus we see no indication of
                                              any major problem  with the halogenation technique.
                                              Work on t-butyl  radical  is in progress.

                                                 There are other specific elementary processes
                                              for which the  rate constants (and products) require
                                              study.  It  is  not  my intention to try to develop
                                              a list in this  paper, but this workshop as  a  whole
                                              might consider doing so.
                                                     58

-------
4.   How Accurately (do we need to know whatever
    it is that we need to know?

   This question cannot be answered in a general
way.  It is optimistic to think that we can know
thermochemical values for the free radicals to
better than ± 1 kcal  mol"1 in AHf and ± 1 cal
mol"1 deg"1 for S°.   Errors of tnis size will
limit possible accuracy of estimates of rate
constants to an order of magnitude.  If the data
being modeled justify higher accuracy, individual
rate constants will  have to be measured.  In
general, rate constants can be expected to be
measured to accuracies on the order of a factor
of two but some are hard to obtain at all.

   As far as  I can tell, the limiting problem in
current smog modeling is as much with the data
to be modeled as the kinetic and/or thermochemical
data.

5.  With respect to this particular topic what do
    you see as the research priorities?

    I have really addressed these previously, but
to  summarize:

    General
   Tests of predictions using thermochemical
   kinetics

   Thermochemistry of free radicals

   Product studies

   Reactions in aromatic systems

   Specific (incomplete)

   Branching ratios in OH reactions with
   unsaturates

   Heats of formation of prototype alky!  radicals

   This Workshop should develop a list  like this
 combining all topics to eliminate overlap.

 6.  Are there speculative problem areas that
    should be given some attention?

   In the general area of organic free  radical
 kinetics, the question of perturbation  of
 "elementary" reactions by the formation of weakly
 bound complexes is one that  I find perturbing.
 Thus, if:
    A + BC *•  A-BC -> products

 and   k    >  k2
   kexp't  = Kikz

 This allows for low and/or negative activation
 energies, since E   .    AEi + R2 and AE: < 0. If
 E   > E2, A_  > Afx|ttr k_  > k2, thus:
   If this type of mechanism is common, it must
be taken into account and what will be needed  is
a method for recognizing and predicting such
occurrences.

   Acknowledgment.  I have profited greatly from
the work of my colleagues, Dr. John R. Barker
and Dr. Alan C. Baldwin.  Conversations with
these coworkers and Dr. Dale G. Hendry have been
very valuable.

   Support from SRI International with some help
from the Environmental Protection Agency (Contract
No. 68-02-2427) is gratefully pointed out.

   Some parts of this paper are taken from
previous reports and publications.


References
           < A
[1]  Benson, S. W., Cruickshank, F. R.,
     Golden, D. M., Haugen, G. R., O'Neal,
     H. E., Rodgers, A. S., Shaw, R.,  and
     Walsh, R., Chem. Rev. 69, 279 (1969).

[2]  Benson, S. W., Thermochemical Kinetics,
     2nd Ed. (John Wiley and Sons, Inc.,
     New York, 1976).

[3]  Golden, D. M., Solly, R. K., and  Benson,
     S. W., J. Phys. Chem. 75_, 1333 (1971).

[4]  Smith, G. P. and Golden, D.  M.,  Int._
     J. Chem. Kinetics (to be published).

[5]  Barker, J. R., Benson, S. W., Mendenhall,
     G. D., and Golden, D. M., EPA-600/3-77-10,
     Grant No. R802288, October  1977.

[6]  (a)  Batt, L., McCulloch, R. D.,  and
     Milne, R. T., Int. J. Chem. Kinetics
     6, 945 (1974).

     (b)  Batt, L., McCulloch, R. D.,  and
     Milne, R. T., Int. J. Chem. Kinetics
     Symposium No. 1, 441  (1975).

     (c)  Batt, L. and Milne, R. T. Int. J.
     Chem. Kinetics 8, 59  (1976).

     (d)  Batt, L. and McCulloch, R. D.,
     Int. J. Chem. Kinetics 8, 911 (1976).

[7]  Mendenhall,  G. D., Golden,  D. M.,  and
     Benson, S. W., Int. J. Chem. Kinetics
     I, 725 (1975).

[8]  Berces, T. and Trotman-Dickenson,  A.  F.,
     J. Chem. Soc.  83, 348 (1961).

[9]  Heicklen, J.  and  Johnston,  H. S.,
     J. Amer. Chem. Soc. 84. 4030  (1962).

[10] Kerr, J. A.  and Parsonage,  M. J.,
     Evaluated Kinetic Data on Gas-Phase
     Addition Reactions  (CRC  Press, Cleveland,
     Ohio, 1972).

[11] Cadman, P.,  Trotman-Dickenson, A.  F.,
     and White, A. J., J.  Chem.  Soc.  (A),
     2296  (1971).
                                                    59

-------
                                                  88B,  88(B),

                                                  89A.  89(a),
[12] Bires, F.  W. ,  Danby, C. J., and Hinshel-
     wood, C.  M. ,  Proc.  Roy. Soc.  (London)
     A239, 154
[13] Quee, N.  J.  and Thynne, J. C. J.,
     Trans.  Faraday Soc.  63. 2970 (1967).

[14] McMillan, G.  R., J.  Amer.  Chem.  Soc.  82,
     2422 (1960).

[15] Leggett,  C.  and Thynne, J. C. J., J_._
     Chem. Soc.  (A), 1188 (1970).

[16] Ferguson, J.  M. and Phillips, L.,
     J. Chem.  Soc.  87. 4416 (1965).

[17] Cox, 0. L.,  Livermore, R.  A., and
     Phillips, L., J. Chem. Soc.  (B),
     245  (1966).

[18] East R. L.  and Phillips, L., J.  Chem.
     Soc^lAl, 1939 (1967).

[19] Emanuel ,  G., Aerospace Report No. TR-
     0200(4240-20)-5.

[20] Stull, D. R. , Westrum, E.  F. , Jr., and
     Sinke, G. C., The Chemical Thermodynamics
     of Organic Compounds (John Wiley and
     Sons, Inc.,  New York, 1969); Thermo-
     chemistry of Organic and Orqanometallic
     Compounds (Academic Press,  Inc., New York,
     1970).

[21] Barker, J.  R. , Benson, S.  W., and Golden,
     D. M., Int.  J. Chem. Kinetics 9, 31
     (1977).

[22] "Semenov Rule," see Laidler, K.  J. ,
     Chemical  Kinetics, p. 132 (McGraw-Hill, Inc.
     New  York, 1965).

[23] Carter. W.  P. L., Darnall  ,  K. R., Lloyd,
     A. C., Winer, A. M. , and Pitts,  J. N.,
     Jr., Chem.  Phys. Letters 42, 22  (1976).

[24] Whitten,  G.  Z. and Hugo, H.  H. ,  SAI
     Report EF76-126, Draft Final Report
     (1976).

[25] Benson, S.  W. O'Neal, H. E., Kinetics
     Data on Gas-Phase Unimolecular Reactions,
     NSRDS-NBS 21 (U. S. Government Printing
     Office, Washington, D.C., 1970).
                     Appendix
      Application of Thermochemical Kinetics
       to the Analysis of Some Recent Data


    Niki et al.  (J. Phys. Chem. 82, 135 (1978))
 present data  involving the analysis of a complex
 mechanism which  leads them to the  conclusion that
 in  the HO radical-initiated oxidation of the
 ethylene-NO system the radical HOCH2CH26 cleaves
 in  preference to  reacting with tropospheric 02
 concentrations.   This conclusion is difficult to
 justify, viz:
     HOCH2CH26 ?  CH2OH + CH20
              -i


AH°   -42 ± 2   -6.1 ± 2  -26.0

S°     76.4      59.0      53.3

Cp     21.7
                                       AH°   9.9 ± 3

                                       ASS   34.9
                                         K
 Ej      12.8 + 0.71 (9.9 + 3)   19.8 ± 2


 For  k_  ,  similar  to  C2H5  +  C2Hi,  and  log  A     8.0

 .MogAi   8.0  t-  (34.9    8.35)/4.58    13.8

  log kj    13.8   19.8/0
                                     C     8
 RRK  correction  for fall-off   s    —2-=	  = 7


   -jp—  ^ 0.75  1 atm,  300 K
     Kco

 .'. ki(l  atm, 300  K)   0.2 s"1, within  a  factor of
 50.


   HOCH2CH26 +  02 *  HOCH2CH + H02


   AH°     -42   0       -75     5       AH^   -28

 log  A2      8.3

 Three methods for estimating  E2:

 (a)  E2a   4.0

 (b)  E2b   10.5 + 0.25(AH°)    3.5

 (c)  E2(.   11.5 + 0.29(AH°)   4.0


 Three estimates for  k  :

 (a)  k2a   2.4  x  10  M'V1

 (b)  k2b   5.8  x  10  M^s"1
(c)  k2
           7.0 x  10
For a 20 percent mixture of 02 at 700 torr, the
effective first-order rate constants are:

(a)  ki    1.8 x 103
                                                                               average   3.8 x 103s~'
(b)  kib   4.3 x lO's"1

(c)  k2(.   5.2 x lO's"1

Thus, k1  (1 atm, 300 K) £ 0.2 s"1 and k2  (300  K)
^ 4 x 10  s-1 indicate that the decomposition path-
way 1S expected to be totally negligible, if our
estimation techniques are okay.   For kj, there is
an uncertainty of a factor of 50, and that for k,
is probably a factor of 20.   Thus, the ratio of
k2/kj  can range as shown:
                                                   60

-------
            1o9 l-r-J = 4.3 ± 3
   Another interesting question is raised by the
work of Herron and Huie (J.  Amer.  Chem. Soc. 9£,
5430 (1977)).   They found from a study of ozone-
alkene reactions that they could explain their
data best by invoking the production of vibration-
ally excited formic acid which decomposes by
several  steps, the principal one of which is
     HCOOH
 CO + H20
   If the pathway for creating of HCOOH involves
the rearrangement of the intermediate CH200 via
   "CH200
   0
   /\
H2C-0
"OCH20"
we can calculate that the internal  energy in
formic acid must be ca.  150 kcal  mol"1.
[•CH200-]  is the least stable of the above species
(AHf ^ 48  kcal  mol"1 calculated from BDE(H-CH200-)
= 93 kcal  mol"1.   Since the ring closing this
species undergoes has about 10 kcal  mol  1 activa-
tion energy, the excitation relative to  HCOOH
(AHf    90.5 kcal  mol"1) is -v 150  kcal  mol"1].

   It is possible to use quantum RRK theory to
estimate the rate constant for all  pathways
providing  that  we can write the Arrhenius
parameters:
  k ^ A
           n.'  (n-m + s-1]
          (n-tn):  (n  + s  -1)!
n   E/hv;  m   Ea  t/hv;  s  = number of oscillators
for HCOOH; s  = §;  and v (geometric mean) =
1345 cm"1.

    These  calculations  favor  the  production  of OH
 and HCO radicals, but  all  the  rate  constants  are
 so fast that  this simple  calculation  may  not  be
             AH/kcal mol"1  E /kcal mol"1  log A/s"1 logtk/s"1]
                         a
 HCOOH •* H2 + C02


     •» H20 + CO
      HCO + H
         0
         II
      H + COM

       S
      HC + OH
    3.6


    6.3



   106.6


    92.6


   108.9
     > 50


     > 30



      107


       93


      109
13


12



14


14


16
i 11.7


' 10.3



 10.3


 11.1


 12.3
able to discriminate properly between the first
and last reaction above.   However, similar cal-
culations for larger species do not have this
difficulty.   Furthermore, in the case of larger
species deactivation by collision must be taken
into account.
               Summary of Session

   Parkes opened the discussion by re-iterating
Golden's contention that there is no  use trying  to
understand the reactions of one radical in  a  class
in isolation from the other members of that class.

   Batt followed with extensive comments on his
work on alkoxy radical reactions  (see contributed
comments).  In response to a question by Benson  as
to whether the production of hot  radicals by
photolysis could explain the various  divergent
results, Batt replied that his experiments  were
carried out in an atmosphere Of CFc, using 366 nm
radiation for the photolysis, so  that any excess
energy in the radicals should be  quenced rapidly.
Cox pointed out that they had done some photolysis
experiments using methyl nitrate  at low concentra-
tion ("c 10 ppm) and get similar results to  those
of Batt and those of Heicklen, both carried out
at higher concentrations.  This suggests that hot
radical production in the photolysis  is not the
reason for the discrepancy.

   Ravishankara stated that when methyl nitrate  is
photolyzed in the banded region,  the excess energy
is predicted to go to NO, whereas in the
continuum, CH30 comes off with excess energy and
NO is in the ground state.

   In discussion on the rate constant for the
reaction CH30 + 02 -» CH20 + H02,  it was apparent
that not only was the value not completely  agreed
upon, even the extent of the discrepancy was
subject to some controversy.  Much of this
appeared to be due to different conceptions of
what is reasonable agreement.  Cox felt that,
compared to other reactions of this type, it is
known rather badly.  There is a discrepancy of at
least a factor of three in the room temperature
value.  Batt is reporting 109'51 mol'-'s 1 for the
A factor, while Golden spoke of 108'5, which
appeared to agree with the other  data.  Batt
pointed out that, due to the pressure dependence
of CH30 + NO, Heicklen's value at room temperature
could be high, bringing the A factors into  better
agreement.

   Tsang pointed out that, for a  given class of
reactions, the pre-exponential factor for
decomposition is a constant, which is a valid base
point for comparison.

   The discussion of alkyl peroxy radicals  was
initiated by Benson (see contributed  discussion)
who proposed a complete new mechanism for their
self reaction.  Briefly, for alkyl peroxy radicals
with an a-hydrogen, the reaction  proceeds through
a radical disproportionation to produce a carbene
peroxide radical and a hydroperoxide: 2RCH202 •*
RCH202H + RCHOO.  Heicklen asked  how  peroxides
(ROOR), which he has observed in  these systems,
are formed.  Huie asked if secondary  ozonides had
been observed in any of these systems.  Benson
replied that the reaction of an aldehyde and  the
carbene peroxy radical to form a  secondary  ozonide
is slow, so it might not compete.  Huie questioned
this, pointing out that work from both his
laboratory and Niki's showed secondary ozonide
formation at low aldehyde concentrations.
                                                    61

-------
   The focus of the discussion was shifted to H02,
as the simplest peroxy radical.  Golden, noted
that two unpublished papers suggest that its heat
of formation is 5 kcal/mol lower than presently
accepted.

   Calvert brought up the question of hydration
with reference to H02 reactions.  Recent work on
the recombination of H02 suggested that H02 was
hydrated.  Calvert asked if other peroxy radicals
could be hydrated, or even if other classes of
radicals, like alkoxy and hydroxyl, could also be
hydrated.

   Cox (see contributed comments) discussed his
work on H02 recombination, which suggests the
involvement of an H20i, intermediate.  His results
are consistent with hydration of HOZ.

   Tsang asked how much peroxy recombination
reactions affected photochemical smog.  Apparently,
they are not important since as Calvert pointed
out, in an urban environment, NO was always being
pumped in.  The importance of these reactions in
the clean troposphere was noted by Warneck.

   Tsang discussed the discrepancy in alkyl
radical studies (see contributed remarks).   In
addition, he emphasized that the systematic
approach to radical  chemistry outlined by Golden
is the only way to solve these problems.

   Basco presented results on ethyl  radical re-
combination by flash photolysis, which is in
agreement with the value reported by Parkes.   Also,
rate constants measured for the methyl peroxy
and ethyl peroxy radicals agree with the  values
of Parkes.  Calvert noted that this rate  constants
also agreed with those of Parkes and Hochanadel,
but the extinction coefficients appear to differ,
which suggest the agreement in rate constants may
benfortuitious.

   Basco and Parkes  both stated that they have
obtained a spectrum for the acetyl  radical, which
should allow them to carry out kinetic measure-
ments.

   In closing, Tsang asked if spectroscopic
techniques might be  useful in studying the  proper-
ties of large organic radicals.   It was agreed
that uv spectroscopy would not separate these
species, but infrared spectroscopy looked
promising.


                    Comments

L. Batt, Chemistry Department, University of
Aberdeen, Aberdeen,  Scotland  AB9 2UE

   Our work on the decomposition of  alkyl nitrites
(RONO)  has resulted  in  the determination  of values
for k2 and k6  (tables  1  and  2  respectively):
                                      RO + NO
                                                    RO „ + HNO
                                                      -M
                                                                          (6)
      RONO

      RO + NO

      RO

      RO + 02

      RONO
RO + NO

RONO

Products
RO

RO
_H

_H
       H02

       HNO
(1)

(2)

(3)

(4)

(5)
                                These studies show that reaction (6) accounts for
                                the production of nitroxyl  (HNO) rather than
                                reaction (5).

                                     Table 1.  Rate constants for the reaction
                                               RO + NO * RONO (2).
R
Me
Et
i Pr
s Bu
t Bu
log k2
(M"1 s"1)
10.1 ± 0.6
10.3 ± 0.4
10.5 ± 0.4
10.4 ± 0.4
10.5 ± 0.2
                                     Table 2.  Rate constants for the reaction
                                               RO + NO -<- RO M + HNO (6).
                                                           -n
R
Me
Et
i Pr
s Bu
log k6
(M"1 s"1)
9.3 ± 0.6
9.8 ± 0.4
9.8 ± 0.4
9.8 ± 0.4
                                  Table 3.   Arrhenius parameter for reaction (3).
Reaction
t-AmO ->
s-BuO -*
t-BuO ->
i-Pro ->•
EtO ^
MeO +

M2Ka ^
ACHb ^
MzKan
ACHb ^
CH20 J
CH20 H

- Et
- Et
H Me
- Et
H Me
H H
log A
± 0
14
14
15
14
15
14
(s'1)
.5
.8
.9
.5
.6
.0
.2
E(kcal mo!"1)
± 1
13.
15.
17.
17.
19.
27.
8
3
0
2
8
5
  ,M2K   acetone.
  DACH   acetaldehyde.

   Values of k3 have been obtained by allowing
reactions (2) and (3) to compete (table 3) except
for R   Me.   Here the value of k3 has been
obtained via a thermochemical kinetic argument.
These values may be contrasted with Golden's
estimated values.  Figure 1 shows the first,
proper and unequivocal evidence for the pressure
dependence of k3(t-BuO).  (This means that table 3
represents limiting values).  Table 4 indicates
the variation of k3 at 160 °C as a function of
pressure and shows that k3 is within a factor of
2 of its limiting value at a pressure of 1
atmosphere of carbon tetrafluoride.  Suitable
                                                   62

-------
    6.2
    6.0
  S? 5.8
    5.6
                           I
                200        400        600

                    CF4 PRESSURE (Torr)
                                               800
    Fig.  1.   Pressure dependence of k3 (130 °C).


Table 4.   Arrhenius parameters for k3(t-BuO) as
          a function of CFi, pressure.
log A3
11.45
14.7
15.5
E3
(kcal mol"1)
11.4
15.5
17.0
log
5.7
6.8
7.0
k3
no CFi,
1 atm CHi,
P -+ 00
conclusions may  be  drawn for smaller and larger
alkoxyl  radicals!

   Using essentially  Group Aojditivity Rules,  we
have calculated  values  of AH3 for alkoxyl  radicals
that appeared  in Niki's schemes.   We used  figure 2
to determined  values  of E3 (table 5).  By  a consid-
eration  of the reverse  step we calculate that A3
                              "
for these  radicals  is
a similar  conclusion.
10
                                   Golden came to
                      AH°Kcal mol'1

   Fig. 2.  E3 values determined from table 5.
                                                         Table 5.  Thermochemistry for some  alkoxyl
                                                                   radicals.
                                                                                                    AH?
                                                                                            (kcal mol"1)
                                 OH

                                 CH2CH20  ->- CH2OH + CH20

                                   39.9       -4         -27.7     18        8.2


                                    OH  CH3

                                 CH3CH  CHO -> CH3CH OH + CH3CHO

                                   57.7       -13.3      -39.7     16        4.7


                                 OH

                                 CH20  -c CH20 + HO

                                   31.8       -27.7        9.4     28.5    23.5


                                 OH

                                 CH20 + 02 •*  HCOOH + H02

                                   31.8       -90.5        5.0       1    -53.7
                                                        Our  values  for  k2  have  allowed  us  to determined
                                                     values  for  k2'  (table 6):
                                                            RO +  N02
                                                    RONO
                              We allowed reactions (21)  and (4)  to compete in
                              order to  determine a value for ki,  where R   Me.

                                  Table 6.   Rate constants for the reaction
                                            RO + N02 H- RQN02 (21).
R
log k2,

(M'1 s'1)
                                                          Me

                                                          Et

                                                          t-Bu
                                                             9.8 ± 0.4

                                                             9.9 + 0.4

                                                            10.2 ± 0.4
                                                     By using dimethyl  peroxide  as  a  thermal  source and
                                                     methyl  nitrite  as  a  photochemical  source of
                                                     methoxyl radicals, we  were  able  to cover a tempera-
                                                     ture range  from 200  °C to  (Scottish)  room tempera-
                                                     ture.   All  methoxyl  radical  sources now give essen-
                                                     tially  the  same result that ki,   109'5  • 10"5/6
                                                     M^s"1.  It is  too premature to  qualify these re-
                                                     sults with  error limits.  This result may also be
                                                     contrasted  to that of  Golden's.   One other result
                                                     from this study is the ratio k6'/k2'  <.0.l:
                                                          MeO + N02
                                                     CH20 + HONO
                                                                                               (6-:
                                 We  conclude that for the alkoxyl  radicals that
                              occur  in  Niki's scheme, decomposition will compete
                              with difficulty if at all  with their reaction with
                              oxygen at room temperature.  In the series R = Me,
                                                  63

-------
Et, i-Pr, and t-Bu where we consider the two
possible competing reactions of decomposition and
reaction with oxygen at room temperatures:

   (a)  MeO reacts exclusively with oxygen

   (b)  t-BuO decomposes exclusively

   (c)  EtO and i-Pro react via both paths.
W. Tsang, Center for Thermodynamics and Molecular
Science, National  Bureau of Standards, Washington,
D.C.   20234

   Dr. Golden has demonstrated the important role
that the thermochemistry unstable organic inter-
mediate can play in providing a basis for the
estimation of the rates of a variety of reactions
of importance in air pollution.  It is necessary
to point out that for proper application under
ambient conditions highly accurate enthalpies are
necessary.   A change of 1.4 kcals in activation
energy is equivalent to a factor of 10 in rate
constant.  In this respect some of our recent
results are especially disturbing (W. Tsang,
Int'l. Journal of Chemical  Kinetics, JO., 821
(1978)).  These investigations demonstrated the
concordance of all existing results on the
symmetrical decomposition and combination of the

alkanes: n-butane * 2 ethyl, 2,3-dimethylbutane J

2 isopropyl and hexamethylethane *- 2 t-butyl,
over a temperature range of 350-1200 K.  Unfortuna-
tely,  this very satisfactory agreement between
four completely disparate type of experiments
(comparative rate single pulse shock tube, radical
buffer, very low pressure pyrolysis and modulation
spectroscopy) also leads to the conclusion that
the commonly accepted heats of formation of ethyl,
isopropyl and t-butyl radical are'10, 10 and 20
kJ higher than currently accepted numbers from
metathesis reactions.  The implication of these
results is that if there are serious questions for
such simple radical species, then, what degree
of confidence can one have for more complex
systems?  Thus the utilization of estimation
schemes may be badly flawed by the uncertainty in
the data base.
Sidney W. Benson, Chemistry Department, University
of Southern California, Los Angeles, CA  90007

   Although bimolecular reactions of the two R02'
radicals do not  seem to be of importance in the
modeling of the  tropospheric photochemistry, their
self-reactions provide an important clue about an
intermediate which  is assuming more and more	
importance, namely  the carbene dioxide R'R"COO
more commonly referred to as the Criegee
zwitterion.

   During the past  two years, under Army sponsor-
ship, we have been  reexploring in our laboratories
at USC the reactions of interest in ignition and
combustion.  In  these systems the self-reactions
of R02' radicals assumes great importance.   It has
been known for some time that when R represents a
tertiary carbon  grouping such as Me3C or F3C,
these self-reactions follow the path:
                2R02
                  2RO  +  02
    this overall  reaction proceeds through the
 formation of a weakly bonded tetraoxide:
 (AHr ^ 9 kcai;
                2R02 Z ROOOOR + 9 kcal
 RzO,, is stable only at T < -70 °C but at least two
 groups of workers have studied the reversible
 equilibrium in solution below 200 K and made
 measurements of AH and AS.  For R02 radicals where
 R does not contain a tertiary C atom but has
 instead a-H atoms such as R'CH200" or R'R"CHOO',
 the equilibrium has not been observed but only a
 very rapid reaction leading to termination.  The
 major products observed for such terminations are
 the conjugate alcohol-ketone (or aldehyde) pair.
 6. Russell some time ago proposed the following
 mechanism for it:
                                    H   0
                +              +   / \ / \
      2R'CH202' ? R'CH.O.CHzR1 <-  R-C   0
                                      H.   0

                                       ''0

                                       CH2R'
                                  .	I
     R'CHO + 02  + HOCH2R'
   This is a 1,5 H transfer reaction which has
never been observed in saturated molecules but only
in molecules containing at least one multiple (pi)
bond.  There are many objections to it and in
previous publications I have discussed some of
them.  Its longevity may be attributed chiefly to
its apparent ability to predict the major termina-
tion products.

   In our recent reexamination of these and related
ozone systems, my colleague Dr. P. Nangia and I
have come to the conclusion that this reaction
proceeds instead through an atypical radical
disproportionation to produce the Criegee
zwitterion and a hydroperoxide:
2R'CH202
                 R'CH202H + R'CHOO
We estimate that AH2 % -26 kcal/mol with an activa-
tion energy of about 1 to 2 kcal/mol.  This
estimate is in excellent agreement with the ab
initio calculations recently made by Goddard et
al. on the stability of the zwitterion.  The
reaction

          R'CHOO ->• RCHO + 0(3P)

is endothermic by only 6 to 8  kcal but must have
an appreciable barrier of perhaps 15 kcal  in
excess of this because it is a spin forbidden
process.  Thus the zwitterion may have a reasonable
life time at ambient temperatures in excess of
1 second.

   The zwitterion is expected  to react relatively
rapidly with aldehyde to form  secondary ozonides
and we estimate the mechanism to be a concerted
one with a low activation energy.
                                                   64

-------
                                  ,0-0
R'CHOO + RCHO -*• R1 - C
                    /
                  H
             0
                                         C - R
                                          \
                                           H
    This ozonide will slowly decompose into an
 aldehyde R'CHO or RCHO and the isomeric, biradical
 R'CH02, which can rapidly and exothermically
 isomerize to the carboxylic acid.
                    o-
                    0-
                             R'C
          .0
          0 - H
    It  is our feeling that this  is  the mechanism
 for formic acid production which has been observed
 in both smog chamber experiments as well as  in the
 ambient atmosphere during smog  periods.  There is
 no kinetically acceptable way in which  the
 precursor HCO radicals which have  been  supposed  to
 be the source of  formic acid in the ambient
 atmosphere can do anything but  produce  peroxy-
 formic acid and that only slowly relative to
 H02 +  CO production.  HC03H on  decomposition will
 not produce HCOOH.
    We also  estimate  that  RCHOO  can  donate  its
 weakly-bonded  oxygen atom to  many  species  such  as
 NO, N02,  RO, R02,  S02,  and possibly S03.   It can
 also in  principle  react with  02 to  form 03 and
 RCHO.  However,  this reaction is spin  forbidden
 and may  have an  appreciable activation energy.
 Despite  this it  may  play  a role in  03  production
 in a number of oxidation  experiments which have
 been reported  some time ago.

    The only pathways suggested  so  far  for  the
 zwitterions in smog  episodes  is from the relative-
 ly slow  secondary  reactions of  03  with olefins.
 However,  it does not seem unreasonable to  suppose
 that the exothermic  reactions of a-H containing
 peroxy radicals  with N02  or with N03 may also
 occur:
      R'CH,0,
                  N02
                  NO 3
R'CH02 +
                                 HONO + 14  kcal
                                 HON02 +  41  kcal
 In the later portions of the smog reaction when NO
 has decreased and N02 and 03 have begun to peak,
 such reactions may become significant.  In this
 period of the overall reaction H02 reactions with
 R02' can also produce zwitterion as well as the
 more familiar ROOH.

   It is our intention to publish all of these
 considerations and their related antecedents in a
 forthcoming publication now in preparation.
 Richard A. Cox, U.K.A.E.A., Environmental and
 Medical Sciences Division, A.E.R.E., Harwell,
 Oxfordshire 0X12 ORA, England

   In  connection with the problem of disproportion/
combination  of peroxy-radicals we have some
experimental  information which may suggest the
involvement  of an H20i, intermediate in the reaction
H02 + H02   H202 + 02(1).  The overall  rate
constant ki exhibits a substantial negative  tempera-
ture dependence (exp(+1250/T)) and a  small
pressure effect (30 percent decrease  between  760
and 40 Torr).  Other unpublished results  from
Burrows and Thrush (Cambridge, U.K.)  suggests an
even lower rate constant at ^ 5 Torr.  These
effects may possible be understood in terms of the
formation, by the combination of 2 H02 radicals,
of an H^ molecule which can be vibrationally
relaxed by energy transfer:
                                    H02
                                                                  H02
                                                          H202 + 02
                                  If formation of products H202 + 02 (as opposed
                               to redissociation to 2H02) from the vibrationally
                               relaxed H20., molecule is relatively more favored
                               then from HiOi,*, factors favoring population of
                               the lower vibrational levels of H^ will tend to
                               increase the overall rate constant.
                               William P. L. Carter, Statewide Air Pollution
                               Research Center, University of California,
                               Riverside, CA  92521

                                  There is inadequate data concerning the effect
                               of non-hydrocarbon substituents on the decomposi-
                               tions of alkoxy radicals.  Substituted alkoxy
                               radicals of many types are formed in polluted
                               tropospheric systems, and in some cases, it is
                               uncertain whether decomposition of reaction with
                               02 predominates.  As discussed in the previous
                               section uncertainties in the decomposition rates
                               of e-hydroxy-alkoxy radicals present serious
                               problems in developing models for the OH-olefin-
                               NO  system.  In addition to 0-hydroxy-alkoxy
                               raaicals, our detailed propene + n-butane-NOx-air
                               model1 predicts formation of species of the types
                                                           R-CHCH2ON02,  RCH-C-R',  and
                                                                  R-CH-CHO
                               and it is probable that in a more complete detailed
                               smog model, other types of substituted alkoxy
                               radicals would be predicted to be formed.   It is
                               commonly assumed that decompositions of these
                               species predominates, but this requires experiment-
                               al  verification.
                                 Carter,  W.  P.  L.,  Lloyd,  A.  C.,  Sprung, J.  L.,
                                 and Pitts,  J.  N.,  Jr.,  Computer  modeling of smog
                                 chamber  data:  progress  in validation of a
                                 detailed mechanism for  the photooxidation of
                                 propene  and n-butane in photochemical  smog, Int.
                                 J.  Chem.  Kinetics  11. 45  (1979).
                                                   65

-------
                Recommendations

   An understanding of the chemical kinetic behavi-
our of large organic free radicals is necessary to
describe the oxidation of hydrocarbons released
into the troposphere.  Before we describe specific
areas where we consider research is necessary
there are some general points that should be made
points that arise from the particular complexity
of the problem.

   1.  The majority of the existing chemical
kinetic data on the reactions of organic free
radicals have been obtained from measurements of
the ratios of the rates of competing processes -
very often the ratio of a propagation rate constant
to a termination.  It is also true that in practice
such ratios or combinations of rate constants are
often the controlling parameters.  It would be
generally useful therefore if experimentalists
published clearly the actual ratios that they have
measured as well as the rate constants that they
have deduced from them.  Further, the users of rate
data, particularly modelers, should state what is
needed for given situation, absolute or relative
rate constants.

   2.  There are too many hydrocarbons and too many
resultant radicals to be able to hope to study all
their potential reactions individually.  This means
that we fully support the thesis of Golden in his
talk that we should consider the behaviour of
classes of radicals.  We should make sure that any
new results are consistent with observations from
other members of the same class and examine devia-
tions carefully.  At the present time, this means
that the thermodynamic properties of the key
radical classes such as alkyl, alkoxy and alkyl
peroxy need to be firmyl reestablished after the
recent upheavals.  The kinetic parameters from the
different types of processes have also to be
measured.  At the present time we are beginning
to obtain absolute rate information for the first
members of the different classes.  We will have to
extrapolate to larger radicals that we cannot
easily produce in the laboratory to estimate
firstly, rate constants for processes that are
identical to those undergone by smaller radicals,
e.g., abstraction from a substrate hydrocarbon,
and secondly, to predict rate constants for
processes such as alkoxy rearrangements or
cleavages that the smaller species cannot them-
selves undergo.  The results of recent product
studies show such processes are important and the
only way they will be consistently and rationally
modeled will be by the establishment of a sound
and consistent data base for the radicals that
can be studied.

   3.  The techniques for making careful systematic
measurements for individual organic free radicals
in the gas phase at temperatures of atmospheric
interest are now becoming available.  The infrared
and ultraviolet absorption spectra of several
larger alkyl and alkyl peroxy radicals have been
detected and used for kinetic studies.  The agree-
ment between flash photolytic and modulation
methods has been good.  So too, to a large degree,
has been the similarity for, for example, peroxy
radical combination in the gas and liquid phases.
Kinetic studies using direct detection of alkoxy
radicals would make a key contribution that would
help complete the stody of aliphatic radical
chemistry.  Oxidation chemistry can now do more  •
than just elucidate the admittedly complex mechah-
isms and can produce directly measured quantitative
data.

   4.  These developments in detecting transients
have been accompanied by advances in the methods
of detecting stable products.  Long path ir has
been extended by using Fourier transform techniques
and the chromatography of peroxides is becoming
more reliable.  Such studies tell the paths that
the radical reactions must follow.  New techniques
such as laser pyrolysis when they are applied to
larger molecules will confirm whether these
suggested paths are reasonable.

   5.  Classical determinations of radical thermo-
chemistry could perhaps be supplemented usefully
by structural information from the spectroscopic
information that is now appearing for larger
organic radicals.

   6.  The existing theory of chemical kinetics,
particularly transition-state theory and its
deriatives provides a useful tool for rationalizing
the kinetic information (see 2).  It is important
to emphasis however the degree of accuracy in
energy measurements that is needed for prediction
at ambient temperatures using rate expression in
the Arrhenius form.

   The following specific points were raised during
the discussion:

   1.  Alkyl Radicals.   The thermochemistry of
these key initial building blocks is the subject
of controversy at the present time and must be
clarified before we can proceed with confidence
with more complex radicals.   There is no longer any
dispute about the order of magnitude of the rate
constants of combination, but there is uncertainty
about the heat of formation and the structural
details of the radicals.

   2.  A1koxy Radicals.   Here there is a glaring
need for a direct detection technique useful for
kinetic studies.  The cleavage reactions appear to
be relatively satisfactorily described but the
rates of reaction with Qz need measurement.  This
means in turn consistency between measurements
against different reference reactions e.g. alkoxy
+ hydrocarbon rate constants.  The rate parameters
for the recently proposed isomerizations of the
large alkoxy radicals need testing preferably
directly,  or if not, against an improced data base
of alkoxy measurements.   Recently it has been
claimed that alkoxy radicals have been detected by
emission   if confirmed this could produce a break-
through.

   3.  Alkyl Peroxy.  The key reactor here with NO
has so far proved too fast for direct measurement
but in the light of the changed values for H02 +
NO needs to be checked.  The rates of radical
combination that are important in lightly  polluted
situations are in the process of being firmly
established but the mechanistic explanation of the
results is speculative.  Further information  about
the different channels in the combination  is
needed together with better  product data.
                                                   66

-------
   4.  Acetyl  and  Acetyl  Peroxy.   The absence  of         5.   Information  from Solution.  There  is  a  vast
discussion  here  on what are precursors on  the         body of data on low temperature  radical chemistry
route  to  PAN  illustrates  an important area of         in  solution.  This  could  form a  useful  basis for
ignorance.  The  recent detection  of acetyl radicals   ideas,  comparisons  with gas  phase work  and surface
should allow  a start in this area and permit         processes.
checking  of the  classical  works  in the field.
                                                   67

-------
  Session IV
NOX Chemistry

-------
                TROPOSPHERIC CHEMISTRY OF NITROGEN OXIDES   A SUMMARY OF THE STATUS OF
                                        CHEMICAL KINETIC DATA
                                            Richard A. Cox

                                  Environmental and Medical Sciences
                                          Division, A.E.R.E.
                                    Harwell, Oxfordshire OX 12 ORA
                                            United Kingdom
        This paper is a review of the tropospheric chemistry of nitrogen oxides.  The  important
     atmospheric reactions and the photolysis of these compounds are discussed and problem areas
     are emphasized.

     Keywords:  Nitrates; nitrites; nitrogen oxides; photolysis; reactions; review; tropospheric
                 chemistry.
                 1.   Introduction

   An important role for nitrogen  oxides in the
chemistry of the lower atmosphere, both from the
point of view of urban atmospheres (Leighton,
1961) and natural trace gas budgets, (Junge 1963,
Levy 1972)  has been recognised for some time now.
The term 'NO '  in air chemistry has usually been
synonymous  with the commonly known oxides of
nitrogen, NO and N02, but recent fashion in ter-
minology refers to total 'odd nitrogen' species
which includes, in addition to NO  and N02, the
higher oxides of nitrogen,  N203, N20i,, N03, N205
and also the oxyacids of nitrogen, HONO (nitrous)
and HON02 (nitric).  A significant role is also
now believed to be played by peroxynitric acid,
H02N02.

   In any model of the chemical transformations
in urban air, the chemistry of the organic
derivatives of the oxyacids, alkyl nitrites, alkyl
nitrates and the peroxynitrates must be considered.
Especially  important are the peroxyacylnitrates
(PAN's) which observational data show to be one
of the most important compounds in photochemical
smog.

   An important atmospheric nitrogen oxide, not
normally included under the terminology 'NO  ',
is nitrous  oxide N20.  Present knowledge do.ls not
point to a  role for this oxide in  the tropospkeric
gaseous nitrogen cycle.  However,  observational
data suggests that there is a sizeable unidentified
sink for N20 in the troposphere.  It may be
appropriate, therefore, to  consider any chemical
kinetic data which might relate to this problem.

   Finally, reduced nitrogen compounds, NH3 and
its derivatives, should be  mentioned, since the
problem of  coupling of the  NH3 and NO  cycles has
been raised from time to time (Robinson and
Robbins, 1971).  NH3 undeniably plays an important
role in the aerosol and precipitation chemistry of
nitrates.  Relating more to the current chemical
kinetic data assessment, is the problem of oxida-
tion of NH3 to NO (or N02).

   2.  Importance of NOV in Atmospheric Chemistry
                       A

   Nitrogen oxides and related species are import-
ant atmospheric pollutants in their own right,
e.g. the toxicity of N02 and the corrosive nature
of N02 and nitric acid toward many types of
materials.  However for the atmospheric chemist
it is the interaction of NO  with other chemical
species in the atmosphere and the resulting
influence on the basic trace-gas cycles and the
formation of secondary pollutants, which is of
interest.  It is in the solution of problems
arising from these interactions where chemical
kinetics can play a role.  These problems are
primarily related, both in the natural and the
polluted troposphere, to the photochemical oxida-
tion of hydrocarbons.

   It is now well-established that the atmospheric
oxidation of hydrocarbons and relative substances
(i.e. oxygenated and halogenated organics),
proceeds by a photochemically initiated free
radical chain process.  The chain carrying radicals
are OH, H02 and their organic analogues RO and
R02.  Nitric oxide, NO, is involved in this chain
process through its ability to convert the
relatively inactive R02 radicals to active RO
species via the general atom transfer reaction:

       R02 + NO = RO + N02   (R = H, alkyl, etc.)

Nitrogen dioxide, on the other hand, acts as a
chain 'terminating species through its recombina-
tion reactions with RO and R02 radicals, e.g.

       HO + N02 -* HON02

       R02 + N02 J R02N02.
                                                   71

-------
In the lower atmosphere HON02 formation provides
an efficient sink for both N02 and HO  species.
Formation of peroxynitric acid and the peroxy-
nitrates acts as a more or less temporary sink,
since the products are thermally unstable and
redissociate to R02 and N02.

   The other important role of N02 arises from its
rapid photo-dissociation to give ground state
atomic oxygen and nitric oxide:

       N02 + hv(X <_  400 nm)   0(3P) + NO

followed by

                  0 + 02 + M   03 + M  .

This  provides  the  source  of ozone  in  photochemical
smog  and  also  in the  natural  troposphere.  Also
the primary  source of OH  radicals  in  the  lower
atmosphere,  is through  production  of  excited  atomic
oxygen,  0 *D,  from the  photolysis  of  03  in the
ultra-violet region at  X  £ 310  nm:

       03 + hv   0(1D) + 02 X < 310 nm

   0(1D) + H20   20H

Thus the involvement of N02 in the tropospheric
ozone budget, has a direct bearing on the average
concentration of OH in the lower atmosphere and
consequently on the atmospheric residence times of
a variety of trace-gases.

   A model of atmospheric NO  chemistry is there-
fore necessary to (a) formulate a realistic model
of photochemical smog on which to base control
strategy and (b) to provide an understanding of
the global tropospheric trace-gas cycles.  The
objective of such a model is to predict the total
budget of NO  from source to sink, and the
distribution of NO  among the various chemical
species in time ana in space.  This distribution
depends on the nature and strength of the sources,
the chemical interactions within the atmosphere.  ..
and the role of the various sink mechanisms.

   3.  Cycle of NO  Through the Lower Atmosphere

   The currently accepted picture of the cycle of
gaseous-nitrogen oxides through the lower       	
atmosphere is similar to  that formulated originally
by Robinson  and Robbins (1971).  It was based on
observations and measurements of the distribution
of atmospheric NO , chemical reaction rate data,
and meterological factors.   Figure 1  shows a
schematic illustration of the cycle.
LIGHTNING
  Fig. 1.  Life-cycle of nitrogen oxides in the
           troposphere.
   The main source of NO  is believed  to  be
emission of NO from the ground, either from  man-
made scwrces, mainly combustion, or from  soil
processes.  An additional source of  NOX is fixed
atmospheric N2 from lightning. There  is current
argument about the magnitude, both relative  and
absolute, of these sources.

   Once in the atmosphere, chemical oxidation  of
NO to N02 occurs rapidly, primarily through  the
reaction

       03 + NO = N02 + 02.

In daylight NO is reformed by photolysis  of  N02,
but is also oxidised by photochemically generated
radicals, i.e.

       R02 + NO + RO + N02.

   Removal of N02 is primarily via formation of
nitric acid, with alternative pathways via organic
nitrates and pernitrates.  Formation  of peroxy-
nitric acid and PANs is reversible but this  can
act as a sink if PANs are removed, for example,
by absorption at the ground.  Additional  removal
of N02 can occur by reaction with 03  to give N03.
N03 is rapidly photodissociated (in dayl'ight)  but
also reacts with NO (to reform N02) or with  N02
to give N20s.  The latter reaction is  reversible
and so N02, N03, N205 and 03 can exist in
equilibrium.  N205 can also be converted  to  HN03
by heterogeneous reaction with water.

   The nitric acid, N205 and organic nitrates can
all be removed from the atmosphere by  absorption
at the ground (dry deposition) or by  incorporation
into the precipitation elements   aerosol
particles, cloud and fog droplets, which  eventually
leads to rain-out.

   4.  Status of Chemical Kinetic Data

   Accurate chemical kinetic data is clearly
required for the primary chemical processes
involved in the transformation of NO to nitric
acid.  Also of interest is data relating  to all
possible minor interactions which would influence
the basic atmospheric NO  cycles, or which produce
unusual secondary pollutants in urban  air.   Due
to the recent stimulus in the field of atmospheric
kinetics, data for some of these processes is now
rather well  known.  For some processes  more and
better data is badly needed, and these  become
self-evident during any detailed discussion of the
data base.  In the following paragraphs the status
of the data base is very briefly indicated for
some specific areas of NO  chemistry mentioned
above.   The topics covereS should not  be  considered
exhaustive,  but rather a minimum set necessary for
modelling the basic NOX cycle.

   A.'. Photochemical data - for N02, N03, N205,
       HONO, HON02 and H02N02

   In order to calculate photodissociation rates
for a species in the atmosphere, a knowledge of
the absorption cross-section a as a function of
wavelength and the quantum yield(s) of the photo-
dissociation pathway(s), $ls is required.  This
is then combined with suitably averaged data for
the photon-flux in the atmosphere to obtain  the
                                                   72

-------
'J value1  which is essentially a first order rate
constant for photochemical removal of that species.

   Reasonably reliable data are now available for
o and $1 for the photodissociation of N02, HOMO
and HONO;  over the important wavelength regions

       N02 + hv = 0 + NO
      MONO + hv   OH + NO
  HON02
                  OH + N02
Data on absorption cross-sections and on the
various possible dissociation pathways of N03,
N205 and H02N02 is much less satisfactory.  Data
for NOs is particularly important since recent
work has indicated that photodissociation can
follow two different pathways, depending on wave-
length:
N03
               NO
               N02
                       02
                        0
Only a limited amount of data is available on the
absorption cross sections and dissociation quantum
yields of the organic nitrites and nitrates.
B.
       Reactions of NO  with odd-oxygen species
                      A
   The reactions of NO and N02 with the odd oxygen
 species 0 and 03 have long been recognized as
 important for aeronomy and reliable rate data is
 now available for the following reactions
       0 + NO(+M)    N02

          0 + N02    NO + 02

      0 + N02(+M)    N03(+M)

          03 + NO    N02 + 02

          03+ N02    N03 + 02


   Less well known is the chemistry of N03 and N205
and a number of investigations have sought to
define the rate constants for the following
processes, all of  which are needed for modelling
NOV in the urban atmosphere:
  A

       N03 + NO   2N02

                (M)
       N03 + N02 = N205

Due to the chemical  complexity of these systems,
there is.some uncertainty in the kinetic parameters.
N03 may also undergo  H-abstraction reactions  with
organic molecules  and some kinetic parameters for
these reactions have  been reported.   The effective
rate of the reaction  of N205 with water which is
probably heterogeneous,  is rather uncertain at this
time and could be  of  importance in the overall NOX
budget in the lower atmosphere.   A quantitative
treatment of the rate of heterogeneous removal of
gaseous species on aerosol  particles and cloud and
fog droplets,  which is acceptable to many modelers,
has yet to be formulated.
   C.
    Reaction  of NO  with  odd-hydrogen species
   The coupling of the NO  and HO  cycles is one
of the most important aspects of atmospheric free-
radical chemistry.  The reactions of hydroxyl  (HO)
radicals with NO  species has been widely studied
in response to problems of aeronomy, and a
reasonably good data base is available here.   The
important reactions are:

       HO + NO (+M)   HONO (+M)

       HO + N02(+M)   HON02(+M)

          HO + HN03 = H20 + N03

          HO + HONO   H20 + N02

Note that the M dependent reactions are in the
transition region between third-order and second-
order kinetics at the pressures encountered in
the troposphere.  If the actual measurements of
the rate constants as a function of pressure for
M   Air are not available, the rate constants  in
the transition region can be estimated from a
knowledge of the third order low pressure rate
constants, k,,,, and the high pressure second-
order rate constant k^.  These two rate constants
kjTI and k  therefore°°comprise a minimum data
set for thTs type of reaction.  The temperature
dependence of these association reactions is also
important since in the low pressure regime they
usually exhibit a significant negative temperature
coefficient.  This can be important in modelling
NO  circulation in the global troposphere.
  X

   The reactions of the hydroperoxyl  radical, H02
with NO and N02 are also very important.   Kinetic
information on H02 reactions is not as complete
as that for HO reactions but application of new
free radical detection techniques to kinetic
studies of H02 has led to a significant improvement
of the reliability of the data base.   Recently
new data has been reported for

       H02 + NO + HO + N02

       H02 + N02(+M) J H02N02 + M
                                                      which allows a more quantitative appraisal  of the
                                                      role of these reactions.   These studies have
                                                      shown in particular that  the reaction of H02 (and
                                                      probably other peroxy radicals) with NO are much
                                                      more rapid than had hitherto been believed, and
                                                      they then assumed much greater significance in
                                                      atmospheric free radical  chemistry.   The formation
                                                      of peroxynitric acid is reversible and associated
                                                      data for its thermal decomposition which has
                                                      recently been obtained allows the role of this
                                                      'new'  species to be assessed with some confidence.
                                                         D.
                                                          Reactions  of  NOV with  organic  radicals
                                                                         A
                                                  The possible reactions of NO and N02 with
                                               organic radicals are numerous.   However reactions
                                               with organic peroxyradicals appear to be the most
                                               significant for atmospheric chemistry, and these
                                               are exemplified by the reactions of NO and N02
                                               with peroxyacetyl  radicals:

                                                      CH3C(0)00 + NO -> CH3 + C02 + N02

                                                      CH3C(0)00 + N02 J CH3C002N02(PAN)

                                               These reactions govern the formation of peroxy-
                                               acetylnitrate in urban air and  show clearly the
                                               competition between the chain carrying reaction
                                                   73

-------
involving NO and the chain terminating step
involving N02.   Also since PAN can decompose back
to CH3C002 and N02, it only acts as a temporary
sink for radicals.  Reasonably reliable kinetic
data for the thermal decomposition of PAN are now
available and the relative rates of reaction of
the peroxyacetyl radicals with NO and N02 are
moderately well-defined.

   There is little or no kinetic information
concerning the analogous reactions of other organic
peroxy radicals with NO and N02, or the thermal
stability of the peroxynitrate species produced
in the reaction with N02.  Several of the peroxy-
nitrates have been identified in the laboratory by
infra-red spectroscopy and they may therefore
play a significant role in urban smog formation.
Evaluation of the rate parameters for the
analogous reactions of other organic peroxy
radicals with NO and N02 is necessary to define
the chemistry of the breakdown of individual
organic species during atmospheric photo-
oxidation.

References

    Junge, C. E., Air Chemistry and Radioactivity
    (Academic Press, New York, 1963).

    Leighton, P. A., Photochemistry of Air
    Pollution  (Academic  Press, New York, 1961).

    Levy,  H.  II.,  Photochemistry of the lower
    troposphere,  Planet. Space Sci. 20_, 919-935,
    (1972).

    Robinson,  E.  and Robbins,  R. C.,  Sources,
    abundance and  fate of  Gaseous Atmospheric
    Pollutants-Supplement,  American Petroleum
    Institute Publication  No.  4015, April  (1971).
               Summary of Session

    This  session was  concerned with  the various
 simple reactions of  NO, N02, N03 and N205, few of
 which are  well understood.  The conversion of
 N02 into N205  and  its  subsequent reaction with
 water to give  nitric acid  is of great importance.
 Levine noted that  if the N205-H20 rate constant
 were fast  enough it  would  play a significant role
 in  N02 removal.  O'Brien discussed  smog chamber
 data in  which  the  fate of  N03 depended on the
 substate present,  suggesting aerosols may be
 involved in some cases in  HN03 formation.  Stedman
 noted that N02 disappearance rates  at night
 corresponded to the  N02-03  rate, but that the
 products were  not  able to  regenerate the reactants,
 which suggested that HN03  might be  a product.
 Benson suggested an  additional possible product,
 pernitric  acid, arising from the reaction of OH
 with NO3.

    The photolysis  of N02 is still a subject of
 controversy since  new  data  (reported by Whitten)
 suggest  that the quantum yield is less that one
 around 380 to  < 400  nm.  It was also mentioned by
 Demerjian  that the same workers measured different
 absorption coefficients than the NBS workers
 although the latter  values  are still to be
 preferred.
   Basco commented on his earlier work on  the
flash photolysis of N02 in which the 02  product,
arising from the secondary 0 + N02 reaction, was
monitored.  The 02 should be observed up to the
12th vibrational level corresponding to  46 kcal  ,
excess energy.  In fact it was found up  to the
16th level which is 12 kcal higher with  radiation
greater than 400 nm.  Whether the photoexcited
N02 implied by these results plays a role  in
atmospheric chemistry was not discussed.

   The new high value of Howard for the  rate
constant of H02 + NO has been used by most
modelers for the ROZ + NO reaction.  Golden pointed
out that the H02 reaction has a negative temperar
ture coefficient which implies a bound state.   The
transfer of rate constants may thus be invalid.
Heicklen raised the possible role of pernitrous
acid in this reaction.  Hendry presented data on
the pyrolysis of nitrates which support  the high
values for R02 + NO.  Parkes also noted  that in
flash photolysis systems the t-butyl peroxy and
methyl peroxy radicals could never be seen in the
presence of NO implying a very high rate constant
( -\j 10"11 cm3 molec :s x).  Another unresolved
question was whether adducts are formed  i.e.,
R02 + NO -»• ROONO which can rearrange and cleave.
Carter presented smog-chamber results which
suggest that the formation of alkyl nitrates is an
important reaction.  Heicklen also suggested the
possibility of a reaction R02 + NO -»• MONO +
aldehyde involving a six member cyclic transition
state.

   The reactions of RO with NO and N02 were also
considered.  The reaction of RO with NO  leading
to the nitrate is not too important in the
atmosphere (but could be in the laboratory) since
the nitrate is rapidly photolyzed.  However, as
Heicklen noted, part of the reaction leads to
HNO + aldehyde.   It seems to be agreed  that the
principal  fate of RO in the atmosphere  is either
isomerization, scission,  or reaction with 02 -  as
discussed in the Free Radical  session.

   The more general  problems of modeling were
discussed by Dodge in terms of current  problems in
NO  chemistry.  Questions as to whether the correct
photolysis rates are used and how well  the models
should be expected to fit the observations were
brought up but not resolved.  Carter discussed
radical  initiation in smog chambers, emphasizing
that it probably arises from contamination of the
chamber.   The real  problem lies in its  unpredict-
able nature.  Solution of this problem must be
given the highest priority.


                    Comments

Robert J. O'Brien, Patrick J. Green, and Richard
M. Doty, Department of Chemistry, Portland State
University, Portland, Oregon  97207

   We have been analyzing several of the UCR smog
chamber experiments for their nitrogen balance.
The details of this analysis will be published in
the near future and only the results will  be
discussed here.

   For the case of a hydrocarbon which reacts only,
with hydroxyl radical and negligibly with  other
                                                    74

-------
free radicals or ozone we may derive an equation
for the NO  balance based on the  following
mechanism:
? Nfl., + OH
•3 ar\n + n
H NO, + MO-
M


     5  N205 + H20

     6  N03 + NO  -
                     surface
                                loss

                                HN03

                                N03 + 02

                                N205


                                2HN03

                                2N02
   Under the assumption that the OH concentration
 is given by
                    _J	      d[HC]
                                  dt
we may obtain  the  equation

     ^initial    £NOXJ

      k2  r [N02]
                 d[HC]  + k3
                            f [N°2][°3]dt
   This equation assumes all N03 forms nitric
 acid, and that the total NOX loss  (except for
 PNA) is accounted for in HN03.
   200
    160
 -5  120
 :=  so
    40
             EC-77
                                    = 2.32
          J	I
                        I    I     I    I     I
              20
                       40

                       [N02]
                                60
                            d[T]ppb
                                         80
Fig.  1.   Correlation  of  NOX  loss with  reaction  2
         for UCR  run  EC-77.  T  = toluene.

   Figure 1 shows a plot of the left hand side
of eq. (1) vs.:
                      [N02]
                   TO ~m
                             d[T]
for UCR run EC-77 which is a toluene (T) run which
made no ozone.  The slope of this line, (k2/ki),
 2  200 -
                                                      Fig.  2.   Correlation  of NOX  loss  with  reaction  2
                                                               for UCR run  EC-80.   T    toluene.

                                                     is equal to  2.3 which is in good  agreement with
                                                     literature values for k2 and  ki.  Figure 2 shows
                                                     a similar plot for EC-80, a toluene run which made
                                                     ozone.  The  initial slop agrees with the previous
                                                     run and the  upward curvature  coincides with the
                                                     appearance of ozone.  The deviation between the
                                                      experimental data and the extrapolated straight
                                                      line  is  correlated with the  second  term  on  the
                                                      right hand  side  of eq.  (1)  in figure 3 which  is
                                                      a plot of
                                                       [NOX]1
                                                               [N0x]
                  [PAN]   1.9
                                                              vs.  / [03][N02J  dt
   An excellent linear relationship is obtained.
If all N03 formed HN03 via the above mechanism we
would expect the slope of figure 3 to be 2 k3 or
.100 ppm 1 min"1.  The actual slope is .032 ppnf1
min'1.

   Figure 4 gives a plot similar to figures 1 and
2 for EC-83, a run at zero relative humidity which
made 0.42 ppm 03 with 2 ppm initial NO .  Note
that there is an insignificant derivation from
linearity indicating little conversion of N03 to
HN03 in the absence of H20.

   Table 1 gives a summary of 10 UCR toluene runs.
The values of k2/ki obtained agree well with
literature values and have a standard deviation of
10 percent.  The values obtained for the slope  of
the curves similar to figure 3 show greater scatter
and indicate that about 1/4 to 1/3 of the N03
formed is converted to HN03.  However, the kinetic
analysis is not altogether clear as to the meaning
of the linearity and value of the slope in these
plots.
                                                   75

-------
        200
Table 1  Summary of experimental  and computed data for UCR toluene  reactions.
                                                    Slope of
      Initial  Initial  Initial Maximum  Maximum Relative        figure 3
     1000
                            (N02) (03) dt pprrT mm
     Fig. 3.   Correlation of  upward  curvature from
               figure 2 with reaction 3.   t  = time.
                                        PPb
  Fig.  4.   Correlation  of NOX  loss  with  reaction 2
             for UCR run  EC-83 a dry  reaction.
             T   toluene.

    Figure  5 shows a  plot similar  to figure  1  for a
UCR butane run,  EC-41.   Although  considerable
ozone was  formed, no  upward  curvature  is observed
in  this  or any other  butane  run,  indicating  no
conversion of N03 to  HN03.   The slope  of the plots
for 10 UCR butane runs  are  summarized  in table 2
and give a value of  k2/ki of 3.7  ±  9 percent (rel.
std.  deviation).   This  value  is in  good  agreement
with  some  literature  values.
                                                               Rxn No.
                                                                 EC-

                                                                77

                                                                78

                                                                79

                                                                80

                                                                81

                                                                82

                                                                83
       NO/NO;
       ratio

       8.9

       2.2

       4.2

       4.2

       4.3

       2.0

       2.1

       4.9

       4.7

       5.!
NO
ppS

574

100

100

500

500

1000

2000

470

520

490
toluene
 ppb

 276

 230

 980

 1000

 2000

 1900

 5600

 970

 1900

 1100
0,
ppb

12

92

96

27

313

365

420

230

290

300
PAN
ppb

 2

13

15

47

61

59

65

43

50

37
humidity k2/ki    plots
       ratio  ppnr'min
                   -i
 40

 40

 40

 40

 40

 40

  0

 70

 48

 35
                                                                                                     Average

                                                                                      Relative standard deviation
2.3

2.6

2.7

1.9

2.5

1.9

1.6

2.1

2.2

2.0

2.3

10S
0.026

0.147

0.032

0.041

0.022

0.004

0.019

0.046

0.022

0.029

 40?
                                                                  ~  100 -
                                                                Fig.  5.   Correlation  of NOX  loss  with  reaction 2
                                                                           for UCR run  EC-41.   B    butane.
                                                                        Table 2.  Summary of experimental and computed
                                                                                data for UCR butane reactions.
Rxn No.
EC-
39
41
42
43
44
45
46
47
48
49
Initial
NO
pp6
0.6
0.593
0.6
0.137
1.26
0.614
0.587
0.599
0.594
0.611
Initial
butane
ppb
2.2
4.03
0.385
0.38
3.92
1.94
4.00
3.9
1.94
4.12
Maximum
0,
ppb-
0.073
0.237
0.006
0.12
0.015
0.138
0.252
0.255
0.163
0.286
k:A>
ratio
3.8
3.56
3.73
3.6
4.4
3.6
4.04
4.17
3.77
3.8
                               Average  3.7

                Relative standard deviation   9%
                                                            76

-------
   The very interesting tentative conclusion we
reach  is  that N03 in the presence of water is
converted to HN03 in the toluene runs but not in
the butane runs.   The explanation may lie in
aerosol  formation which occurs with toluene but
not with  butane.

   For runs which form no ozone, or before ozone
buildup occurs in any run, we are able to use an
analysis  similar to the one in figure 1 to account
for all  NO  loss by the reaction OH + N02 + M ->•
HN03.   The slope of the line in this plot yields
a rate constant ratio which shows only a 10 per-
cent standard deviation in the two groups of ten
experiments (toluene and butane) which we have
analyzed.  This rate constant ratio may then be
used for computer modelling of the series of
reactions with the confidence that the predicted
HC and NO  loss rates will agree within ± 10 per-
cent  (stdv dev.) of the experimental values,
provided the correct OH concentration profile is
obtained.  For the case of butane the relative
HC and NO  loss rates are not effected by 03 or
NOs formation and are controlled only by OH.
Donald H.  Stedman,  University of Michigan,
(c/o NCAR,  Boulder,  Colorado  80303)

   It has  been frequently observed in our data
and in others from  rural  air that N02 decays at
night and  does not  come back in the morning.  The
decay rate is roughly equal  to the rate of N02 +
N03.  The  product of the reaction is not observed
so it is not stopping at N03, presumably going on
to N205.  When the  sun comes up in the morning
the N02 thus lost does not reappear so the sink
cannot be  photolyzed, thus it has to be going
further to some product such as nitric acid.
 William P. L. Carter, Statewide Air Pollution
 Research Center, University of California, River-
 side, California  92521

   Alkyl nitrate yields observed in alkane-NO -.
 air smog chamber runs suggest that the following
 reactions
          R02 + NO
                    M
RON02
 are important sources of alkyl nitrates when the
 alkyl group, R, is sufficiently large [1].  No
 other mechanism for alkyl nitrate formation can
 explain the observed near-independence of the
 RON02 yield on initial NO  levels, or the fact
 that the possibility of H shift isomerizations of
 some alkoxy radicals do not result in significant-
 ly reduced yields of the corresponding alkyl
 nitrate.  We consider it unlikely that the high
 yields of alkyl nitrates observed in the smog
 chamber runs [1] could be due to heterogeneous
 reactions of NO  with organic products, since the
 measured alkyl  filtrate levels do not significant-
 ly decline following NO  consumption.

   If these reactions are as significant as our
 smog chamber results suggest, then they would
                                have  the effect of making larger alkanes act as
                                radical  inhibitors in  photochemical  smog systems,
                                which  has  significant  implications concerning
                                their  reactivity and effects  on smog formation
                                rates.

                                   Unfortunately, the  only evidence  for these
                                reactions  comes from smog chamber data, and as
                                far as  I  know,  our model  is the only current one
                                which  includes  them.   Direct  laboratory studies,
                                obtaining  unambiguous  mechanistic and kinetic
                                data  concerning the reactions of R02 radicals
                                with  NO,  are  clearly required.   It is
                                particularly  important that a wide variety of
                                R02 radicals  be studied,  in order to clearly
                                establish  substituent  and size effects.

                                Reference

                                [1]  Darnall, K.  R., Carter,  W.  P. L.,  Winer, A.
                                     C., and  Pitts,  J.  N.,  Jr.,  J. Phys.  Chem.  80.
                                     1948  (1976).
                                Dale G.  Hendry,  SRI International,  Menlo Park,
                                California  94025

                                   Peroxyacly nitrates (PANs) play an important
                                part in influencing the OH radical  concentration
                                in the atmosphere.   Under conditions where PANs
                                increase in concentration they act as radical
                                sources.  This effect is associated with the
                                facile equilibrium
                                        ROON02
                   ROD- + N02
where R   CH3C(0)02, CH3CH2C(0)02, etc.

   Peroxyalkyl nitrates are a second type of
peroxynitrate  (where R = CH3, CH3CH2, etc.) that
could also be important in affecting the OH con-
centration.  Richard Kenley and I are currently
studying the decomposition of these types of
compounds and, in the case of peroxy-t-butyl
nitrate, find evidence for a 20-23 kcal mol 0-N
bond strength, which is considerably weaker than
found for peroxyacetyl nitrate (27 kcal/mol).

   When reactions 1 and -1 for peroxyalkyl
nitrates are included in atmospheric models for
propene, n-butane, and toluene, we find they have
no significant effect on the overall chemistry if
it is assumed that log AI = 16.5 s"1, EI = 23
kcal/mol, log A-i = 9.0 s'1, and E-i = 0 kcal/mol.
However other reactions of peroxyalkyl nitrates
beside reactions 1 and -1 may be important.  For
example if the reaction
                                  ROONO-,
                 0-0
                /     ^
               C         N-0
             ' \     //
                 H   0
C = 0 + HON02
                               competed with reaction -1 then the formation  of
                               peroxyalkyl nitrates could be important  radical
                               sinks.  Additional information is needed on the
                               chemistry of peroxyalkyl nitrates to  develop
                               reliable atmospheric chemistry models.
                                                   77

-------
Marcia C. Dodge, Environmental Protection Agency,
Research Triangle Park, North Carolina  27711

   We are encountering difficulties  in our model-
ing efforts that appear to be related to NO
chemistry.  In our modeling of smog  chemabeP data
for the  propylene-NO  system, we are able to
obtain respectable ffts for all species except
03.  As  an example, in one particular run
conducted in the UCR evacuable chamber, we
achieved excellent fits for NO, N02, PAN, and
propylene.  The predicted concentrations of
formaldehyde and acetaldehyde were a little low
in this  run, but well within the experimental
uncertainty of the measurements.  Despite the
good agreement obtained for these six species,
the peak 03 concentration for this particular
simulation was 42 percent too high.  The
experimental peak occurred at 0.38 ppm whereas
the simulated 03 max occurred at 0.54 ppm.
Similar  results are found for the other propylene/
NO  runs and for several other hydrocarbon/NO
systems  as wel1.

    In our modeling we are using Howard's value of
8.1 x 10"12cm3/molecule-second for the rate of
reaction of H02 with NO.  Prior to Howard's
determination, we were using a rate  constant of
1.4 x 10~12cm3/molecule-second for this reaction.
If we repeat the propylene-NO  run described above
using the old, much slower rate constant for the
H02 + NO reaction, we obtain a predicted 03
maximum  of 0.39 ppm, in excellent agreement with
the experimental value of 0.38 ppm.

   We have tried to off-set the effect of the new,
fast rate constant for the H02 + NO  reaction.
Although we have varied all rate constants in the
mechanism within their limits of uncertainty, we
have not been able to reduce the simulated 03
levels without destroying the fits for the other
species. With the new H02-N0 rate constant, too
much NO  is consumed by H02 late in the reaction
and 03 continues to build in the simulations long
after the time 03 was observed to level off in
the smog chamber.  We are tempted to conclude
from this that there may be competing reactions
for H02  in addition to the H02-N0 reaction or
Howard's rate constant for the H02-N0 reaction
does not apply at atmospheric pressure.  Perhaps
the rate constants for the reaction  of H02 with
itself or with 03, at atmospheric pressure and
in the presence of water vapor, are  higher than
the values currently accepted for these reactions.
It  is also conceivable that some of  the HOONO
intermediate formed in the H02 + NO  reaction may
be stabilized at atmospheric pressure so that
the effective rate of formation of OH and N02 is
less than that measured by Howard at low
pressure.  It is difficult to say what the source
of discrepancy is, but the fact that many
modelers are not able to handle the  new, fast
H02-N0 rate constant suggests that the role of
NO  in smog chemistry is not yet fully understood.
  A
William P. L. Carter, Statewide Air Pollution
Reseat*eh Center, University of California, River-
side, California  92521

   Perhaps the most  important single uncertainty
in NO  chemistry affecting the problem of
developing unambiguously validated models.for
tropospheric chemistry concerns  initiation in
smog chamber systems.  As is now well known,
model simulations which assume radical initiation'
only from known processes predict overall trans-
formation rates in hydrocarbon-NO -air systems
far slower than those experimentally observed in
smog chambers.  It is highly probable that this
excess radical initiation observed in smog chamber
is due to some aspect of heterogeneous NO
chemistry, since of  the known species formed in
these systems which  can photolyze to give radicals,
only nitrogen-containing species, specifically-'
nitrous acid or alky! nitrites,  photo!ize
sufficiently rapidly that contamination by
currently undetectable amounts could give the
necessary rates of radical input [1].  At least
in the UCR chambers, oxygenate contamination is
far lower than the levels required to give the
necessary radical initiation.

   This excess radical initiation in smog chamber
systems is probably  due to a chamber contamination
effect, and not to some unknown  omission in the
homogeneous mechanisms, or to MONO being
inadvertently injected along with NO .  Evidence
for this was obtained in experiments performed
at UCR employing large Teflon bags inside the
black-light irradiate all glass  chamber.   It was
observed that the overall reactivity of hydro-
carbon-NO -air photolyses in a new, clean bag was
far less $han the reactivity subsequently observed
in the same bag after only a few smog simulation
experiments were performed in it [2].  If the
excess radical initiation were due to HONO
injection or to some deficiency  in the model, and
not to chamber contamination, then high reactivity
should have been observed in the clean, as well as
in the dirty bag.

   In terms of model validation, the most serious
problem caused by this chamber radical source is
due to the fact that it is unpredictable, and
must be represented  in models by some type of
adjustable parameter.  This means that aspects of
the model concerning radical initiation or
termination cannot be unambiguously validated.
A mechanism with erroneously high radical input
in the homogeneous chemistry (such as those
assuming 100 percent fragmentation to radicals in
the 03-olefin system), or with erroneously low
radical termination  rates, can be made to fit
the smog chamber data by suitably reducing the
adjustable chamber radical input parameter, and
vice-versa.  These erroneous mechanisms, which
appear to be "validated" by smog chamber experi-
ments, will then give erroneous  predictions in
ambient air simulations, where the compensating
chamber radical parameter is removed from the1
model.  Thus, the occurrence of  this chamber
radical input phenomenon is clearly  a very serious
problem,  and studies aimed at resolving  it should
be given  very high priority.
                                                    78

-------
 References

 [1]  Carter, W. P. L., Lloyd, A. C., Sprung,
     J. L., and Pitts, J. N., Jr., Computer
     modeling of smog chamber data: Progress
     in validation of a detailed mechanism
     for the photooxidation of propene and
     n-butane in photochemical smog,
     Int. J. Chem. Kinetics 1J_. 45 (1979).

 [2]  Darnall, K. R., Winer, A. M., and
     Pitts, J. N., Jr., unpublished results.
W. Tsang, Center for Thermodynamics and Molecular
Science, National Bureau of Standards, Washington,
D.C.  20234
    It is clear that within the forseeable future
 EPA will place a great  deal of dependence on smog
 chamber  data with regard to pollution  regulations
 and abatement strategies.  It is therefore
 extremely disturbing to learn of the  large  un-
 certainties and  irreproducibilities in  such
 experiments.  One hopes that if environmental
 decisions are to have any basis in science  that a
 vigorous program for the proper validation  of
 smog chamber data would be instituted  as soon as
 possible.   In particular, investigations on the
 nature  of the surface processes, occurring  on smog
 chamber  walls should have first priority.
               Recommendations

   Although significant advancements  have  been
made recently  in understanding  the  role  of nitro-
gen oxides in  photochemical  smog, there  are  a
number  of key  processes in the  NO   cycle for which
rel'iable rate  data do not exist.  In  the session
on NO  chemistry, six major  areas for further
research were  identified.  In order of decreasing
priority, the  recommended areas  for investigation
are the following:

   (1)  Reactions of Alkylperoxy Radicals

       The oxidation of NO  by  alkylperoxy
radicals,

        R02 + NO + RO + N02,

is an important process in photochemical smog  for-
mation.  Rate  constnats, however, have not been
measured for this family of  reactions.   It is
generally assumed that the rate of  reaction  of R02
radicals with  NO is comparable  to the rate of
reaction of H02 radicals with NO.   This  assumption
may not be valid.  Because of the importance of
R02-N0  reactions, the temperature and pressure
dependency of  these reactions should  be  determined.
As a first priority,  it is recommended that  the
rate of oxidation of  NO by methylperoxy  radicals
be investigated.  Methylperoxy  radicals  are  the
most prevalent of the R02 radicals  and the fate  of
the CH302 radical is  important  to an  understanding
of the  background troposphere.

   In addition to the oxidation of  NO by R02
radicals,  it  has  been postulated that longer chain
alkylperoxy  radicals  may  add to NO  to form excited
complexes that decompose to alkyl  nitrates:
         R02 + NO -*  (R02NO)
RON02
Since alkyl nitrate formation is radical terminat-
ing, the rearrangement shown in the  above equation
could have a significant impact on smog chemistry.
It is important, therefore, to determine the
extent to which this chain-terminating reaction
can occur.

   Another process of potential importance in the
polluted atmosphere is the reaction  of alkylperoxy
radicals with N02 to form peroxynitrates,

         R02 + NO  j R02N02,

and the subsequent decomposition of  the nitrates.
If this class of reactions is analogous to the
H02 + N02 reaction, the alkyl peroxynitrates would
be too short-lived to be of importance in smog
chemistry.  They could, however, play a role at
higher altitudes and at low temperatures and are
worthy of study for this reason.

   The possibility of the alternative channel of
decomposition,

         R02N02 •*• RCHO + HON02,

should also be investigated.  Decomposition to
nitric acid could be important even  if this route
occurs to only a small extent.

   To model the behavior of the alkyl peroxynitrates
(and H02N02 as well), information is needed on the
absorption cross-sections and the products arising
from the photodissociation of these  nitrates.  The
photochemistry of these species, however, is
expected to be of greater importance in the strato-
sphere than in the lower troposphere.

   (2)  Chemistry of Peroxyacyl Nitrates

        It was recently determined that PAN under-
goes rapid thermal decomposition:

         PAN -* CH3C(0)02 + N02.

The rate of this decomposition as a  function of
temperature is reasonably well-known.  Reliable
kinetic data are also available on the relative
rates of reaction of the peroxyacetyl radical with
NO and N02,

         CH3C(0)02 + NO ->• CH3 + C02 + N02

         CH3C(0)02 + NO •> PAN

Rate data, however, do not exist for the analogous
reactions of the other peroxyacyl radicals.
Atmospheric observations indicate that the PAN-type
compounds are the most stable of the organic peroxy-
nitrates and, therefore, merit study.  As a first
order of priority, the thermal stability of the
higher analogs of PAN should be determined.  If
these compounds are sufficiently stable, the
relative rates of the reactions of the correspond-
ing peroxyacyl radicals with NO and N02 should be
measured.   The PAN-type compounds recommended for
study are peroxypropionylnitrate (PPN) and peroxy-
benzoyl  nitrate (PBzN), important for understanding
                                                   79

-------
the fate of aromatics in smog.

   (3)  Reactions of OH and H02 Radicals with NO
        and N02
and N02,
The reactions of hydroxyl radicals with NO


 OH + NO + (M) + HONO + (M)

 OH + N02 + (M) •> HON02 + (M),
have been studied by a number of investigators.
Uncertainties, however, in the rates of these
reactions under atmospheric conditions still exist.
These pressure-dependent reactions are in the
changeover region between third-order and second-
order kinetics at atmospheric pressure.  Because of
this pressure effect, additional studies of these
reactions are warranted.

   New  rate data have been reported recently for
the reaction

         H02 + NO -* OH + N02

The new rate constant is significantly higher than
previously determined values for the rate of this
reaction.  The faster rate constant has a substan-
tial impact on the predictions of photochemical
models.

   Direct determinations of the H02-N0 rate
constant have been made only at reduced pressure.
Because of the significant impact of this reaction
on model calculations of smog formation, the effect
of pressure on the reaction rate should be
determined.  In addition, the effect of water
vapor on the rate of this important reaction
should  be elucidated.

   (4)  N03 Chemistry

        The reactions of N03 have been extensively
studied by Johnston.  Recently he corrected his
previous rate data for the reactions,
          N03 +  NO -

          N03 +  N02
             2N02

            + N205
 An  independent  confirmation  of  the  pressure  and
 temperature  dependency  of  these reactions  is
 recommended.

    A confirmation  of the recently published  absorp-
 tion cross-section and  quantum  yields  for  N03
 photolysis  is  also desirable.   Two  pathways  have
 been identified for this photodissociation:


          N03 + hv  ->- NO  + 02

          N03 + hv  -+• N02 +  0

 The significance of these  processes in the perturb-
 ed  troposphere appears  to  be minimal;  however,  N03
 dissociation could be of importance to an  under-
 standing of the natural troposphere.
    (5)  Reactions  of Alkoxy  Radicals

        Alkoxy  radicals  in the  lower  troposphere
 undergo bimolecular reaction with  Qz,  NO,  and  N0'2.
 To  assess the importance of  the alkoxy-NO
 reactions,  it is necessary to know the rate of
 reaction of RO  radicals  with NO and N02 relative
 to  the rate of  reaction  with 02.   Such relative
 rate  data do not exist.

    Reaction of  alkoxy  radicals  with N02 can proceed
 by  two pathways:
          RO

          RO
                                                             N02

                                                             N02
RON02

RCHO + HONO
The rates of the additipn_and^bstraction reactions
for the various alkoxy radicals are reasonably
well-known.   Rate constants for the alkoxy-N02
reactions relative to 02, however, are less well-
defined.

   Reaction of alkoxy radicals with NO also can
proceed by two pathways:

         RO + NO ^ RONO

         RO + NO ^ RCHO + HNO

It is generally assumed that the alkyl nitrites
rapidly photolyze and, therefore, do not serve as
effective sinks for alkoxy radicals.  The rate of
photolysis, however, is uncertain and merits study.

   Only one determination has been made of the
relative rate of the addition reaction versus the
abstraction reaction to form an aldehyde (or a
ketone) and HNO.   An independent confirmation of
the importance of the abstraction pathway is
recommended.

   (6)  Heterogeneous Reactions

        The assumption is generally made that
heterogeneous processes are unimportant in the
atmosphere.   Little quantitative information is
available to support this statement.  As an
example, the rate of removal  of N205 by water in
the lower troposphere,

         N205 + H20 •* 2HN03

is uncertain.   This heterogeneous reaction could be
significant and deserves additional study.  The
rate of heterogeneous removal of other gaseous
species by aerosols or fog droplets could also be
significant and merits attention.  An estimate of
the dry deposition velocities of various reactive
NO  species is also needed to assess the
significance of such removal  processes in the
lower troposphere.

   In addition to characterizing the role of
heterogeneous processes in the atmosphere, it is
also necessary to determine the degree to which
heterogeneous processes affect the results of smog
chamber studies.   Kinetic mechanisms for photo-
                                                    80

-------
chemical  smog are tested primarily using data
collected in smog chambers and, therefore, it is
important to fully characterize surface effects
and other chamber-related phenomena.   These
phenomena include the heterogeneous formation of
HN03 and  MONO within chambers and the absorption
and desorption of reaction species from the
chamber walls.  One common problem encountered in
modeling  chamber data is that it is difficult to
reproduce the observed initial  rate of hydrocarbon
and NO disappearance.  The very rapid initial
decay of  these species in smog  chambers suggests
that there is a nonhomogeneous  source of free
radicals  present at the onset of irradiation.  It
is possible that radicals may be produced from
contaminants on the chamber walls or they may arise
from the photolysis of nitrous acid.  (There is
evidence to suggest that HONO may form during
loading of smog chambers).  It is important to the
modeling effort to characterize this radical
initiation process.

   Summary:  A number of the key processes  suggest-
ed for study involve organic peroxy  radicals.  It
is not recommended that kinetic studies be  conduct-
ed on every member of each family of reactions.
Only enough members of each class of reactions
should be studied to establish a representative
data base.  This data base should then be used
to generalize rates for the other members of the
series using established thermochemical estimation
techniques.
                                                    81

-------
Session V
Aromatics

-------
                        REACTIONS OF AROMATIC COMPOUNDS  IN THE ATMOSPHERE
                                          Dale G. Hendry

                                 Physical Organic Chemistry Group
                                       Chemistry Laboratory
                                        SRI International
                                  Menlo Park, California  94025
       This paper is a review of the tropospheric chemistry of aromatic  compounds.   The
    reactivity of aromatic compounds is discussed and rate constants  for their  reactions
    with OH are tabulated.  The reaction mechanisms are discussed  in  detail.

    Key words:  Aromatics; free radicals; mechanism;  reactions;  tropospheric chemistry.
               1.  Introduction

   Single-ring aromatic compounds make up a high
proportion of the carbon found in the urban
atmosphere.  Table 1 summarizes data reported on
the composition of the atmosphere in Los Angeles,
California [I]1, and Manhattan, New York [2].
From  these studies we see that benzene and alkyl-


 Table 1.  Atmospheric concentrations  of single-ring
          aromatic  hydrocarbons.

                   Concentration,  ppm  (ppmC)
Los Angeles
(1973)
0.008 (0.048)
0.020 (0.140)
0.020 (0.160)
0.004 (0.032)
0.001 (0.009)
0.005 (0.050)
0.058 (0.439)
0.259 (1.15)
0.087 (0.271)
2.1 (2.1)
0.038 (0.076)
0.404 (1.86)
Manhattan
(1969)
0.0043 (0.
0.013 (0.
0.012 (0.
0.
(0.
(0.
(0.

0.022 (0.
(0.
022)a
094)
094)
083
294)
371 )b
096)

044)
761 )b
 Estimated.

 Not included:  ca. 0.038 ppmC ethane, ca. 0.043
 ppmC propane and ca. 0.022 ppmC  benzene.
 Figures  in brackets  indicate  literature references
 at  the end of  this paper.
substituted benzene compose roughly 24 carbon  per-
cent of the total hydrocarbon in Los Angeles and
about 37 carbon percent in Manhattan.

   The aromatic compounds in the atmosphere come
from gasoline [3], in which they are used to
enhance the octane rating.  Gasoline itself is
composed of 30 to 40 percent aromatic hydrocarbons
and apparoximately 6 to 8 percent toluene.  Hydro-
carbons emitted in automobile exhaust are composed
of 6 to 8 percent toluene.

   We have known for some time from smog chamber
reactivity studies that the alkyl-substituted
benzenes are reactive in promoting oxidation of
NO to NOz and formation of ozone [4].  However,
only in the last few years has an effort been made
to understand specifically how these compounds
react.  This effort has been very productive,
largely because it builds on an existing background
of moderately well understood smog chemistry of
the alkanes and alkenes.  The total conversion of
the aromatics to H20, CO, and C02 is a complex
process, of which we understand only the initial
steps.

   2.  Initial Reactions of Alkylbenzenes

   Table 2 summarizes the possible reactions of
toluene, a representative aromatic hydrocarbon,
with the oxidizing species known to be present in
the atmosphere.  Best values of rate constants and
approximate concentrations are included for estima-
ting the rate of loss of toluene by the various
processes.  The data in table 2 show clearly that
the only important reaction of toluene in the
atmosphere is with OH.  The contribution of the
reactions with 0 atom and 03 are about 10"1* and
10"3 of that of OH reaction.  The reactions of R02'
proceed extremely slowly and can account for only
10"8 of the total consumption of toluene.

   Rate constants for the reaction of OH with
various alky! benzenes are summarized in table 3.
We are fortunate to have several techniques for
measuring the rate constants for reaction of OH
with aromatic hydrocarbons.  The agreement between
                                                  85

-------
  Table 2.  Reactions of toluene in  the atmosphere.

             Concentration           k,_  _   T
  Reactant   molec  cm'3 (ppm)   cm3mo1ec'1s {   to!,  s
    OH
   [5-7]   5.0 x 106  (2 x 1Q-7)  6.4 *  10~12  3.1 x
    [8]    2.5 x 10"  (1 x 10"9)  7.5 x
                                           5.3 x 1Q8
    [9J    1.5 x 1012  (6  x 10"2) 3.4 x 10"22  2.0 x 109

R02-  (H02-)
   [10]    2.5 x io9 (1 x io~")  1.7 x 10~22  2.3 < 1012
   Table 3.   Reported rate constants for reaction of
             OH plus aromatic  hydrocarbon.

                    Perry  Hansen   Doyle   Davis
Compound
Benzene
Tol uene
o-Xylene
m-Xylene
p_-Xylene
1 ,2,3-Mesitylene
1 ,2,4-Mesitylene
1 ,3,5-Mesitylene
J;
1.
6,
14.
24.
15.
33.
40.
62,
LJ_
,20
.40
,3
,0
.3
,3
,0
.4
_I<
1.
5.
15.
23.
12,
26,
33,
47.
LL
.24
.78
,3
,6
.2
,4
,5
.2
_LL
!I
< 3.8
4
12.
23.
12,
23,
33,
52
.2
,8
.2
.3
,0
.0
.0
[5]
1.59
6.11
12.4
20.5
10.5



the reported values is very good, and we have
confidence in these numbers.  Two basic techniques
have been used: measurement of the decay of a
pulse-generated OH concentration by resonance
fluorescence [5-7] and determination of the rate
of disappearance of the hydrocarbon relative to
a standard hydrocarbon under conditions where OH
is the sole reactive species [11].

   3.   Products of OH-Aromatic Reactions

   Two reactive pathways are expected for the
reaction of aromatics with OH radical.  For
toluene, these two pathways are H-atom transfer
from the methyl group (reaction 1) and addition
to the ring (reaction 2).
   From the pressure dependence at 25  °C,  Davis
et al. [5] suggested that ki/(ki + k2) was  less
than 0.5.  Perry et al. [7] found that the
toluene-OH reaction, as well as other  aromatic-OH
reactions, was strongly temperature dependent.   In
fact, at higher temperatures  the  apparent first-
order rate constants were  found to be lower than
the room temperature value  because of the rever-
sibility of reaction (2).   From extrapolation of
the data at high temperature  (where only reaction
1 is important) to lower temperatures, the ratio
ki/(ki + k2) at 25 °C could be estimated.   For
toluene, Perry et al. postulate ki/ki + k2)
0.16 + °-07  The relatively large uncertainty
       0.05.
arises from the uncertainties associated with
extrapolating  the high  temperature data to
determine the  value  of  ki  at  25  °C.

    In our laboratory we have  been investigating
the products of reaction of aromatic hydrocarbons
and OH in a discharge flow system [12,13].  The
products were  collected in  cold traps and on solid
adsorbents.  The product distributions were
determined as  a function of hydrocarbon, N02, and
02 pressures.  Table 4  summarizes some of the data
obtained as a  function  of  N02 pressures.  The
fraction of products resulting from reaction (1)
is a measure of ki/(ki  + k2)  and  remains constant
over the range of conditions.  For toluene we
obtain 0.15 j^0.02,  which  agrees  very well with
the best value reported by  Perry  et al.  [7].   The
ki/(ki + k2) values  for various aromatic hydro-
carbons obtained by  these  two methods are
summarized in  table  5.
                                                           Table 4.  Product distribution  for the reaction
                                                                    of toluene plus  OHa.

                                                                             N02,  10'11* molec/cm3

                                                              Products      0.71    1.04    1.39   1.75
                                                         C6H5CHO

                                                         C6H5CH2OH
  o-HOC6HlfCH3
                  11.7    9.9    9.7    10.4

                   3.5    3.3    4.4    4.6

                  33.3   37.6   39.9    47.3

                  40.3   37.3   35.8    29.0

                   6.4    6.8    6.7    5.5

                   4.3    4.8    3.2    2.9

                  51.0   48.9   45.7    37.4

CH3C6H302           0.4    0.3    0.3    0.7

C6H5CHO +

  C6H5CH2OH/

  Total  products   15.2   13.2   14.1    14.6
  aOxygen:   9.7  x  io16 molec/cm3; toluene:   3  x  IO15
   molec/cm3;  total  pressure:  8.8 Torr


   We find rn-nitrotoluene to  be  a  major product;
however, the concentration  varies  with  the 02/N02
ratio.  Thus  the  intermediate formed in reaction
(2)  appears to  react by two parallel pathways.
                                                    86

-------
                                            (4)
According to  this  mechanism, the relative amounts
of the sum of the  cresols compared with m-nitro-
toluene are:
[N02J

 NO
          [in-nitrotoluene]

             [cresols]
Analyzing  our data  according to this expression
gives  k^/ka = 4  x 103.

   Using this value of  k^/ka, we estimated the
percent yield of m- toluene as a function of N02 in
the atmosphere,  as  shown in table 6.  Thus, at
the N02 concentrations  generally used in smog
chamber experiments, m-nitrotoluene can account
for 1  to 20  percent of  the toluene.  At concentra-
tions  of N02  found  in the atmosphere, however,
significantly less  than 1 percent toluene will
give m-nitrotoluene.
    Table 5.  Reaction of aromatic hydrocarbons
             with OH.
                  jci/i>i + k?)

Hydrocarbon   This work    Perry [7]

Benzene     < 0.05         0.01-0.13

Toluene
             0.15 ± 0.02  0.01-0.23
 p-Xylene
             0.15 ± 0.02  0.04-0.14
 Mesitylene   0.021  ±0.0060.01-0.04
Atmospheric
 products
NO? < 1  ppm

100% Phenol

15% Benzalde-
hyde, 85%
cresol isomers

15% Methyl-
benzaldehyde,
85% 2,5-di-
methylphenol

2% Dimethyl-
benzaldehyde,
98% trimethyl-
phenol
   Table 5 also lists the products expected  to  be
 formed under atmospheric conditions.  These
 estimates are corrected for the N02/02  ratio and
 the high radical concentration.  Thus,  for the
 intermediates formed in reactions  (1) and  (2),  the
 reactions important in the atmosphere are
 reactions (3) and (5).
   In addition to the meta-nitrotoluene  ortho-
and para-isomers have also been reported  in  smog
chamber experiments [14,15].  These isomers
could potentially be formed from the meta-OH-adduct
of toluene in sequences similar to reaction  (4).
However, since very little m-cresol is formed,
this route does not seem reasonable.   In  many
cases, we have observed NO  to be an effective
nitrating agent upon condensing our reaction
                         Table  6.  Calculated yields of m-nitroluene
                                  as a function N02 concentration
                                  in the atmosphere.

                                      N02
                             10"12 cm3molec'1 (ppm)    m-Nitrotoluene, %
1.0
3.0
10.0
30.0
100.0
300.0
(0.04)
(0.12)
(0.40)
(1.2)
(4.0)
(12.0)
0.07
0.2
0.7
2.4
6.6
24.0
                    mixtures, and we believe  that  many nitro products
                    observed in smog chambers  may reflect heterogeneous
                    reactions either during the  actual  chamber
                    reaction or during trapping  out  of the products.
                    Since phenolic compounds  are especially susceptible
                    to heterogeneous nitration,  the  origin of nitro-
                    phenols must be interpreted  with extreme caution.

                              4.   Reactions of Initial  Products

                       Benzaldehyde Reactions.   Two  processes appear
                    to be important for  the reaction of the benzal-
                    dehyde formed from toluene in  the atmosphere:   the
                    reaction with OH and photolysis.

                       Niki et al . [16]  recently reported the rate
                    constant for the reaction
                    as  k    1.3  x  10"11  cm3  molec^s"1, which is
                    identical within  the experimental  uncertainties to
                    rate  constant for other aldehyde-OH reactions.
                    Addition of OH to  the ring  as  observed for toluene
                    (reaction (2))is  expected to occur no faster than
                    addition_to benzene, where  knH   1.2 x 10"12 cm3
                    molec^s l.   Thus  the attack of OH is expected to
                    be  largely  at the  aldehydic position.

                        Using our  discharge  flow systems, we have  found
                    that  the reaction produces  phenol as the only gas
                    phase product [17].  Thus the initial reactions  of
                    the benzaldehyde  with OH is
                                                                   (7)
                                                      followed by the reactions
                                    C<0)0,
                                                                  (10)
                                                    87

-------
In our flow system phenoxy is converted  to  phenol,
presumably by a wall reaction with benzaldehyde
[17].  Considerable amounts of wall products  are
also found, but field ionization mass spectrometry
indicates that the products are largely  various
states of oxidized phenol including as much as
10 percent nitrophenol.

   Niki et al. [18] found o- and £-nitrophenol in
the Cl2-catalyzed reaction of benzaldehyde  in
presence of air and N02.  The benzoyl radical is
first formed by the reactions of Cl atom
                                              (ID
              I'hCHO + Cl	B-l'liCO + HCI            (12)

which  is  followed by  reactions  (9)  and (10).   A
significant  fraction  of the  benzaldehyde appears
as  nitrophenols,  presumably  by  reaction of
phenoxy  radical with  N02.
   The second important reaction of benzaldehyde,
photolysis, has been studied by Berger et al.  [19]
in the gas phase and in the absence of oxygen.
Three photolytic reactions are possible.
                          I'hH + CO

                         • I'll- + HCO

                         Ph<:o + H-
(15)


(16)


(17)
Although  reaction  (15)  is  energetically  favorable
at all wavelengths of the  visible  spectrum,  its
measured  quantum yield  is  significant  only at wave-
lengths less  than  300 nm  [19].   Reaction (16) is
energetically possible  only  below  300  nm.   Thus,
only  reaction (17),  which  has  an energy  cut-off
at 330 nm,  appears to be  important in  the solar
spectrum.   However,  the possibility of the
generation  of a  triplet excited state  that reacts
with  oxygen above  330 nm cannot be ruled out.
                   0,
                       l'hC(0)0,-
Such a  reaction would  be  an  important  source  of
radicals  and therefore critical  for modeling
purposes.

    In both  the  reaction paths  for benzaldehyde,
the  phenoxy radical  is eventually formed.   In the
atmosphere, the fate of phenoxy, we believe,  is
determined  by the  reaction with  oxygen as  well  as
by  reaction (13).
                                               (19)
However,  thermochemical  calculations  indicate  that
the  DH°(C-02)  in  the  resulting  peroxy radical  is
small  and that the  reaction  will  be  reversible.
The  importance of reaction  (19)  therefore  depends
        on how fast the peroxy radical is  trapped  by
        reaction with NO.
                                                      (20)
        The competition between the formation of nitro-
        phenol  (reaction 13) and reaction (20) is
                                          _[NOJ


                                           [NO]
                       (21)
        The following estimates are applicable: ki3 =
        1.7 x 10"12 cm3 molec"1 s"1, k20 = 5.0 x 10~12
        cm molec'1 s"1, ki9 = 6.0 x TO"13 cm3 molec"1 s"1
        and [02]   0.01 fl in air at 1 atm.  If k_19 =  '.
        107 s l which is consistent with our estimate of
        DH°(C-02) = 10 kcal/mol, the two processes will
        compete equally at N02/N0   1.   At very high
        ratios of NOZ/NO, however, the formation of nitro-
        phenol will predominate.  If the DH°(C-02) is much
        weaker than 10 kcal/mol (k 19 » 107 s"1), reaction
        (13) will predominate under most atmospheric
        conditions, but if it is much stronger than
        10 kcal/mol (k 19 » 107 s"1),  reaction (13) will
        be unimportant'and reactions (19) and (20) will
        predominate.   The relative importance of reactions
        (13) and (19)-(20) is very critical because
        reaction (13) is a termination  reaction whereas
        (19)-(20) will  lead to ring degradation and
        further oxidation of NO by reactions of the
        following type.
                                                        (23)
            00-

           HCCHCHO
 00
 II
HCC.H + IICO
                          (25)
       This  reaction  sequence is  speculative, although
       each  step  can  be  justified in most cases by
       analysis of  competing  reactions.   It does suggest
       that  a-dicarbonyl  compounds should be important
       secondary  products.  These compounds absorb light
       very  strongly  in  the solar spectrum and can be
       significant  source of  radicals [20,21].

           Cresol  Reactions.   The  reaction of OH plus
       o-cresol was studied by Perry et al. [22] over the
       temperature  range 300  to 435 K (reactions for £-
       and m-cresols  are expected to be similar).  Rate
       constants  were reported for two processes:  (1) a.
       nonreversible  reaction believed to be hydrogen   ,
       abstraction  (k   2.6 x 10"12 cm3 molec"' s"1) and
       (2) a reversible  reaction  believed to be addition
       to the ring  (k --  3.1 x 10"11 cm3 molec"1 s"1).  We
       postulate  the  following reaction pathways.
                                                    88

-------
                                             (26)
                                             (27)
                                             (28)
                                             (29)
 We are beginning to investigate the  products  of
 reaction of p_-cresol plus OH in our  discharge
 flow system to determine if the above  reaction
 routes are valid.

   Other homoegeneous reactions of cresols  to be
 considered are the reactions with 03 and  0  atoms.
 For £-cresol we have obtained a second-order  rate
 constant for reaction with ozone equal  to about
 1.4 x 10"18 cm3 molec"1 s"1.  At 0.05  ppm 03,
 this reaction is about 1 percent of  the OH-cresol
 reaction, assuming the p_- anc' P_-cresol  have the
 same reactivities.  While the reaction may  prove
 unimportant as a loss mechanism for  cresols,  it
 can be a dominant source of free radicals at  high
 ozone concentrations if it produces  radicals
 efficienctly.  We hope to determine  if this is  the
 case in our studies of cresol-03 reactions.

   Atkinson and Pitts [23] studied the  reaction of
 0 atom plus o-cresol_and found it_to have a rate
constant of 5.8 x 10"13 cm3 molec"1 s"
                                          Since
   Reaction  (26)  should lead to hydroxybenzaldehyde,
reaction  (27)  should  be followed by reactions of
the type  proposed for the  simple phenoxy radical,
and reaction  (28)  will  lead to dihydroxytoluenes.
                                                                                                  (a)
                                                                60     101
                                                                               Time (rain)
                                                       Fig.  1.  (a)  Simulation of SAPRC EC-86:
                                                               Toluene (*   experimental, T = simulation).
                                                                                                  (b)
                                                                               Time (min)
                                                      Fig. 1
the OH reaction is 100 times faster and since  OH
is 100 times more abundant than 0 atom, this
reaction with o-cresol is insignificant.

   5.   Modeling of Toluene Smog Chamber Data

   The Statewide Air Pollution Research Center
(SAPRC) at the University of California,  Riverside     ,
has carried out a series of runs with toluene  in       '
their smog chamber facility.  Concentrations of        |
toluene range from 0.2 to 2.0 ppm while the NO        :
concentration was varied from 0.1 to 1.0  ppm.  We      i
have developed a mechanism to simulate these data.     |
The mechanism includes the standard inorganic          I
reactions and those organic reactions which have
been discussed in previous sections.  We  have  also
included reactions for formation and decomposition
of the major peroxynitrates as well as the termina-
tion reactions of H02- with RO- and R02-  radicals.

   Figures 1 and 2 show simulation and experimental
data for SAPRC Runs EC-77 and EC-86.  Run EC-77
was with 0.28 ppm toluene and 0.58 ppm NO , which      Fig.  1
run EC-86 was with 1.09 ppm toluene, 0.49xppm  NO
and 0.16 ppm formaldehyde.  The agreement betweefi
the simulation and experiment data is very good.
(b) Simulation of SAPRC EC-86:
NO (*   experimental,  I   simulation)  and
N02 (+ = experimental, 2 = simulation).
                                                                 (c)
                                                          33 33333 3333
                                                                                Time (min)
                                                              (c) Simulation of SAPRC EC-77:
                                                              Ozone (*   experimental, 3   simulation)
                                                              and Formaldehyde (+ = experimental,
                                                              F = simulation).
                                                   89

-------
         (d)
     C CCH63 B B 9 B B t
                                                          (c)
                                                                             3  *  FFF
                100     IS
                          Time (min)
Fig.  1.  (d) Simulation of SAPRC EC-77:
        Cresol-total (C   simulation), Dihydroxy
        toluene (D = simulation, and Benzaldehyde
        (B   simulation).
                                                                      100    15
                                                                                 ion    250
                                                                                Time (min)
                                                 Fig.  2.  (c)  Simulation of SAPRC EC-86:
                                                         Ozone (*   experimental, 3   simulation)
                                                         and  Formaldehyde (+   experimental,
                                                         F   simulation).
                                          (a)
                                                          (d)
                         Time (rain)
                                                                                Time (min)
Fig.
2.  (a) Simulation of SAPRC EC-86:                Fig. 2. (d) Simulation of SAPRC EC-86:
   Toluene (*   experimental, T   simulation).           Benzaldehyde (+   experimental,  B  =  simula-
                                                         tion) and PAN (* = experimental, P   sim-
                                                         ulation).
                                          (b)
                           Time  (min)
 Fig.
2. (b)
   NO  i
   N02
Simulation of SAPRC EC-86:
* = experimental, 1 = simulation) and
(+   experimental, 2 = simulation).
There are discrepancies in the  formaldehyde  and
PAN values which in part may be due  to  experimental
uncertainties.  The reason for  the over prediction
in EC-86 of the ozone near its  maximum  is  not
clear.  However, the effect is'  also  seen in  the
simulation of smog chamber data for  other  hydro-
carbons, and thus may or may not be  due to the
actual toluene mechanism.

   6.  Conclusions

   The reaction of toluene in the atmosphere is
very complex.  We have  a good understanding  of the
initial reactions, but  we still need to determine
the fate of the primary and secondary products.
We can model the toluene smog chamber reasonably
well, but out mechanism includes speculation
regarding many of the  intermediates.

   More smog chamber studies  are needed to identify
the products of toluene reaction and their yields.
                                                    90

-------
Currently, data show very low material balances
which may be indicative of the formation of
aerosols or deposition of products on the chamber
walls.  The chamber data should be obtained over
a wide range of conditions because the sensitivity
of individual  reactions varies with the conditions.
Thus, by using a wide range of conditions,
different parts of the model  can be tested.

   The current study of individual reactions of
various intermediates should be continued.  This
work has been  one of the most helpful sources of
information in developing the toluene mechanism.

   Finally, the inability to simulate the ozone
data in smog chamber runs, indicate a need for a
better understanding of the chemistry that controls
the ozone concentration.  Since this effect appears
to be common to the simulation of data for other
hydrocarbons,  the problem may not be solely with
the organic part of the mechanism.
 References

 [1]  Calvert,  J.  G.,  Environ.  Sci.  Tech.  10,  256
      (1976).

 [2]  Lonneman,  W.  A.,  Kopczynski,  S.  L.,  Danley,
      P.  E.,  and Sutterfield,  F.  D.,  Environ.  Sci.
      Tech.  8,  229  (1974).

 [3]  Crabtree,  J.  H.,  private  communication.

 [4]  Heuss,  J.  M.  and  Glasson,  W.  A.,  Environ.
      Sci. Tech.  2.,  1109  (1968).

 [5]  Davis,  D.  D.,  Bellinger,  W.,  and  Fischer,  S.,
      J.  Phys.  Chem.  79_,  293 (1975);  Davis,  D.  D.,
      Investigation  of  Important  Hydroxyl  Radical
      Reactions  in  the  Perturbed  Troposphere,
      EPA-600/3-77-11  (October  1977).

 [6]  Hansen,  D.  A.,  Atkinson,  R.  and  Pitts,
      J.  N.,  Jr.,  J.  Phys.  Chem.  7£,  1763  (1975).

 [7]  Perry,  R.  A.,  Atkinson,  R.  and  Pitts,
      J.  N., Jr., J.  Phys.  Chem.  81_, 296 (1977).

 [8]  Atkinson,  R.  and  Pitts,  Jr.,  J.  N.,  J.  Phys.
      Chem.  79_,  295  (1975).

 [9]  Nakagawa,  T.  U.,  Andrews,  L.  J.,  and Keefer,
      R.  M., J.  Amer. Chem.  Soc.  82_, 269 (1960).

 [10]  Hendry,  D.  G.,  Mill,  T.,  Piszkiewicz,  L.,
      Howard,  J.  A.,  and  Eigenmann,  H.  K., J.  Phys.
      Chem.  Ref.  Data 3,  937 (1974).

 [11]  Doyle, G.  J.,  Lloyd,  A.  C.,  Darnell, K.  R.,
      Winer, A.  M.,  and Pitts,  J.  N., Jr., Environ.
      Sci. Tech. i,  237 (1975).

 [12]  Kenley,  R.  A.,  Davenport,  J.  E.,  and _Hendry,
      D.  G., J.  Phys. Chem.  82,  1095-1096  (1978).

 [13]  Kenley,  R.  A.  and Hendry,  D.  G.,  manuscript
      in  preparation.
[14] O'Brien, R. J., Green, P. J., and Doty, R. A.,
     Interaction of Oxides of Nitrogen with
     Aromatic Hydrocarbons, 175th National
     Meeting, of the American Chemical Society,
     March 1978.

[15] Fitz, D. R., Grosjean, D., Van Cauwenberghe,
     K., and Pitts, J. N., Jr., Photo-oxidation
     Products of Toluene-NO  Mixtures Under
     Simulated Atmospheric Conditions, 175th
     Meeting of the American Chemical Society,
     March 1978.

[16] Niki, H., Maker, P. D., Savage, C. M., and
    ' Breitenbach, L. P., J. Phys. Chem. 82_, 132
     (1978).

[17] Kenley, R. A.,  Lan, B., and Herdry,  D.  G.,
     unpublished data.

[18] Niki, H., Maker, P. F., Savage,  C. M.,  and
     Breitenbach, L.  P., Fourier Transform IR
     Studies of Gaseous and Particulate Nitro-
     geneous Compounds of Atmospheric Interest,
     175th National  Meeting of the American
     Chemical Society, March 1978.

[19] Berger, M., Goldblatt, I. L., and Steel,  C.,
     J. Amer. Chem.  Soc. 95_, 1717 (1973).

[20] Porter, G. B.,  J. Chem.  Phys. 32_, 1587 (1960).

[21] Bouchy, M. and  Andre, J.  C., Molec.  Photochem.
     8, 345 (1977).

[22] Perry, R. A., Atkinson, R., and Pitts,  J.  N.,
     Jr., J. Phys. Chem. 81_, 1607 (1977).

[23] Atkinson, R. and Pitts, J.  N., Jr.,  J.  Phys.
     Chem. 79_, 541 (1975).
              Summary of  Session
   The presentation by Hendry emphasized the
importance of aromatic compounds in the chemistry
of urban air pollution.  Single ring aromatic
compounds account for 25-40 percent of the carbon
species found in urban air.  From chamber studies
these compounds are known to be reactive in the
production of ozone (03).  Therefore a knowledge
of the atmospheric chemistry of simple aromatics
is required for inclusion of these compounds in
tropospheric models to predict their role and
contribution to photochemical smog formation.
The importance of a better understanding of the
chemistry was illustrated in comments by Atkinson
and Hendry on the uniqueness of the 03 formation
curve and the current inability to simulate 03
smog chamber data.

   The major theme of the discussion and the
majority of the uncertainties centered around
mechanisms of reactions of primary and secondary
aromatics in the atmosphere.  There was general
agreement that the initial reaction can be account-
ed for almost solely by attack of the hydroxyl  (OH)
radical.  For methyl substituted benzenes, the
accepted mechanisms are hydrogen abstraction at
the methyl group and OH addition at the ortho
position.  However, there was a degree of
                                                    91

-------
uncertainty regarding the branching ratio for the
two pathways, with the only significant work being
done on toluene.  In addition Atkinson pointed out
the thermodynamic favorability of OH addition at
the methyl position.  Inclusion of this pathway
could alter mechanistic consideration of other
investigators.

   There was considerable discussion and some lack
of agreement on product yields.  For toluene,
O'Brien reported lower product yields than Hendry
for benzaldehyde and o-cresol by factors of
approximately 4 and 10 respectively.  There was
some speculation that Hendry's results were higher
because he based his yields on the total amount of
gas-phase carbon analyzed.  In any event none of
the investigators have yet analyzed aerosol carbon
or carbon on the walls of reaction chambers.  Both
O'Brien and Atkinson reported carbon balances well
below 100 percent.

   The question of the mechanism(s) of ring opening
was raised several times.  Tentative mechanisms
were proposed in the papers by Hendry and
Atkinson.  Evidence for ring opening was given by
the observation of peroxyacetyl nitrate and carbon
monoxide by Atkinson.  Ring opening could be of
considerable importance as a source of free
radicals and simple oxgenated products.

   The status of the uncertainty on reaction
mechanisms may be illustrated by the fact that the
only work reported on basic mechanisms was the
low pressure flow tube studies of Hendry.   Most
of the mechanistic work reported was on toluene.
The reactions and fate of aromatic products was
largely unconsidered.  In his paper Hendry
discussed the attack of OH on benzaldehyde and the
cresols and subsequent reaction pathways.

   With regard to reaction kinetics, relatively
good agreement was reported for the OH reaction
with the primary aromatics-benzene, toluene and
xylenes.  Little data are available on the higher
homologues or on OH reaction with aromatic
products.  O'Brien reported ratio measurements,
relative to toluene, for o-cresol and benzaldehyde.
Atkinson presented reaction rate data for the
cresols.

   It was evident from the discussion that some
problems exist with regard to analytical  measure-
ments of products.  All of the analyses reported
during the discussions were performed by gas
chromatography (GC).  O'Brien reported some
difficulty with some product measurements at low
concentrations, e.g. cresols.  No analyses were
reported by other techniques such as mass
spectroscopy or Fourier transform infrared
spectroscopy.  Either of these techniques could
give better time resolution and the possibility
of observing intermediates.  Finally there is the
larger question of the amount and nature of
products in the aerosol phase.
                    Comments

 Roger Atkinson, Statewide Air Pollution Research
 Center, University of California, Riverside,
 California  92521

    I would like to make three points:

    1)  Besides the two initial  reactions of the OH
 radical with the substituted aromatics (taking
 toluene as an example),
 OH
                                    (ii)
    OH radical  addition at the 1-position leads  to
 the formation  of phenol  and CH3  radicals
OH 4-
                                        + CH,
   Elimination of CHa from radical III can be
calculated to be ^9 kcal mol"1 endothermic.  This
together with an activation energy for the
addition of CH3 radicals to toluene of ^4 kcal
mol"1 [1], leads to an activation energy of
"" 13 kcal mol"1 for reaction (3).  Hence reaction
(3) will be favored over elimination of an OH
radical (analogous to reaction (1)) from this
OH-toluene adduct.  The occurrence of this
reaction pathway would hence mean that the values
of ki and k1/(k1 + k2) obtained by Perry, Atkinson
and Pitts [2,3] are upper limits.  This may be
especially true for o-xylene where, by analogy
with the 0(3P) atom reaction [4], OH radical
addition at the methyl substituted positions is
likely to be appreciable, and for which the report-
ed value of ki/(ki + k2) appears to be high, with
a low value of Ei6, compared to the other
aromatic hydrocarbons.

   2)  At the Statewide Air Pollution Research
Center, University of California, Riverside, we
[5] have recently determined rate constants for
the reaction of OH radicals with o-, m- and
p-cresol from the rates of disappearance of the
cresols and n-butane in irradiated NO -organic-air
mixtures of atmospheric pressure and 300 ± 1 K.
Using a value of k(OH + n-butane) of 2.73 x 10"12
cm3 molec ' s l at 300 K [6] rate constants k
(cm3 molec l s ') of (4.7 ± 0.4) x 10"12;
(6.7 ± 0.7) x 10 12 and (5.2 ± 0.5) x 10"12 were
obtained [5] for o-cresol, m-cresol and p-cresol.
Further experiments [7] have shown that the NO
photooxidations of the cresols form hydroxy-  x
nitroluenes as the major observed gas phase
                                                  92

-------
aromatic products, the 2,3- and 2,5-isomers from
o-cresol, the 3,2- and 3,4-isomers from m-cresol,
and the 4,3-isomer from p-cresol.

   3)   Ring  cleavage in the OH-aromatic systems
may occur,  in part,  through the reaction sequence
   (ii)
                                    Robert  J.  O'Brien,  Patrick  J.  Green,  and Richard
                                    M.  Doty,  Department of  Chemistry,  Portland State
                                    University,  Portland, Oregon   97207

                                       The  product  formation  and  dynamics for the
                                    toluene (T)  system  under  ambient conditions are of
                                    great current interest.   For  the simple case where
                                    hydroxyl  radical determines both formation and
                                    loss of "stable" products (P.j):
                                          HO,
                                                               T + OH
                                                               Pi  + OH
                                                             loss
                                   the maximum production concentration  will  be given
                                   by
                              (in)

 Radical (III)  may then  react with  NO  to  form  N02and
                                             [Pmax]

                                              [T]
                                            o:
the radical
                                   where a. is the yield of product P.  in  the  primary
                                   reaction step and R. = t>./a is the ratio  of the OH
H t which would probably undergo   rate constant with the product, P.,  to  the  rate
ring opening,  leading to a variety of oxygenated
species.   H  atom abstraction from radical  (II) by
Oa to form o-cresol is ^ 26 kcal mol"1 exothermic,
with the  bond  strength of the C-H bond at which
abstraction  takes place being ^ 20 kcal mol"1.

   From group  additivity calculations, using a
CQQ-H bond energy of 90 kcal mol"1 (the same as
that for  HOO-H, AH.(III) -v -6 kcal mol"1.  As
AH.(II) ^ 1  kcal mil"1; formation of (III) from
(II) is i* 7  kcal mol"1 exothermic and hence
radicals  (III) and (II) will be in equilibrium.
Hence this reaction pathway leading to ring
opening is expected to become more important at
lower temperature, and vice-versa.

References

[1]  Cher, M., Hollingsworth, C. S., and Sicilio,
     F.,  J.  Phys. Chem.. 70., 877 (1966).

[2]  Perry,  R. A., Atkinson, R., and Pitts, J. N.,
     Jr.,  J. Phys. Chem.. 81_, 296 (1978).

[3]  Perry,  R. A., Atkinson, R., and Pitts, J. N.,
     Jr.,  J. Phys. Chem.. 81_> 1607 (1978).

[4]  Grovenstein, E., Jr. and Mosher, A. J.,
     J. Amer.  Chem. Soc.. 92, 3810 (1970).

[5]  Atkinson, R., Darnall, K.  R., and Pitts,
     J. N.,  Jr., J. Phys. Chem., submitted for
     publication (1978).

[6]  Perry,  R. A., Atkinson, R., and Pitts, J. N.,
     Jr.,  J. Chem. Phys.. 64, 5314  (1976).

[7]  Darnall,  K. R., Atkinson,  R., Glangetas,  A.,
     Winer,  A. M., and Pitts, J. N., Jr.,
     unpublished data  (1978).
constant with toluene.

   If we assume pseudo first order  loss  for
toluene (constant OH concentration) we may
integrate the rate expression to  determine the
length of time required to achieve  maximum

concentration (tTax) in terms of  the  toluene
lifetime
                                                   •   I III UCI MO \J I
                                                   This is given by
                                                      (Rri;
                                                              In Ri
                                            (2)
                                   A plot of this function is given in figure  1.

                                        100|
                                                      Fig.  1.   Variation of the time for a product to
                                                               reach maximum concentration relative to
                                                               the  toluene  lifetime (tmax/TT) with the
                                                               relative  reactivity with OH radical (R.)
                                                               as given  by  eq.  (2).                   "•
                                                   93

-------
     From  this  plot we may  determine  (for example)
  that  for a  typical  atmospheric  lifetime of
  toluene  of  10 hours the concentration  of ^cresol
  (which reacts six times faster  than  toluene with
  OH) should  reach a  maximum  in 3.5 hours.   If  the
  yield of ^-cresol is 5 percent  (see  below) and
  the ambient toluene concentration is .020  ppm,
  this  maximum  concentration  calculated  from eq.  (1)
  is about 0.2  ppb.

    Assuming pseudo  first  order  toluene  loss the
  variation of  a product with toluene  concentration
  is given by
     Pi
R-l
                               (3)
    The general form of this equation for various
 values of R. is shown in figure 2.
    80 r-
  a.
  a.
  1
  o

  .1
  o
  o
                         R=0.1
                      Toluene ppm

 Fig.  2.   Variation  of a  product  concentration with
          toluene  concentration for  different values
          of R., assuming a  =  10  percent.


   The values of a- may  be  obtained from the
 individual  rate measurements  of Atkinson and
 others.   However, considering the errors present
 in each measurement a separate measurement of the
 ratio itself may be preferable.  We have made
 such measurements by irradiating a mixture of P.
 and toluene at about a 10 to 1 ratio (P./T).  A
 plot of In P. vs.  In T gives the value of R. as
 the slope.  This analysis is not sufficient if
 the product photolyzes to any appreciable extent.
 For the case of o-cresol  the photolytic lifetime
 in our reaction vessel is 101* minutes and is
 probably negligible.

   For the more general  case of non  first order
 toluene loss (variable OH concentration)  we may
 still  derive an expression  to analyze production
 formation and loss.   For  the same mechanism given
above  (reactions a and b) it can  be  shown that
                                           A[P.J  + R.
                                                       To  [T]
                        d[T] -'o1 A[T]
(4)
   For the simple case where a product is tptally
unreactive, R.   0 and a plot of Pi vs. T will
give a straight line with slope = a.  For the
case where the product does react further, the
second term on the left hand side of eq. (4)
corrects for this loss of product.  The variation
of jij-cresol and of benzaldehyde for one of our
experiments are shown in figures 3 and 4.  This
experiment was carried out by irradiating toluene
and N02 each at about 4 ppm in a 250 L evacuable
glass vessel with a mixture of fluorescent black
lights and sun lamps.

   The yield of each product may be determined
from the slope of these plots.  For benzaldehyde
we obtain a 2.5 percent yield and for o-cresol a
5 percent yield.  These yields are much lower
than those measured by Hendry in his low pressure
flow system.

   We have been initially skeptical of our. low
yields, especially for o-cresol  since it is about
ten times lower than the yield reported by Hendry.
To double check this result we have carried out
experiments in which we start with a mixture of
toluene and ^-cresol (4 ppm and 1 ppm) respective-
ly).  The decay of jg-cresol is then modified by
formation of o^-cresol from toluene.  Equation (3)
holds for any initial product concentration so we
have plotted the data for this experiment in the
required form in figure 5.  The yield of jg-cresol
is found from the slope to be 5 percent, in
agreement with the other experiments.  This
experiment has the advantage of generating a large,
                                                          160 i-
                                                                        Slope=.05
                                          a
                                          a.
                                          a.

                                          8
                                          0)
                                                       1000    2000
                                                                       3000
                                       4000    5000  I
                                                               Toluene ppb
                                        Fig.  3.
        Analysis of cresol formation from  toluene.
        Lower data points are a plot of o-cresol
        vs. toluene.  Upper data points with  line
        are a plot of (C) + 6 / (C)/(T) d(T)  vs.
        (T) in accordance with eq.  (s).  (T =
        toluene; C - jj-cresol).
                                                   94

-------
 a
 a.
 a
 o
 •a
2
S
     601-
     50 -
     40 -
     30
     20
     10
                                Slope = .025
              1000
                      2000     3000

                      Toluene ppb
                                      4000
  5000
Fig.  4.
         Analysis of benzaldehyde formation from
         toluene.  Lower data points are a plot of
         benzaldehyde vs. toluene.  Upper data
         points with straight line are a plot of
         (B) + 2.3 / (B)/(T) d(T) in accordance
         with eq. (3).   (T = toluene; B = benz-
         aldehyde) .
Q.
Q.

O
in
                  8
                         9
                                10
                                      11
12
                  Toluene ppm
Fig.  5.
         Analysis of competitive disappearance of
         a mixture of 11  ppm toluene + 2 ppm
         o-cresol.  Data  plotted as (C) + 6 / (C)/
         0") d(T) in accordance with eq. (3).
         (T   toluene; C    o-cresol).

 unambiguous, initial cresol peak with the gas
 chromatograph.

   We currently have no explanation for the lower
 product yields for toluene but the most obvious
 explanation would lie in  the difference in
 pressure between the two  measurements.  Our yields
of nitrotoluene are  in agreement with  those ob-
                                                      tained from Hendry's work and would be negligible
                                                      at sub ppm N02 concentrations.  Our mass balance
                                                      for gaseous toluene products is then well below
                                                      100 percent.  The carbon balance may possibly be
                                                      accounted for by aerosol analysis which we will do
                                                      in the future.
         A.  R.  Ravishankara,  Applied Science Laboratories,
         EES,  Georgia Institute of Technology,  Atlanta,
         Georgia  30332

           We have noticed a decrease in the net rate constant
         for m-xylene  (drops at  lower pressures e.g. '^ 3 Torr
         of Ar).   In Hendry's experiments there could be an
         overemphasis  on the abstraction route since his flow
         tube pressures are not  very high.
             Recommendations
                                                     Importance
   Study of the atmospheric reaction processes of
aromatic hydrocarbons is in its early stages.   Our
current knowledge about these compounds is rather
primitive compared to the alkanes and alkenes.
However, aromatics are major components of urban
atmospheres and elucidation of their reaction
pathways is essential for the following reasons.

   1)  Oxidant formation.  As major urban hydro-
carbons which react relatively rapidly with
hydroxyl radical, aromatics will  contribute
directly to ozone formation and buildup and they
seem to generate appreciable quantities of PAN,
itself a harmful oxidant.

   2)  Direct health effects.  Oxidant products  of
aromatic hydrocarbons are poorly characterized
and present a potential health hazard of
undetermined magnitude.

   3)  Aerosol formation.  Gas phase mass balances
for smog chamber experiments with aromatics are
very poor and may indicate appreciable aerosol
formation.  If so, the aerosol so formed may
contribute to a heterogeneous component of tropo-
spheric chemistry which is currently unrecognized;
this heterogeneous component may well impact other
areas   in particular NO  conversion to nitric
acid or free radical loss processes.

Current Status
           1)  Rate constants.  Considerable work has been
        done to determine the reaction rate constants for
        hydroxyl radicals with the chief aromatic
        constituents of the atmosphere.  Agreement between
        various groups is quite good so this question is
        resolved.  Ozone and other free radicals (HOz,
        N03) are known to react slowly with aromatics and
        are therefore, at present of minor importance.
        Relative rates of ring addition versus side chain
        abstraction, while less certain than the overall
        rates, are also fairly well settled.

           Rate constants for reaction of OH with some of
        the more important reaction products of aromatic
        hydrocarbons (cresols, benzaldehyde, etc.) have
                                                  95

-------
 also  been measured.

    2)   Product  identification.  Major  products  of
 the reaction  of OH with  toluene which  have  been
 identified  are  the following:  cresols,  nitro-
 toluenes, benzaldehyde,  benzyl alcohol,  benzyl
 nitrate, peroxybenzoyl nitrate, peroxyacetyl
 nitrate, and  carbon monoxide.  Yields  of these
 compounds have  been determined at low  pressure  and
 are becoming  available at atmospheric  pressure  as
 well.   Currently the low pressure yields are
 considerably  higher on an absolute basis than
 those at high pressure, but on a relative basis
 are in good agreement.  Products have  been
 identified  for  the reactions of some other
 aromatics as well.

   Products  of the subsequent reactions of the
primary products are  known in a few  cases.

   The hydroxyl  radical  is a key  component in
controlling  loss rates of the primary products but
other  processes, such  as  direct photolysis or
 reaction with 03, R02, RO, N03, etc.  may be
 important as well.

   3)   Ozone formation.   Only a limited amount of
work has been done on  modeling the ozone profiles
 in aromatic/NO  or mixed  hydrocarbon/NO  systems
which  include aromatics,  because  of  the general
 lack of knowledge about the detailed reactions
 involved.  However, the limited work done to date
 indicates that ozone  profiles are different than
 those  in nonaromatic  systems and  in  some cases are
 difficult to model  unless unique  radical-radical
 reactions are invoked.

   4)   Analytical techniques.   Current studies of
 aromatic hydrocarbon  systems are  severely hampered
 by a lack of versatile techniques  for analyzing
 the high molecular weight products involved.
 Techniques  which have  been employed  include gas
 chromatography, gas chromatography-mass spectro-
metry and to a limited extent Fourier transform
 infrared spectrometry.  These techniques are
 difficult to employ when  they are successful, and
 are often unsuccessful.   Much time has been spent
 in adapting these techniques for use in the study
 of aromatic hydrocarbons, but they still suffer
 from some inherent problems.

 Recommendations

   1)   Absolute yields of the major known primary
 products of aromatic-OH reactions should be
 determined at atmospheric pressure.   These
 aromatics would include as a minimal set benzene,
 toluene, the xylenes,  trimethyl benzene and some
 alkyl  benzenes such as ethyl benzene.

   2)   Rate constants  for the various processes
 these products  undergo should be determined.
Although a large number of compounds are  involved,
competitive kinetic studies employing several
compounds simultaneously may suffice in some cases.
This would reduce the total number of necessary
experiments.

   3)  A carbon mass balance for the gaseous
products including CO and C02 should be obtained
for the major aromatics.  The mass balance should
include the carbon content of any aerosol formed.

   4)  New analytical techniques should be investi-
gated for application to the study of aromatics.
These techniques would be doubly useful because
they would be equally applicable to the study of
higher moleculas weight alkanes and alkenes.
Techniques which might be investigated include
improvement of gc sampling techniques and separa-
tion efficiency on the column, direct mass spectral
analysis employing non-framentation ionization,
liquid chromatography, and field desorption mass
spectrometry.

    In all these techniques every attempt  should
be made to work at  realistic reactant concentra-
tions and total pressures and to induce minimal
sample alteration.   However, some low pressure
techniques may have  to  be employed  (e.g., direct
ms  sampling)  because  of the lack of  any other
viable alternatives  for direct  analysis of inter-
mediates.  The current  advancement of knowledge
in  this area  is now  limited by  analytical
methodology.  Advancement of knowledge in the
alkane and alkene systems will  soon  suffer the
same  fate, as the chemistry of  the low molecular
weight compounds becomes worked out, and  higher
members of the series are studied.

    5)  Heterogeneous  processes  may be of  great
importance in the aromatic hydrocarbon systems.
The  impact of these  processes may well extend
beyond the purely aromatic systems and influence
the  chemistry of NO   and of free radicals
generated from othe$  classes of compounds.  An
attempt should be made  to assess the significance
of  these  processes  on the overall chemistry of
the  troposphere.

    6)  Compouter modeling of the aromatic hydro-
carbon system should  be continued in order to
assess the ozone forming potential of these
hydrocarbons.   It is  expected that these  modeling
efforts will  become  more meaningful  as more
basic rate and product  data become available.

    Recommendations  1  to 3 may be expected to be
completed with current  funding  in the  next year
or  two.   Recommendations 4 and  5 are much more
ambitious and will  require a  long term committment
and considerable additional funds for  instrument
development.
                                                   96

-------
 Session VI
SOX Chemistry

-------
                                             SO  CHEMISTRY
                                            Jack G.  Calvert
                                          Chemistry  Department
                                       The Ohio  State  University
                                         Columbus, Ohio   43210
         An  evaluation has been made of the existing kinetic data related to the elementary,
      homogeneous reactions of S02 within the troposphere.  A set of recommended values of the rate
      constants for these reactions is presented.  The results show that the direct photooxidation
      of S02 by way of the electronically excited states of S02 is relatively unimportant for most
      conditions which occur within the troposphere.  The oxidation of S02 within the natural
      troposphere is expected to occur largely by way of reactions 39, 31, and 33, with reaction  39
      being  the dominant path:  HO + S02 (+M) ->• HOS02 (+M) (39); H02 + S02 -> HO + S03 (31); CH302 +
      S02 ->  CH30 + S03  (33).   For certain  special  conditions  within  the troposphere the  oxidation
      of S02 by way of the products of the ozone-olefin  reaction may be significant.   Also the
      reaction of 0(3P) with  S02  may contribute  somewhat to  the  S02  removal  in  N02-rich, 02-deficient
      stack  gases in sunlight during the early stages of dilution of the plume.

         The complete paper upon  which most of the talk  was  based is:  "Mechanism of the  Homogeneous
      Oxidation of Sulfur Dioxide in the Troposphere", by Jack G.  Calvert,  Fu,  Su,  Jan W.  Bottenheim,
      and Otto P. Strausz which appeared in the  Preprint Volume  I,  Plenary  Papers,  International
      Symposium on Sulfur in  the  Atmosphere,  July, 1977,  and  which  was  presented at the  Symposium on
      September 7-14, 1977, Dubrovnik, Yugoslavia.  The  paper has been  published in Atmospheric
      Environment, 1_2,  197 (1978).

      Key words:   Kinetics; photochemistry;  review;  sulfur dioxide;  troposphere.
            Summary of Session
   The discussion  was  concerned  with  two  basic
 problems - what is the mechanism of conversion of
 S02 in the atmosphere, and  what  do we know  about
 aerosol formation  arising from SO and  NO
 reactions.  There  appeared  to be a consensus  that
 the reaction of S02 with OH is the most important
 homogeneous mechanism, but  there is still great
 interest in quantifying the role of H02,  R02, and
 the Criegee intermediate  in this process.   There
 is considerable experimental  work underway  on
 aerosol formation, the incorporation  of NO   in
 aerosols, and the  role of  specific radicals in
 aerosol formation.

   Whitten opened  the  discussion with a descrip-
 tions of his modeling  results for the Los Angeles
 basin.  He pointed out that while OH  levels
 could be reduced by reducing hydrocarbon or NO
 levels, they were  relatively constant for a given
 HC/NO  ratio.  Thus if the  reaction of OH with
 S02 controls the S02 level, control of the  HC or
NO  levels will not necessarily  have  any affect
on the OH levels if the HC/NOY  ratio  remains
fixed.  A similar conclusion was reached in the
case of H02 radical levels, although the new
value for the rate of NOX+ H02 means that H02
levels are reduced to the point where the
H02 + S02 reaction is probably unimportant.

   Whitten also reported modeling studies which
suggest that a fairly rapid conversion of S02 to
sulfate takes place in fog droplets.  The observa-
tions used in this study were based on one days
sulfate collections from 14 monitor stations  (24
hour averages).  A photochemical model was not
sufficient to account for the S02 conversion  rate.

   Miller reported on smog chamber results which
supported  Whittens observations.  He  also  discussed
measurements  which show  that  the  rate constant  for
H02  + S02  is  no  greater  than  1  ppm"1  min"1 and  for
CH302 +  S02  is no  greater  than  2  ppnr1 min"1.   The
role of  NOX  in aerosol formation  was  also  discussed.

   Heicklen  suggested that excited molecules  such
as N02,  formed by  irradiation at  wavelengths  above
400  nm,  might react with S02.   Also photo  excited
aldehydes  and ketones might  be  important.  Cox
reported on  the  photolysis of MONO  in the  presence
of S02.   The end product was an aerosol.   With
                                                   99

-------
added NO  some of the NO  was incorporated in the
aerosol ,xbut since NO  Cctn lead to aerosol forma-
tion without S02 being" present, it was difficult
to say how much of the lost NOX was incorporated.

   Jeffries reported smog chamber results on the
effect of added CO on S02 conversion rates in
natural background air, which indicate that OH is
the only important oxidizing species for S02.
Ravishankara noted that aerosols are readily
formed in uv flashed H20-S02 mixtures, probably
by photolysis of H20.  Huie raised the question
of the reactivity of the Criegee intermediate with
respect to S02, pointing out that if the Criegee
intermediate isomerized to a dioxirane intermediate
it probably decomposes through a "hot" acid or
ester before it has time to react with SOa.
However, Calvert noted, one would expect for
larger olefins an increasing chance of stabilizing
the Criegee intermediate and observing some direct
chemistry.

   Niki noted that generating H02 by reacting Cl
with H2 in the presence of 02 and S02 led to H202
but no conversion of the S02.  If NO was added
(to drive the H02 to OH), then sulfuric acid
aerosol is formed.  In the case of CH302 some
conversion of S02 is observed and there is
evidence to suggest that sulfones are possibly
formed.  Jeffries asked if Niki has any evidence
that water intercepted the Criegee intermediate to
make acetic acid.  Niki has not yet carried out
an experiment with added H20.  Jeffries noted that
in earlier work they had not seen acetic but had
seen formic acid in the reaction of ozone with
propylene.
rate  for CH302 + S02  is  not  greater  than  2  pprrf1
min  '

  .When we  irradiate  just  S02 with nitrous  acid we
observe an  NO  loss in excess of  that  for the
experiment  without S02.  The amount  of the  NO  loss
corresponds to the amount  of H2SOi, formed.   We
suspected that either NO or  N02 might  be  incorporat-
ed in the aerosol phase.   It has  been  suggested,
for example, that HSCU might react rapidly  with
N02 to give aerosol mixtures of sulfuric  and nitric
acids.  However recent chemical analyses  of filter
collections from such reactions show very low
nitrate levels relative  to sulfate.  Thus,  if such
reactions occur, the  nitric  acid  apparently ends  up
in the gas  phase.

   The experiments were conducted with  the  following
concentration ranges: S02  (400-500 ppb),  HN02 (100-
200 ppb), NO and N02  (20-200 ppb), CO  (2-200 ppm)
and CHi, (200 to 900 ppm).
Richard A.  Cox, U.K.A.E.A., Environmental and
Medical Sciences, A.E.R.E, Harwell, Oxfordshire
0X12 ORA, England

   As  part of an extensive research program on
kinetics of reactions of S02 in the atmosphere,
we have  used the photolysis of HONO to produce
OH radicals and allowed them to react with S02
at 25  °C and 1 atm  pressure in air.  We have
drawn  the following conclusions:

     (a) Reaction of OH with S02 occurs with a
                   Comments
 David  F. Miller, Battelle-Columbus Laboratories,
 Columbus,  Ohio  43201

   Gary Whitten presented modeling results relating
OH concentrations to initial concentrations of
NMHC (nonmethane hydrocarbons) and NO .   According
to his model, OH concentrations are predicted to
be nearly  constant for any NMHC/NO  ratio.  I'd
like to add that our smog chemaber results,
reposted last year in Dubrovnik, led to the same
conclusion.  This finding has a very important
implication regarding precursor controls designed
for limiting ozone.  Because S02 competes with
NMHC and NO  for OH, proportional control of NMHC
and NO  cou'fd result in an increase in the conver-
sion of S02 to sulfate.

   Secondly, I'd like to comment on some smog
chamber experiments in which we've irradiated
mixtures of nitrous acid with S02 and either CO ur
 CHi, to estimate the S02 oxidation rates attribut-
able to H02 and CH302.  Although our analyses of
 the data are less than satisfactory, primarily
 because of so much uncertainty about the nitrogen
 oxides chemistry, we estimate that the  rate  for
 H02 +  S02  is not greater than 1  ppm"1 min  1  and the
     (b) This reaction proceeds by addition to
give HOS02 and the subsequent radical chemistry
leads to a short chain reaction in which NO is
oxidized to N02.  This chain reaction is
inhibited by N02.

     (c) The final reaction product is an aerosol.
physically resembling model H2SO^ aerosols.

     (d) In the N02   inhibited system, the
aerosol contains   SO -NO  species yielding
equimolar proportions of   SOi/~ and N03  on
hydrolysis in water.
 Harvey  E. Jeffries, Department of  Environmental
 Science and  Engineering, University of  North
 Carolina, Chapel Hill, North Carolina   27514

    In a series of experiments performed  in  UNC's
 outdoor aerosol chamber, S02 (at % 0.3  ppm)
 oxidation in natural background air (<  20 ppb
 NO   < 50 ppb C organics) was followed  by observ-
 ing* aerosol number  (by CN), aerosol volume  (by
 EAA), and aerosol sulfur content (by XRF analysis
 of  filters).  Runs  were repeated with various
 additional amounts  of CO added (5, 10,  15,  25
 ppm).   The  additional CO  resulted  in  delays in
 time of CN  peak, small  increases  in 03  produced,
 and reduced  aerosol volume  (by both EAA +  XRF).
                                                   100

-------
In one run, no CO was added initially, but when
a steady rate of increase in aerosol  volume had
been established, 25 ppm of CO was injected;
aerosol  volume production (i.e. growth) was
totally  stopped within 4 minutes.   CO's role  in
this otherwise low concentration system is to
convert  OH to H02.   It seems clear that OH was
by far the major oxidizing species.   It is
expected that under urban conditions,  however,
(i.e.  higher NO concentrations) the effects of
CO would not be observed because the  higher NO
converts H02 to OH.
A. R. Ravishankara,  Applied Science Laboratories,
EES, Georgia Institute of Technology,  Atlanta,
Georgia  30332

   We have noticed formation of aerosols directly
in our system when ^ 300 mTorr of H20 and S02 are
photolyzed.  This mechanism could be important
where water concentrations are high i.e., very
quick oxidation of sulfur dioxide leading directly
to aerosol.  (Even a mixture of S02, 03 and H20
gives aerosols).   The water concentration needed
to get this aerosol  formation seemed rather
magical   aerosols formed only after a critical
amount of water was  present.
               Recommendations
   It is  recommended that kinetic and chemical
data regarding S02 chemistry in the troposphere
be obtained.  The classes of reactions are of six
types, with the first four of these being almost
equally important.

   1.  Of most importance is obtaining both
product and rate information of HO S02 and RO S02
with H20, NO  , 02, hydrocarbons, NR3, and comBina-
tions of these gases.

   2.  Of essentially equal importance is obtaining
information on the fate of S02 in 03-olefin-02
reactions.  There are three subsections of this
 problem which  should  be  attacked  in  the  following
 order:

       a)  characterize  the intermediates which
           which react with S02
       b)  obtain products and rate  coefficients
           for the reactions of these intermediates
           with S02
       c)  study the effect of adducts such as
           H20, NO , hydrocarbons, NH3,  and
           combinations  of these  gases.

   3.  More data is needed on the rate coefficients
 (and products) for the reactions  of  H02, HO, and
 0(3P) with S02.  These data should include
 pressure, temperature, and humidity  studies.  In
 the case of H02, there is a large uncertainty in
 the rate constant.  With regard to HO and 0( P),
 fairly reliable values exist.  However because of
 the importance of the HO radical, which  appears_to
 be the most important species for S02 removal, it
 is important to have as  accurate  a rate  coefficient
 as possible.

   4.  The reactions of  R02 radicals with SOZ
 should be investigated to determine  products and
 rate coefficients at a variety of pressures,
 temperatures, and humidities, and in the presence
 of NO , 02, NH3 and hydrocarbons.  The reactions
 of R0xradicals with S02  appear to be unimportant
 in the troposphere, and we do not give a high
 priority to their study.  However it would be
 useful to actually have  rate coefficients for RO
 reactions with S02 to know exactly what  role
 these reactions do play.

   5.  A low priority recommendation is  the study
of the possible reaction of electronically
excited N02 with S02.   There is no evidence that
a reaction occurs, but this should be confirmed.

   6.  The direct photoexcitation of S02 is not
important in the removal of S02 in the troposphere,
and we do not recommend  studies in this  area.
However we do point out  that such reactions may be
 important in the formation of sulfur-containing
organic aerosols.   If so then such reactions could
be of significance in aerosol chemistry.
                                                  101

-------
                              Workshop  Attendees
Larry G. Anderson
General Motors Research Labs.
Environmental Science Department
Warren, MI  48090

Paul Ascher
Northrop Services, Inc.
P. 0. Box 12313
Research Triangle Park,  NC  27709

Roger Atkinson
Statewide Air Pollution  Research Center
University of California
Riverside, CA  92521

N. Basco
University of British Columbia
Chemistry Department
Vancouver V6T 1W5
British Columbia, Canada

L. Batt
Chemistry Department
University of Aberdeen
Meston Walk
Aberdeen, Scotland  AB9  2UE

Sidney W. Benson
Chemistry Department
University of Southern California
Los Angeles,  CA  90007

Jack G.  Calvert
The Ohio State University
Chemistry Department
140 W.  18th Avenue
Columbus, OH   43210

William P. L. Carter
Statewide Air Pollution  Research  Center
University of California
Riverside, CA  92521

C. Richard Cothern
EPA
1705 N.  Stafford  Street
Arlington, VA  22207

Richard A. Cox
U.K.A.E.A.
Environmental and Medical  Sciences
     Division, A.E.R.E.
Harwell, Oxfordshire 0X12 ORA
England

R. J. Cvetanovic
National  Research Council of Canada
Division  of  Chemistry
Ottawa,  Ontario, Canada K1A OR6
 Kenneth  L.  Demerjian
 Environmental  Protection  Agency
 NCHS-C Room 320-B
 Davis Drive
 Research Triangle  Park, NC   27711

 William  B.  DeMore
 Jet Propulsion Laboratory
 4800 Oak Grove Drive
 Pasadena, CA  91103

 Marcia C. Dodge
 Environmental Protection  Agency
 MD-84
 Research Triangle  Park, NC   27711

 William  H.  Duewer
 Lawrence Livermore Laboratory  (L-262)
 P. 0. Box 808
 Livermore,  CA  94550

 A. Fontijn
 Aerochem Research  Lab., Inc.
 P. 0. Box 12
 Princeton, NJ  08540

 David Garvin
National  Bureau of Standards
Washington, DC  20234

Lewis H.  Gevantman
National  Bureau of Standards
Office of Standard Reference Data
Washington, DC  20234

 David Mark Golden
SRI International
333 Ravenswood Drive
Menlo Park, CA  94025

T. E. Graedel
 Bell  Laboratories
 Room 1D-349
 Murray Hill, NJ  07974

 David Gutman
 Illinois Institute of Technology
 Department  of Chemistry
 Chicago, IL 60616

 Robert Hampson
 National Bureau  of Standards
 A145, 222
 Washington, DC   20234

 Julian Heicklen
 The  Pennsylvania State  University
 Department  of  Chemistry
 152  Davey  Lab.
 University  Park, PA   16802
                                       103

-------
Dale 6. Hendry
SRI International
333 Ravenswood Drive
Menlo Park, CA  94025

John T. Herron
National Bureau of Standards
A145, 222
Washington, DC  20234

Jimmie A. Hodgeson
Office of Environmental  Measurements
National Bureau of Standards
Washington, DC  20234

Frank P. Hudson
Department of Energy
(E-201)
Washington, DC  20545

Robert E. Huie
National Bureau of Standards
A145, 222
Washington, DC  20234

Harvey E. Jeffries
University of North Carolina
Department of Environmental  Sciences
     and Engineering
Chapel Hill, NC  27514

J. A. Kerr
Department of Chemistry
The University
Birmingham, B15 2TT
England

Michael J. Kurylo
National Bureau of Standards
A145, 222
Washington, DC  20234

Stuart Z. Levine
Brookhaven National Laboratory
51 Bell Avenue   Building 426
Upton, Long Island, NY  11973

Alan C. Lloyd
Environmental Research and Technical
     Institute
2030 Alameda Padre Serra
Santa Barbara, CA  93103

Richard I. Martinez
National Bureau of Standards
A145, 222
Washington, DC  20234

Thomas J. McGee
University of Maryland
Molecular Physics Bldg.
College Park, MD  20742
 Joe  V.  Michael
 NASA/Goddard  Space  Flight  Center
   and Catholic  University  of America
 Code 691
 Greenbelt, MD  20771

 David F. Miller
 Battelle-- Columbus Laboratory
 505  King Avenue
 Columbus, OH  43201

 Mario J. Molina
 University of California
 Department of Chemistry
 Irvine, CA  92717

 Hiromi Niki
 Scientific Research Lab.
 Ford Motor Company
 P. 0. Box 2053
 Dearborn, MI  48121

 Robert J. O'Brien
 Portland State  University
 Department of Chemistry
 P. 0. Box 751
 Portland, OR  92707

H. Edward O'Neal
San Diego State University
 Department of Chemistry
San Diego, CA  92115

 David A. Parkes
Shell Research B.V.
Badhuiswes 3 Amsterdam N
Postbus 3003
The Netherlands

Robert Allen Perry
ERL/NOAA
Department of Commerce
Radio Building, Room 3522
Boulder, CO  80302

A. R. Ravishankara
Georgia Institute of Technology
Applied Science Lab.,  EES
Georgia Tech.
Atlanta, GA  30332

Keith Schofield
ChemData Research
P. 0. Box 40481
Santa Barbara, CA  93103

John H.  Seinfeld
California Institute of Technology
Pasadena, CA  91125

Donald H. Stedman
University of Michigan
c/o NCAR, P.  0.  Box 3000
Boulder, CO  80303
                                      104

-------
Louis J. Stief
NASA/Goddard Space Flight Center
Code 691
Greenbelt, MD  20771

Fred Stuhl
NOAA/ERL, Aeronomy Laboratory
325 Broadway, R448
Boulder, CO  80303

Wing Tsang
National Bureau of Standards
A145, 222
Washington, DC  20234

Peter Warneck
Max-Pianck-Institut fur Chemie
(Otto-Hahn-Institut)
65 Mainz, Saarstr.  23
Germany
Robert Tony Watson
Jet Propulsion Laboratory
California Institute of Technology
Building 183-601
4800 Oak Grove Drive
Pasadena, CA  91103

Karl Westberg
The Aerospace Corporation
P. 0. Box 92957
Los Angeles, CA  90009

Michael Whitbeck
General Motors Research
Environmental Science Department
Warren, MI  48090

Gary I. Whitten
Systems Applications, Inc.
950 Northgate Drive
San Rafael, CA  94903
                                       105

-------
        Subject Index
        Author Index
Aldehydes, 27
Alkenes (see ozone-alkene)
Alkoxy radicals, 20,34,51,62,65
Alkyl nitrate, 77
Aromatic compounds, 85,92,93,95

Creigee mechanism,  7,21,22
Creigee zwitterion, 64

Dialkenes, 19

Formaldehyde,  46,47
Formic Acid,  47,61
Formyl radical,  47

H02, 34,99
HSCV radicals,  34
Hydroxyl  radical, 7,15,31,47,60
   74,77,85,92,93,95,99,100

Monoterpenes,  19

Nitrogen oxides, 71,74,78
N03 radicals,  34

0 atom, 32,89
OH-olefin, 9,20
Olefins (see  ozone-alkene)
O'Neal-Blumstein mechanism, 22
Oxygenated hydrocarbons, 27
Ozone, 7,20,23,78,89
Ozone-alkene,  7,12,15,20,22,23,61

Peroxyacetyl  nitrate  (PAN), 35,78
Peroxyacyl nitrate, 35,77
Peroxyalkyl  nitrates,  77
Peroxy radicals, 64,65

Sulfur oxides,  99
S02, 99,100
Atkinson, Roger, 15,92

Batt, L., 21,62
Benson, Sidney W., 64

Calvert, Jack G., 99
Carter, William P. L., 20,65,77,78
Cox, Richard A., 65,71,100

Doty, Richard M., 73,74
Dodge, Marcia C., 21,78

Golden, David M., 51
Green, Patrick J., 74,93

Heicklen, Julian, 20,23,46,47
Hendry, Dale G., 77,85

Jeffries, Harvey E., 100

Lloyd, Alan C., 27

Miller, David F., 100

Niki, Hiromi, 7

O'Brien, Robert J., 74,93
O'Neal, H. Edward, 22
    %
Ravishandara, A. R., 47,95,101

Tsang, W., 22,64,79

Whitten, Gary Z., 22
                                        107

-------
NBS-114A IHI..V. a-781
   U.S. DEPT. OF COMM.

  BIBLIOGRAPHIC DATA
        SHEET
1. PUBLICATION OR REPORT NO.

        SP  557
                                                      l.Sev't Accession No,
3. fteciptent's Accession No,
 4. TITLE AND SUBTITLE
    Chemical  Kinetic Data  Needs for Modeling the  Lower
    Troposphere
                                                 5. Publication Date
                                                        August  1979
                                                                         *. Performing Organization Code
 7. AUTHORiS)

     J. T.  Herron,  R.  E.  Huie, and  J. A.  Hodgeson,  editors
                                                 8. Performing Organ. Report No.
 9. PERFORMING ORGANIZATION NAME AND ADDRESS


   NATIONAL BUREAU OF STANDARDS
   DEPARTMENT OF COMMERCE
   WASHINGTON, DC 20234
                                                19, Pmj«ct/Task/Wort< IJnlt No.
                                                11. Contract/Grant No.
12. SPONSORING ORGANIZATION NAME AND COMPLETE ADDRESS (Street. City. State, ZIP)
    Sponsored in  part by
    Environmental  Protection Agency
    Research Triangle Park,  NC   27711
                                                13. Type of Report & Period Covered

                                                           Final
                                                14. Sponsoring Agency Code
 15. SUPPLEMENTARY NOTES

          Library of Congress  Catalog  Number:   79-600125

      Document describes a computer program; SF-185, FIPS Software Summary, is attached.
 16. ABSTRACT (A 200~\vonl or /r«* factual Nummary of most significant information. If document includes a significant bibliography or
          This is  a  report  of the  proceedings of a workshop on  chemical  kinetic data
    needs  for modeling  the lower  troposphere,  held at  Reston,  Virginia, May  15-17,  1978.
    The meeting,  sponsored by the Environmental  Protection Agency and  the National
    Bureau of Standards,  focussed on six key problem areas in  tropospheric
    chemistry:  reactions  of olefins with hydroxyl radicals and ozone,  reactions of
    aldehydes,  free radical reactions,  reactions of oxides of  nitrogen, reactions of
    aromatic compounds,  and reactions of oxides of sulfur.

          The report includes a  summary  and list of major recommendations for further
    work,  review  papers,  discussion summaries,  contributed comments,  recommendations,
    and an attendance list.
 17. KEY WORDS 'six to twelve entries; alphabetical order; capitalize only the tir.it letter at the lirst key word unless a proper ,,nme;
   *<-[ittrtirrn* )

    Aldehydes; aromatics; chemical kinetics;  data needs; free radicals; modeling;
    NO  ;  oleh'ns;  SO  ;  troposphere.
 18. AVAILABILITY
                              .Unlimited
      For Offici.il Distribution. Do Not Release to NTIS
    X Otdei From Sup. of Doc. U.S. Government Printing Office, Washington, DC
      20402, SD Stock No. SN003-003-02111- 3 _

      Older From National Technical Information Service (NTIS), Springfield,
      VA. 22161
                                       19. SECURITY CLASS
                                         (THIS REPORT)
                                                                 UNCLASSIFIED
                                       20. SECURITY CLASS
                                         (THIS PAGE)
                                         UNCLASSIFIED
             21. NO. OF
              PRINTED PAGES

                  105
             2?. Price

                 $4.00
                                                                                     U SCOMM- DC

-------
   There's
   anew
                                           .  . .  the monthly
                                    magazine of the Nation-
                                      al Bureau of Standards.
                                  Still  featured are special ar-
                                  ticles of general interest on
                              current  topics such as consum-
                               er product safety and building
                            technology. In addition, new sec-
                         tions are designed to ...  PROVIDE
                        SCIENTISTS with illustrated discussions
                        of recent technical developments and
                  work in progress . . . INFORM INDUSTRIAL
                MANAGERS of technology transfer activities in
              Federal and private labs. . . DESCRIBE TO MAN-
           UFACTURERS advances in the field of voluntary and
         mandatory standards. The new DIMENSIONS/NBS also
        carries complete listings of upcoming conferences to be
      held at NBS and reports on all the  latest NBS  publications,
    with information on how to order. Finally, each issue carries
  a page of News Briefs, aimed at keeping scientist and consum-
er alike up  to date on major developments at the Nation's physi-
                     cal sciences and measurement laboratory.

              (please detach here)
SUBSCRIPTION ORDER FORM
Enter my Subscription To DIMENSIONS/NBS at $11.00. Add $2.75 for foreign mailing. No additional
postage is required for mailing within the United States or its possessions. Domestic remittances
should be made either by postal money order, express money order, or check. Foreign remittances
should be made either by international money order, draft on an- American bank, or by UNESCO
coupons.

Send Subscription to:
                         NAME-FIRST, LAST

  J  I I  I   I I  M I  I  I  I  I  I  I  I  I  I I  I  I  I I  I   I I   I  I
                COMPANY NAME OR ADDITIONAL ADDRESS LINE
                         STREET ADDRESS
              CITY
                                                    ZIP CODE
                                                    Mil
P«ASE PRINT
                                              CD Remittance Enclosed
                                                 (Make checks payable
                                                 to Superintendent of
                                                 Documents)

                                              O Charge to my Deposit
                                                 Account No.
                                        MAIL ORDER FORM TO:
                                        Superintendent of Documents
                                        Government Printing Office
                                        Washington, D.C. 20402

-------
                                NBS  TECHNICAL  PUBLICATIONS
                    PERIODICALS
JOURNAL OF  RESEARCH—The  Journal of  Research
of the National Bureau of Standards reports NBS research
and development  in  those disciplines of the physical and
engineering sciences in which the Bureau is  active. These
include physics,  chemistry,  engineering,  mathematics, and
computer sciences. Papers cover a broad range of subjects,
with  major emphasis on measurement  methodology, and
the basic technology underlying standardization.  Also in-
cluded from time to time are survey articles on topics closely
related to the Bureau's technical  and scientific programs. As
a special service to subscribers each issue contains  complete
citations to all recent NBS  publications  in NBS and non-
NBS  media. Issued six times  a  year. Annual subscription:
domestic $17.00; foreign $21.25. Single copy, $3.00 domestic;
$3.75 foreign.
Note: The Journal was formerly published in  two  sections:
Section A "Physics and Chemistry" and Section  B  "Mathe-
matical Sciences."
DIMENSIONS/NBS
This  monthly magazine is published to inform scientists,
engineers,  businessmen, industry,  teachers,  students,  and
consumers of the latest advances in science and technology,
with primary emphasis on the  work at NBS. The magazine
highlights and reviews such  issues as energy research, fire
protection, building technology, metric conversion, pollution
abatement, health  and safety, and consumer product per-
formance. In addition, it reports  the results of Bureau pro-
grams in measurement standards and techniques, properties
of matter and materials, engineering standards and  services,
instrumentation, and automatic data processing.
   Annual subscription:  Domestic, $11.00; Foreign $13.75

                 NONPERIODICALS
Monographs—Major  contributions to  the technical liter-
ature on various subjects  related to  the Bureau's scientific
and technical activities.
Handbooks—Recommended codes of engineering  and indus-
trial practice (including safety codes)  developed in coopera-
tion with interested  industries, professional organizations,
and regulatory bodies.
Special  Publications—Include  proceedings  of conferences
sponsored by NBS, NBS annual reports,  and other special
publications appropriate to this grouping such as wall charts,
pocket cards, and bibliographies.
Applied  Mathematics  Series—Mathematical  tables,  man-
uals,  and  studies of special interest to physicists, engineers,
chemists, biologists, mathematicians, computer programmers,
and others engaged in scientific and technical work.
National Standard  Reference Data Series—Provides quanti-
tative  data  on  the  physical  and chemical  properties of
materials, compiled from the world's literature and  critically
evaluated.  Developed  under  a  world-wide program co-
ordinated by NBS. Program under  authority of  National
Standard Data  Act (Public Law 90-396).
 NOTE: At present the principal publication outlet for these
 daia is the Journal of Physical  and  Chemical  Reference
 Data (JPCRD) published quarterly for NBS by the Ameri-
 can Chemical  Society (ACS) and  the American Institute of
 Physics (AIP).  Subscriptions,  reprints,  and  supplements
 available from ACS, 1155 Sixteenth St. N.W., Wash., D.C.
 20056.
 Building Science Series—Disseminates technical information
 developed at the Bureau on building materials, components,
 systems, and whole structures. The series presents research
 results, test methods, and performance criteria related to the
 structural  and environmental functions and the  durability
 and safety  characteristics of building elements  and systems.
 Technical Notes—Studies or reports which  are complete in
 themselves  but  restrictive  in  their treatment of  a  subject
 Analogous  to  monographs but  not  so  comprehensive  in
 scope  or definitive in treatment of the subject area. Often
 serve as a  vehicle for final reports of work performed  at
 NBS under the sponsorship of other government  agencies.
 Voluntary Product Standards—Developed under procedures
 published  by the Department of Commerce  in   Part  10,
 Title 15, of the Code of Federal  Regulations.  The purpose
 of the standards is to establish nationally recognized require-
 ments  for  products, and to provide all concerned interests
 with a basis for common understanding  of the characteristics
 of the products. NBS administers  this program as  a supple-
 ment to the activities  of  the private  sector standardizing
 organizations.
 Consumer  Information  Series—Practical information, based
 on NBS research and experience,  covering areas of interest
 to  the   consumer.  Easily  understandable  language  and
 illustrations provide useful  background knowledge  for shop-
ping in  today's technological marketplace.
Order  above  NBS  publications   from: Superintendent  of
Documents, Government Printing  Office, Washington,  D.C.
20402.
Order following NBS publications—NBSIR's and FIPS from
the National  Technical Information Services,  Springfield,
 Va. 22161.
Federal  Information   Processing  Standards  Publications
(FIPS PUB)—Publications  in this  series collectively consti-
tute the Federal Information Processing Standards Register.
Register serves as the official  source  of information in the
Federal  Government regarding  standards issued  by  NBS
pursuant to the Federal Property  and Administrative  Serv-
ices Act of 1949 as amended, Public Law 89-306 (79 Stat
 1127),   and as  implemented  by  Executive Order  11717
 (38 FR 12315, dated May  11, 1973) and Part 6 of Title  15
 CFR (Code of  Federal  Regulations).
 NBS  Interagency Reports  (NBSIR)—A special  series  of
 interim or  final  reports on work performed  by  NBS  for
 outside  sponsors  (both government  and non-government).
 In  general, initial  distribution is  handled by  the sponsor;
 public distribution is by the National  Technical Information
 Services (Springfield, Va. 22161) in paper copy or microfiche
 form.
                            BIBLIOGRAPHIC  SUBSCRIPTION  SERVICES
Hie following current-awareness and literature-survey bibli-
ographies are issued periodically by  the Bureau:
Cryogenic Data Center Current Awareness Service. A litera-
  ture survey issued biweekly. Annual subscription: Domes-
  tic, $25.00; Foreign, $30.00.
 Liquefied Natural Gas. A literature  survey issued quarterly.
Annual subscription: $20.00.
 Superconducting Devices and Materials. A literature survey
 issued quarterly. Annual subscription: $30.00. Send subscrip-
 tion  orders and remittances for the preceding bibliographic
 services to National Bureau  of Standards, Cryogenic Data
 Center  (736.00)  Boulder, Colorado  80303.

-------