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.
------- |