Citation
A technique for studying the kinetics of nitrate radical reactions with organics

Material Information

Title:
A technique for studying the kinetics of nitrate radical reactions with organics
Creator:
Stevenson, Peter Day
Publication Date:
Language:
English
Physical Description:
xii, 126 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Atmospheric chemistry ( lcsh )
Nitrogen oxides ( lcsh )
Atmospheric chemistry ( fast )
Nitrogen oxides ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 121-126).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Chemistry
Statement of Responsibility:
by Peter Day Stevenson.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
31250634 ( OCLC )
ocm31250634
Classification:
LD1190.L46 1994m .S74 ( lcc )

Full Text
A TECHNIQUE FOR STUDYING THE KINETICS
OF NITRATE RADICAL REACTIONS WITH ORGANICS
B.A., University of Vermont, 1977
B.S., University of Colorado at Denver, 1988
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
of the requirements for the degree of
by
Peter Day Stevenson
in partial fulfillment
Master of Science
Chemistry
1994


This thesis for the Master of Science
degree by-
Peter Day Stevenson
has been approved for the
Department of
Chemistry
by
If.
z, mV
Date


Stevenson, Peter Day (M.S., Chemistry)
A Technique for Studying the Kinetics of Nitrate Radical
Reactions with Organics
Thesis directed by Professor Larry G. Anderson
ABSTRACT
The kinetics of the gas phase nitrate radical (N03)
reactions have become increasingly important in
understanding the atmospheric fate of a variety of gas-
phase organic compounds in the atmosphere at night. In
polluted air, N03 reacts readily with certain organics,
indirectly forming other free radicals and nitrated
species of unknown toxicity and reactivity. Atmospheric
N03 concentrations up to 0.43 ppb have been measured.
Kinetic studies indicate that the rate of some reactions
is sufficiently fast for N03 to be as significant a sink
for organics as OH and 03.
The temperature dependence of the N03 + N02 <-> N205
equilibrium is crucial to understanding the kinetics of
N03 reactions with organics, as N205 can exist at much
higher concentrations than N03, and a slow reaction with
N205 could be interpreted as a fast reaction with N03. The
iii


vast majority of the kinetics information comes from
relative rate studies, and temperature control is usually
stated as 2 K. If the organic under investigation
reacted with N205 to any significant extent, or if the N03-
organic reaction has any temperature dependency, this
lack of temperature control would make it impossible to
obtain any useful kinetics information.
The technique described in this work maintains
temperature control to better than 0.2 K. A gas
chromatograph was used to monitor the rate of organic
reactant decay. Samples were analyzed at fixed time
intervals, and the data were stored on a PC for later
analysis. Experimental data were then compared to model
data, and by taking advantage of the large temperature
dependence of the N03-N205 equilibrium, we have determined
the Arrhenius expression of the temperature dependence of
the N03-propene reaction, in cm3 molecule'1 s'1, to be:
k(303-3i8K) = (1.811.04) x 10"13 exp [ (980348)/T]
IV


The N03-ethene reaction was also studied, and we
verified that the literature values for the rate constant
are in reasonable agreement with the temperature
dependent rate that our technique suggests of:
k(308-3i8K> = (.813-2.03) x 10"16 cm3 molecule"1 s'1
Due to the slow rate at which N03 reacts with ethene,
we could draw no further conclusions.
This abstract accurately represents the content of the
candidate's thesis. I recommend its publication.
Signed
v


ACKNOWLEDGEMENTS
I wish to acknowledge Dr. Larry Anderson's patience
and support, both monetary and otherwise. I also wish to
acknowledge the support I received from my wife Sue and
daughter Sarah during the last few years that I worked on
this project. Finally, I wish to acknowledge the support
I received through all of my schooling from my parents,
Robert Presley Stevenson and Vera Kemp Stevenson, and the
inspiration from my grandfather, Archibald Reed Kemp, who
received his M.S. in Chemistry in 1918 from the
California Institute of Technology.
vi


CONTENTS
CHAPTER
1. INTRODUCTION........................................1
The Atmospheric Importance of N03..................5
Non-methane Hydrocarbons..........................10
N205 Reactions....................................14
Products of N03 and N205
Reactions with Organics...........................16
2 PREVIOUS RESEARCH
TECHNIQUES AND RESULTS.............................21
Absolute Rate Studies.............................22
Ravishankara and Mauldin, Georgia Tech.......24
National Center for Atmospheric Research......26
National Oceanic
and Atmospheric Administration...............2 8
Oxford University.............................30
Statewide Air Pollution Research Center.......32
Max Planck Institute..........................34
Germany and France, Two Labs..................35
Relative Rate Studies............................3 6
Statewide Air Pollution Research Center.......37
Other Techniques..................................41
Niki and Coworkers............................42
vii


University of Technology
and University of Gotberg, Sweden.............43
Commission of European
Communities, Ispra, Italy.....................43
Purpose of this Research..........................44
3 EXPERIMENTAL SYSTEM................................46
Reaction Apparatus and Temperature Control........46
Reactant Preparation..............................49
Data Acquisition by Gas Chromatography............53
Reactants.........................................60
Experimental Procedure............................60
Modelling.........................................62
4 RESULTS AND DISCUSSION.............................63
System Characterization...........................63
Ozone Wall Loss...............................66
Correction Factors............................67
General Model Results.............................70
Determination of Wall Loss........................72
N205 Reaction with Propene........................75
Propene Reaction Results and Discussion...........76
Propene and Ozone Reaction Results............76
Comparison of Ozone-Propene
Experimental and Model Data...................77
Propene-NOj Reaction Results..................80
viii


Comparison of N03-Propene
Experimental and Model Data...................82
Effect of ozone concentration,
temperature, and reactions 4.4,
4.7, and 4.11............................83
The temperature dependence of
the N03-propene reaction rate............86
Data from the 35 C experiments..........96
Ethene Data........................................96
5 CONCLUSIONS............................99
Refinements to the Technique......................105
APPENDIX
A. Supporting Figures and Tables..................107
BIBLIOGRAPHY..........................................121
ix


FIGURES
Figure
1.1 Typical Nighttime Profile of N03....................8
3.1 Teflon Film Reaction Vessel.........................47
3.2 Environmental Chamber...............................48
3.3 Reactant and Diluent Loading System.................50
3.4 Ozone and Oxygen Loading System.....................51
3.5 GC and Data Acquisition System......................54
3.6 Chromatogram of 7 Minute Propene Run................56
3.7 Propene Peak and Integration Parameters.............56
4.1 Ozone-Propene Reaction Data.........................77
4.2 Ozone-Propene Reaction: Effect of Temperature.... 78
4.3 Ozone-Propene Reaction: Effect of Varying k......79
4.4 Ozone-Propene Reaction: Effect of [03]..............79
4.5 N03-Propene Reaction Data...........................81
4.6 N03-Propene Reaction: Data vs. Starting
Rate and Wall Loss..................................84
4.7 N03-Propene Reaction: Effect of [03] and k..........85
4.8 N03-Propene Reaction: Effect of Temperature
and Three Reactions (4.4, 4.7, and 4.11)= 0.........87
4.9 N03-Propene Reaction at 45 C: Effect of
k and N205 Wall Loss................................89
4.10 N03-Propene Reaction at 40 C: Effect of
k and N205 Wall Loss...............................90
x


4.11
NOj-Propene Reaction at 30 C: Effect of
k and N205 Wall Loss........................91
4.12 N03-Propene Experimental Data at 30 C
vs. Best Fit Model Data...........................93
4.13 N03-Propene Experimental Data at 40 C
vs. Best Fit Model Data...........................94
4.14 N03-Propene Experimental Data at 45 C
vs. Best Fit Model Data............................95
5.1 Arrhenius Plot.....................................100
A. 1 Ozone Meter Calibration............................109
A.2 Ozone Wall Loss....................................110
A.3 Corrected Propene Calibration Runs.................Ill
A.4 N03-Propene Reaction at 3 0 C: N03 vs
N205 Wall Loss.....................................113
A.5 N03-Propene Reaction at 40 C:
N03 vs. N2Os Wall Loss.............................114
A. 6 N03-Propene Reaction at 45 C:
N03 vs. N205 Wall Loss.............................115
A. 7 N03-Propene Reaction at 35 C:
Effect of [03] k, and Wall Loss..................116
A.8 Ozone-Ethene Experimental Data vs. Model Data....117
A.9 N03-Ethene Reaction Data...........................118
A. 10 N03-Ethene Average Data at 35 C:
vs. Best Fit Model Data...........................119
A. 11 N03-Ethene Experimental Data at 45 C:
vs. Best Fit Model Data...........................120
xi


TABLES
Table
1.1 Non-methane Hydrocarbon Concentrations..............13
1.2 Measured and Estimated Yields of Products...........18
2.1 Summary of Rate Constants for N03-0rganics..........38
4.1 Chemical Mechanisms Used to Describe the
Reactions of N03-N205 with Propene and Ethene.......71
5.1 Ethene and Propene Rate Constants/Coeffients......101
5.2 Temperature Dependent Rate Coeffients
Determined to Date.................................103
A. 1 Chrom Perfect Report...............................108
A.2 Model of N03-N205 + Propene Reaction
Mechanism at 45 C.................................112
xii


CHAPTER 1
INTRODUCTION
The goal of this research is development of a new
technique for studying nitrate radical kinetics, one that
is more convenient and less expensive then existing
methods, and one that can sort out other potential
reactions such as those of dinitrogen pentoxide. Current
kinetic study methods are either cumbersome and
expensive, or are subject to a number of uncertainties
that cast doubt on the results. This technique utilizes
standard laboratory equipment and computer model-
generated data to study nitrate radical reactions with
organics, specifically propene and ethene. The remainder
of this chapter introduces the nitrate radical and where
it fits into the broader subject of air pollution.
Air pollution is as much a fact of life as breath
itself, and doesn't need to be introduced. The majority
of the human population exists in areas where the air
they breathe is impacted in some way by the far reaching
affects of man's activities on the atmosphere. One of
1


the more dramatic examples of the impact of air pollution
occurred in London after coal replaced wood as a fuel
source. To blend in with soot-covered trees and other
objects, a species of moth actually evolved from being
white, and hence easy for birds to spot, to being almost
entirely black, thereby blending in with the soot that
covered the land.
Many classes of compounds are considered important
in atmospheric chemistry. These include oxygen compounds
such as atomic oxygen (0), molecular oxygen (02), and
ozone (03) and compounds containing oxygen and hydrogen,
such as the OH radical and H02. Hydrocarbons, halogen
containing species, sulfur containing species, and
nitrogen containing species round out the list of classes
of compounds important in the atmosphere.
Of particular importance are the oxides of nitrogen,
which control the destruction and production rates of
ozone, which in turn affects the chemical composition and
thermal characteristics of the atmosphere (1). The
primary oxides of nitrogen are nitric oxide (NO) and
nitrogen dioxide (N02) both of which are commonly
referred to as N0X. Temporary species include the
2


nitrate radical (N03) dinitrogen pentoxide (N205) HONO,
and assorted radicals such as H02N02 and R02N02, which act
as reservoirs of nitrogen oxides. Finally, removal of
nitrogen oxides occurs as nitric acid, HN03, typically by
OH reaction with N02, or through a heterogeneous process
involving water vapor or aerosols and the N205 molecule,
among other less important processes.
In the stratosphere, which is 10 to 50 km in
altitude, N0X catalytically destroys 03. In the
troposphere (0 to 10 km in altitude), however, N0X
increases ozone production, and also affects the OH
radical, which is responsible for initiating the
oxidation of many otherwise stable atmospheric species
such as carbon monoxide (CO) methane (CH4) and non-
methane hydrocarbons (2).
The primary sources of tropospheric NOx (typically
emitted as NO) are combustion of fossil fuel, biomass
burning, lightening, oxidation of ammonia, microbial
activity in soils, the oceans, and input from the
stratosphere. The emitted NO reacts with 03 to form N02
and 02:
NO + 03 > N02 + 02 1.1
3


NO and N02 then rapidly change by a number of
interconvertible reactions. Ozone then results from the
photolysis of N02, forming 0 which then reacts with 02.
NO and N02 may also be converted to HONO, H02N02/ N03, N205,
HN03, and organic nitrates or peroxynitrates (R02N02) .
Whereas the importance of NO and N02 has been known for
some time, atmospheric researchers did not detect N03 in
the atmosphere until 1978 (3). With in-situ measurements
available, it became possible to test atmospheric models
with a reasonable expectation of a sensitive response of
the predicted quantity to changes in mechanism and
kinetics. This in turn stimulated laboratory studies of
the radical, and it is now known to be of particular
importance in the atmosphere. The understanding of the
complexities of the atmosphere can be furthered only with
the ability to test models with laboratory studies and
with atmospheric measurements. This interrelationship
between the three sources of information is crucial to
developing a valid understanding of the chemistry of the
atmosphere.
4


The Atmospheric Importance of N03
At night, the nitrate radical controls the
concentrations of nitrogen-oxygen compounds, acts as a
source of nitric acid, oxidizes a wide range of organic
species, produces peroxy and hydroxy radicals, and yields
nitrated products that may be toxic. The nitrate radical
also reacts very quickly with other radicals, and will
transform other radicals that persist at night because of
their otherwise low reactivity(4).
The existence of the nitrate free-radical (N03) was
first postulated in 1881 by Hautefeuille and Chappuis (5)
as an intermediate in the N205-catalyzed decomposition of
03. The visible band structure of N03 was established in
1938 by Jones and Wulf (6), and in 1962 Ramsay (7)
assigned the major vibrational band features. A few
other studies of N03 absorption cross sections occurred in
the 1960s and early 1970s, but it wasn't until 1974 and
1975 that reactions with certain organics were
investigated by Niki and co-workers (8,9).
Since these pioneering studies, numerous kinetic
studies have indicated that the rate of some N03-organic
reactions are sufficiently fast for N03 to be as
5


significant a sink for organics as 03 and also the OH
radical. Although the OH radical is usually the main
agent of attack on organic species during the day, N03 may
be the most important oxidizing species in the
troposphere at night. For some species, such as dimethyl
sulfide, the reaction at night with N03 may even dominate
over the daytime reaction with OH due to the fast rate of
reaction.
N03 reacts readily with certain organics, indirectly
forming other free radicals and nitrated species of
unknown toxicity and reactivity. A strong visible
absorption peak at 662 nm allows for direct measurement
in the atmosphere, and observers have measured N03 at
night around the globe and in a variety of locales, at
concentrations up to 0.43 ppb (10). These concentrations
are much higher than concentrations of the OH radical, so
even a moderately fast reaction with N03 may account for
the majority of atmospheric decay of a particular gas-
phase organic compound.
N03 is not an important species during the day
because it photolyses rapidly. At night, however, it is
produced by the slow, temperature dependent reaction:
6


o3 + no2 > no3 + 02
1.2
The half life of N02 is about 1.5 hours. After formation
of N03, a rapid equilibration with dinitrogen pentoxide
(N205) occurs:
N03 + N02 (excess) + M <-> N205 + M 1.3
This equilibrium is also temperature dependent: If an
air mass is compressively heated by downward motion, the
equilibrium shifts, giving a source of N03; and, if an air
mass is cooled by upward motion or by radiative heat
loss, the equilibrium shifts toward N205, reducing the
concentration of N03. Under normal atmospheric
conditions, this equilibrium predicts approximately equal
concentrations of N03 and N205 when the N02 concentration
is low and hundreds of times as much N205 as N03 when there
is a significant excess of N02, which is more typical of
urban atmospheric conditions.
Other loss processes for N03 include reaction with
N02 to form NO, 02, and N02, and reaction with NO to form
two molecules of N02. In addition, N205 reacts rapidly
with moist surfaces to form HN03 (11) Humidity in the
form of fog can affect the amount of N03 available
(neither species reacts with water in the gas phase), and
7


this latter reaction is thought to explain the "wall
loss" of N03/N205 in lab studies.
Typical nighttime profiles of N03 concentration begin
by rising just after sunset either to fairly constant
levels for most of the night, or to a peak in the early
morning hours (12). Then N03 rapidly decays at sunrise.
FIG. 1.1: Typical Nighttime Profile of N03 () [with 03
(x) and N02 (.) RH read on left scale.
Adapted from (12)].
Time
8


N02 and 03 concentrations, transport mechanisms, and
relative humidity all affect the N03 concentration. As
the relative humidity increases, the N03 lifetime
decreases. Thus, at relative humidities of less than
50%, N03 radical lifetimes up to about 60 minutes were
observed, while at relative humidities of greater than
50% the lifetimes are always less than 10 minutes (12).
A year-long study of N03 concentrations in the
stratosphere concluded that the data display a maximum in
summer and a minimum in winter, which was in good
agreement with model predictions (13). This annual cycle
is driven by the temperature dependence of the primary
stratospheric N03 formation, reaction 1.2. This in turn
suggests that short-term variability in N03 concentration
in the atmosphere can also be a result of short-term
temperature fluctuations.
Recent concerns regarding indoor air quality, where
concentrations of organics are usually higher than
outdoors due to source emissions from cleaning solvents,
fuels, wood products, synthetic products, etc., have also
included concerns about free radical chemistry, N03
included. Indoor ozone concentrations can be a
9


significant fraction of the outdoor concentration, and if
N02 is present, then N03 will be produced. In the absence
of direct sunlight, N03 could then react with organics
even during the day, producing other radicals, nitrated
species, and nitric acid (14). A recent study (15) has
concluded that an indoor source of nitric acid must be
present during the summer months, which may indicate that
N03 is present indoors also. Therefore, this chemistry
may be an important source of free radicals indoors.
Non-methane Hydrocarbons
Natural and anthropogenic organic materials play an
important role in atmospheric chemistry. Organic
material can be in both the vapor phase or exist as
either viable particulates (molds, fungi, bacteria,
pollen, etc.) or non-viable aerosols. In urban areas
hydrocarbon species, particularly non-methane
hydrocarbons (NMHC), are a key constituent of
photochemical smog. The intermediate products of the
oxidation of NMHC react with N0X species to produce 03.
There are six major sources of vapor phase and
particulate organic carbon (16) They are vegetation,
10


soil and other geologic sources, natural waters, biomass
burning, anthropogenic sources, and in-situ processes.
Dry deposition and wet precipitation remove organic
carbon from the atmosphere. In addition, vapor phase
organics are removed by physical and chemical conversion
to particulates, and by chemical transformations into
gaseous inorganic products. Rough estimates of the vapor
phase organic carbon cycling through the atmosphere are
about 1000 teragrams (1012 grams) of carbon per year
(C/yr) (16). Of this approximately 830 teragrams C/yr is
released by vegetation, primarily in the form of isoprene
and terpenes. Anthropogenic sources, particularly motor
vehicles, contribute an estimated 75 teragrams C/yr.
Non-methane hydrocarbons have a controlling
influence on the production of ozone in the polluted
boundary layer of the atmosphere. In winter,
photochemical degradation is much less efficient, and the
affect of NMHC is more widespread, as they can escape
from the boundary layer in source areas and disperse into
the free troposphere over large parts of the northern
hemisphere. The winter concentrations of NMHC in air
over the North Atlantic Ocean in winter, which of course
11


depends greatly on source, are much higher than the
equivalent summer values, reaching a maximum carbon
concentration of 20 parts per billion by volume (ppbv) in
the form of reactive carbon compounds with lifetimes of
days to months. This forms a reservoir of reactive
chemicals which could influence the extent of ozone
formation in the troposphere (17). Whereas much of the
seasonal chemical behavior of NMHC is accounted for by OH
radical chemistry, there is evidence that N03 chemistry
may play a significant part in the removal of NMHC from
the atmosphere, especially in winter (17).
Table 1.1 presents NMHC data from a variety of
sampling events in both urban and oceanic settings. The
table also presents winter maxima and summer minima, in
parts per trillion by volume (pptv), to illustrate the
seasonal variations of NMHC.
Reaction of N03 with NMHC occurs via two pathways,
addition to an unsaturated carbon atom bond, or hydrogen
abstraction, as typified by the reactions:
N03 + C=C -> C(0N02)C 1.4
N03 + RH -> HN03 +R 1.5
12


Table 1.1: Non-methane Hydrocarbon Concentrations
[adapted from (17)].
Background Average North Atlantic Ocean 39 U.S. Cities London Street
Hydrocarbon Soecies Winter Maximum, PPtv Summer Minimum, PPtv Median ppbv ppbv
Acetylene 673 126 6.5 531
n-Butane 405 25 10.1 147
i-Butane 193 47 3.7 78
n-Pentane 147 8 4.4 38
n-Hexane 65 5 1.8 19
Ethane 2220 1210 11.7 26
Propane 870 85 7.8 15
Ethene 10.7 217
Propene 2.6 54
trans-2-Butene 0.6 9
1-Butene 0 10
i-Butene 1.5 19
Isoprene 5.1
Benzene 230 25 2.1 64
Toluene 115 10 4.8 114
13


Hydrogen abstraction is usually a much slower
process than addition, but can occur simultaneously at
higher temperatures (18,19). Reactions of N03 with
alkenes and substituted aromatics, therefore, are
generally much faster than reactions with similar alkanes
and aromatics. For simplicity, the terms "organic
compounds" or "organic" as used in this report refers to
NMHC.
N-,0 Reactions
N205 concentrations can be ten to thousands of times
greater than N03 concentrations, and ppb levels of N03 may
exist under a variety of conditions (20). Therefore,
reaction of N205 with organics could be important.
Temperature control must be maintained for the
following reasons:
1. A slow reaction with N205 could appear to be a
fast reaction with N03;
2. the N03 equilibrium with N205 is temperature
dependent; and
3. the potential temperature dependence of the
reaction: N03/N205 + organic -> products. 1.6
14


For these reasons, it is obvious that studies of N03/N205
reactions, especially relative rate studies, should
maintain strict temperature control to generate
meaningful kinetics data. They also must be able to
differentiate N03 reactions with organics from N205
reactions with organics, especially if the source of the
N03 is from reaction 1.2 and the reversible reaction 1.3.
Atkinson, et al. has estimated atmospheric
concentrations of N205 from simultaneous measurements of
N03 and N02, and has found that concentrations as high as
about 15 ppb can occur(10). In laboratory studies, an
excess of N02 is often used, forcing the reversible
reaction 1.3 to the right, creating as much as one
thousand times as much N205 as N03.
A number of studies have confirmed that N205 reacts
with NMHC, primarily polycyclic aromatic hydrocarbons
(PAHs) and substituted PAHs. By varying the reaction
temperature and thus varying the concentrations of N03 and
N205 via equation 1.3, by varying the N02 concentration,
and by using an N03 scavenger (2,3-dimethyl-2-butene),
Atkinson, et al. (21,22,23) determined that the reactions
of the PAHs naphthalene and 1- and 2-methyl naphthalene
15


were occurring by an overall process that is kinetically
equivalent to the reaction of the PAH with N205. Further,
it was found that acenaphthene must at least partially
react with N205 (24) .
Products of NO-, and N?0<
Reactions with Organics
The product studies of all N03 reactions with alkenes
indicate the initial addition of N03 is followed by 02
addition to form nitro oxy-alkylperoxy radicals:
N03 + R-CH=CH-R -> R CH CH (0N02) R 1.7
02 + R-CH-CH (0N02) -R -> R-CH(02) -CH(0N02) -R. 1.8
This is followed by reaction with N02 and establishment of
an equilibrium between thermally unstable nitro oxy-alkyl
peroxynitrates:
N02 + R-CH(02) -CH0N02) -R <-> R-CH(02N02) -CH(0N02) -R 1.9
Under atmospheric conditions, the nitro oxy-alkylperoxy
radicals will react to form nitro oxy-alkoxy radicals
mainly by reaction with NO (which under certain lab
conditions will not exist due to the rapid reaction of NO
with N03) :
N0+ R-CH(02N02) -CH(0N02) -R -> R-CH(O)-CH(0N02)-R +2N02 1.10
Decomposition of nitro oxy-alkoxy radicals will lead to
16


the formation of aldehydes and the release of N02:
R-CHO-CH (0N02) -R <-> 2R-CH0+N02, 1.11
and their reaction with 02 forms stable nitrooxy-
aldehyde/-ketone compounds (25):
R-CH(O)-CH(0N02) -R + 02 -> [R-CO-CH (0N02) -R] + H02. 1.12
To date, nothing is known about nitro oxy-
aldehyde/ketone compounds in air. It is well known that
peroxyacetyl nitrate (PAN) is a phytotoxicant, so it is
possible that other products of N03 reactions with alkenes
could also be phytotoxicants.
Table 1.2 summarizes the measured and estimated yields
for the products of N03 reactions with alkenes and
dialkenes. As is indicated on the table, aldehydes and
CO can comprise about 15 to 80% of the products, with the
remainder being unknown or one or more nitro oxy-
aldehyde/-ketones.
Propylene glycol dinitrate was reported as a
significant product of the N03-propene reaction (26).
However, a later study conducted at conditions that
approximated the atmosphere more closely, found that
nitrooxy-acetone was the major nitrate-containing product
(10), which agrees with the above study (25).
17


Table 1.2: Measured and Estimated (est.) Yields of
Products Formed in Dialkenes [adapted the Reactions of N03 with Alkenes and from (25)]
Reactant Products (% vield on a molar basis)
Propene HCHO (8) CH3CHO (12) total nitrates (58 est)
1-Butene HCHO (11) CH3CH2CHO (12) total nitrates (60 est)
trans-2-Butene CH3CH0 (70) CH3C0CH(0N02) CH3 (55) CH3CH (0N02) CH (0N02) CH3 (4)
iso-Butene HCHO (80) CH3COCH3 (85) total nitrates (25 est)
lf 3-Butadiene CO (4) HCHO (12) CH2=CHCHO (12) total nitrates (60 est)
Isoprene CO (4) HCHO (11) total nitrates (80 est)
18


Products from the reaction of N03 with sulfur
compounds have also been studied. Reaction mechanisms
were proposed, and CH3CH2S03H, S02, H2S04, CH3CHO, and
CH3CH20N02 were identified as products in the reaction of
N03 with ethane thiol (CH3CH2SH) diethane sulfide
[ (CH3CH2)2S] and diethane disulfide [(CH3CH2)2S2 (27). The
initial reaction step was found to be a hydrogen atom
abstraction (forming HN03) probably after the formation
of an initial adduct. For all the reactions it appeared
that three radicals [R-S, R-S(O), and R-S(02)] were key
intermediates. Another study regarding N03 reaction with
dimethyl sulfide (CH3SCH3 or DMS) found CH3SN02 as an
intermediate (27), which is interesting due to its sulfur
and nitrogen character.
The aromatics toluene, para-xylene, and ortho-xylene
were reacted with N03, and the products benzaldehyde,
benzylnitrate, nitrotoluene, methyl benzaldehydes, methyl
benzyl nitrates, minor amounts of methyl benzyl alcohol,
and methylnitrotoluene were found (28).
Nitronaphthalenes form from reaction of N205 with
naphthalene (23). 1- and 2-nitronaphthalene contributed
about 25% of the potential product yield, indicating that
19


other products must also be formed. Methyl
nitronaphthalenes formed from the reaction of 1- and 2-
methyl naphthalene with N205 (22). Other nitroarene
products have been found to be 4-nitropyrene (reaction
with pyrene) and 2-nitrofluoranthene (reaction with
fluoranthene) (24).
The health implications of N03 reaction products may
need to be considered. The nitronaphthalenes are both
weak direct and activatable mutagens, and reduction of
any nitro-containing organic could form the corresponding
amine, such as 2-aminonaphthalene, a human carcinogen
(23) .
20


CHAPTER 2
PREVIOUS RESEARCH
TECHNIQUES AND RESULTS
One of two approaches are generally taken for
kinetics studies. Absolute rate studies follow the
concentration change of the species of interest directly.
For N03 studies, the concentration change of N03 is
followed, usually by visible spectroscopy. Relative rate
studies try to determine an "absolute" rate constant by a
"relative" method. By following the rates of two
reactions, the relationship between a known rate constant
(k,) and an unknown rate constant (k2) of the species
being studied can be determined. The known rate constant
is typically generated by an absolute technique, and for
most relative rate studies of N03, the decay of the
organics being studied are followed, usually by gas
chromatography.
21


Absolute Rate Studies
Absolute rate studies, which follow the
concentration changes of N03 directly, are believed to
produce more reliable results than the relative rate
techniques discussed in the following section. However,
absolute rate techniques are more difficult, time
consuming, and can be quite equipment-intensive. For the
study of the nitrate radical reaction with a variety of
organics, almost all of the work has been accomplished
using flow systems. It wasn't until 1985 that
Ravishankara and Mauldin (29) published the results of
their absolute rate study of N03 + trans-2-butene, and
since then absolute studies have been conducted in
Germany and France, as well as England and the U.S.
Temperature control is of lesser importance in
absolute rate studies regarding N03 and N205 because the
source of N03 is controlled, usually only N02 exists as a
"contaminant", the studies are conducted at low
pressures, and N03 is measured directly during the
reaction, thereby minimizing a number of uncertainties.
On the other hand, any temperature dependence that a
reaction has will increase the uncertainty of the data if
22


temperature control is not maintained. To generate the
best data possible, and to investigate the temperature
dependence of the reaction, temperature control must be
maintained. Further, the low pressures used in most
absolute studies make it reasonable to question whether
results apply to tropospheric conditions (19).
Prior studies of the temperature dependence of the
N03 reaction with propene, ethene, 1-butene, chloroform,
methanol, and a series of alkynes and alkanes conducted
by Canosa-Mas, et al. (18,30,31,32,33) were conducted in
the temperature range of 298-523 Kelvin (298 K=25 C) .
Dlugokencky and Howard have also conducted temperature
dependent studies of the reactions of N03-organics
(19,34). Whereas these studies are not conducted
exclusively in the temperature range of the atmosphere,
they have begun to produce the necessary data to begin to
understand the temperature dependence of the N03-organics
reactions.
23


Ravishankara and Mauldin. Georgia Tech.
Ravishankara and Mauldin (29) used a discharge fast
flow apparatus. N03 was detected via long-path laser
absorption at 662 nanometers (run) or via laser-induced
fluorescence. N03 was produced in the main body of a 2.5
centimeter (cm) diameter halocarbon wax-coated flow tube
via the reaction of fluorine (F) atoms with excess HN03:
F + HN03 -> HF + N03. 2.1
A microwave discharge of a 1% F2/He mixture in a Pyrex-
brand side inlet tube produced F atoms, which were
allowed to flow into the flow tube slightly upstream of
the HN03 inlet. This N03 source produces N03 in a
concentration range of (1-30) x 1013 molecules cm'3, with
only N02 as a significant impurity. The reactant, trans-
2-butene, was added through a 0.64 cm diameter movable
injector whose outer surface was coated with halocarbon
wax. The time for reaction between N03 and the butene
could be varied by changing the position of the injector
tip. The minimum distance used was 30 cm (=20
milliseconds). In a given kinetics measurement the N03
concentration was measured at six or seven reaction
distances for each concentration of trans-2-butene.
24


For the system using pulsed laser induced
fluorescence, N03 was produced by the thermal
decomposition of N205 in a side Pyrex injector maintained
at 425 K. A stream of He passing through a N205 trap
maintained at 200 K was passed through this injector to
produce up to 2 x 10n molecules cm3 of N03. This source
also produced 1.5 times as much N02 as N03.
Gas flow was set by mass flow meters and the average
linear flow velocity was around 1300 cm/second. He
diluent gas was used, and trans-2-butene was mixed in He.
Concentration of the reactant was determined from mass
flow rates and the concentration of trans-2-butene in the
stock mixtures. The experiments were conducted at one to
four Torr pressure, and a N03 wall loss of nearly zero was
demonstrated. No mention of temperature control was
provided, and no other information regarding the
construction of the system which could be related to
temperature control was available. The temperature of
the reaction is stated as 298 K in the title, so it can
be inferred that temperature control was in the 1-2 K
range. The reported rate of reaction was (3.780.41) x
10-13 cm3 molecule'1 s'1.
25


National Center for Atmospheric Research
Two N03 reaction kinetics studies were conducted at
the National Center for Atmospheric Research. One was
conducted with an absolute technique, and the other
utilized a technique described later that is neither an
absolute nor a relative rate technique (35,36).
Cantrell, Stockwell, Anderson, Busarow, Perner,
Schmeltekopf, Calvert, and Johnston published a "Kinetic
Study of the N03-CH20 Reaction and Its Possible Role in
Nighttime Tropospheric Chemistry" in 1985 (35). It
apparently is the first report of N03-organics reactions
that considered the importance of temperature. The
reactions were studied in a large glass reactor [6.3
meter (m) length, 445 liter (L) total volume]. Detection
of all reactants, N02, 03, and CH20 [parts per million
(ppm) range], in 700 Torr N2/02 mixtures, was accomplished
spectroscopically by utilizing an internal multiple
reflection system (5.31 m base path) which was adjusted
to 32 passes and a total optical path length of 170
meters. All analyses were based on the infrared bands of
the reactants and products as measured by a Fourier
Transform Infrared Spectrometer (FTIR). N03 could not be
26


measured directly because of its low infrared extinction
coefficient and low concentration, so its concentration
was inferred from the measured rate constant values for
the N03/N2O5 equilibrium and the measured concentrations of
N205 and N02. In a second set of experiments a visible
beam filtered to avoid photolysis of N03 and N02 was used
to measure N03 directly using a differential optical
absorption system (DOAS).
Ozone in oxygen at 5 to 31 ppm was introduced to the
evacuated cell, followed by a measured amount of CH20
vapor was swept in. To initiate the reaction, N02 in dry
nitrogen gas was then thoroughly mixed into the cell at a
final pressure of 700 Torr. Temperature control can be
inferred from tabulated data as 0.1 K per run, but no
further information is presented. The discussion does
not include a discussion of temperature dependence, but
compares data and notes discrepancies between data from
this study and previous studies. The rate of reaction
was found to be (6.3 1.1) x 10"16 cm3 molecule'1 s'1
at 252 C.
In the 19 86 report regarding the N03 reaction with
acetaldehyde (CH3CHO), Cantrell, Davidson, Busarow, and
27


Calvert (36) added N205 in the same concentration range as
the other reactants to the cell described above. An
average temperature is given as 2991 K for five runs,
but the per-run temperature control was not given. The
rate was determined to be (2.1.4) x 10'15 cm3 molecule'1 s'1.
National Oceanic
and Atmospheric Administration
N03 reactions with some atmospheric sulfur compounds
[CH3SCH3 (dimethyl sulfide or DMS) CH3SH (methane thiol) ,
CH3SSCH3 (dimethyl disulfide or DMDS), hydrogen sulfide
(H2S) and sulfur dioxide (S02) ] were studied by
Dlugokencky and Howard (34,37), who also later reported
results of a study of N03 reactions with trans-2-butene,
isoprene, alpha-pinene, and acetaldehyde (13,37). These
studies investigated the temperature and pressure
dependence of the N03-organic reactions. The measured
yields of N02 resulting from the studies of the organics
above decreased with decreasing temperature, indicating
that the rate constants for these four reactions are
temperature dependent. For the atmospheric sulfur
compounds, the rate constant for the reaction with DMS
was found to be temperature dependent but reactions with
28


CH3SH [k= (1.090.17) x 10'12 cm3 molecule'1 s'1] and DMDS
[k=(7.41.5) x 10'13 cm3 molecule"1 s'1] were not. Wall loss
was found to affect the isoprene and alpha-pinene
reactions with N03 at T < 250 K. Table 5.2 in Chapter 5
shows the Arrhenius expressions for the other compounds,
which were found to react with a temperature dependence.
The system used by Dlugokencky and Howard consisted
of two flow tube apparatuses of 2.54 and 4.13 cm inside
diameter (i.d.), which were used to determine the effects
of surface reactions. The reaction zone surfaces of both
flow tubes were coated with halocarbon wax and were
jacketed with circulating silicone oil (T > 298 K) or
ethanol (T < 298 K) from a thermoregulated reservoir to
enable the temperature to be controlled and varied. N03
was produced by the thermal decomposition of N205 at 400
K, and detected by laser induced fluorescence. Total
pressure of the system was about 1 Torr with precise
control, and temperatures were varied between 204 and 384
K, implying 1 K control.
29


Oxford University
Canosa-Mas, et al. published eight papers recently-
regarding N03 reactions with a variety of compounds. The
system they used for most of the studies was a discharge
flow system with a halocarbon wax coated tube heated by
an electric oven regulated electronically to provide
temperatures from room temperature to 523 K. N03 was
produced from the microwave-produced F atom and its
reaction with anhydrous HN03, and was detected by optical
absorption. Low pressures, from 1 to about 5 Torr, were
used with precise control. Injection ports were situated
19.5 cm, 44.0 cm, 70.1 cm, and 95.1 cm (port 1) from the
center of the observation region (a White cell 12 cm in
length).
For the study of the self reaction (38), two
techniques were developed to study slow reactions. The
first technique employed laser flash photolysis to
generate N03 by the reaction of F atoms with HN03, the F
atoms being generated by the photolysis of molecular
fluorine at such a repetition rate that N03 built up to a
steady state concentration. On cessation of photolysis,
N03 was seen to decay over a period of up to 50 seconds.
30


The second technique was a stopped-flow system. A
flow of gas was set up and a steady state of N03 was
observed as in a normal flow system. Then the cell was
isolated, and N03 decay was monitored.
Temperature control is not discussed for either of
the two techniques described above. However, for a
monitoring time of less than one minute, temperature
control can be assumed to be rather precise.
For the study of N03 reaction with OH and H (39) and
with atomic oxygen (40), temperature control is not
mentioned. The study regarding N03 reactions with
chloroform, methanol, HC1, and HBr (32) covered the
temperature range from 298 to 523 K, and all were found
to be temperature dependent with the exception of Hbr,
which was only studied at 298 K.
Canosa-Mas, et al. (18,30,31,33) also studied the
temperature dependencies of the reactions of N03 with
ethene, the most abundant of the unsaturated hydrocarbons
in the troposphere, the alkynes C2H2, C3H4, 1-C4H6, 1-C5H8,
and 1-C6H10, propene, 1-butene, and the alkanes butane, i-
butane, i-pentane, and ethane. In all the reactions,
they found that the variations of the rate constants
31


closely followed the Arrhenius expression, for which they
calculated the parameters A and Ea. A rather thorough
discussion of temperature control was provided. Four
type-K British standard (Ni-Al and Ni-Cr) thermocouples
were attached to the flow tube, two in the middle section
and one at each end of the main heater, and each was
provided with a digital readout. The readings from the
thermocouples were then compared against each other up to
523 K; only at the highest temperatures were differences
larger than 3 K noted. They therefore took the error
limits for the temperature to be 3 K.
The temperature dependent results for propene,
ethene, 1-butene, chloroform, methanol, and the alkynes
and alkanes are presented in Tables 5.1 and 5.2 of
Chapter 5.
Statewide Air Pollution Research Center
Results of two absolute studies were published by
Wallington, et al. (41,42) involving N03 reaction with
numerous sulfur compounds and the equilibrium with N205.
The flash photolysis-long path length optical absorption
apparatus consisted of a 1 m long, 51 mm i.d. quartz tube
32


surrounded by two annular jackets. Distilled water was
circulated through the inner most of these to control the
temperature to within 0.5 K for the range 280-350 K.
Typical residence times were 5 to 15 seconds, so
reactants could be replenished every few flashes. N03 was
produced by photolyzing a molecular fluorine and nitric
acid mixture. N03 radicals were monitored by long
pathlength (8-12 m) optical absorption at 662 nm.
Rate constants for the reactions of N03 with CH3SH,
CH3SCH3, and CH3SSCH3 were determined over the temperature
range 280-350 K at total pressures of 50-100 Torr of
nitrogen diluent, and results can be found on Table 5.2.
Upper limits to the rate constants for reaction with H2S,
S02, and CH3OCH3 were determined at 298 2 K. These were,
in cm3 molecule"1 s'1: 4 x 10"16, 3 x 10"15, and 3 x 10"14
respectively. Reactant concentrations were at least an
order of magnitude greater than the initial N03
concentration of 1-4 x 1012 molecules/cm3.
33


Max Planck Institute
Results of two absolute rate studies using a flash
photolysis, quartz reaction vessel have been published by
Tyndall, et al (43) and Crowley, et al. (44) regarding N03
reaction with dimethyl sulfide and the smallest peroxy
radical, CH302. N03 was produced by photolysis of HN03 to
produce OH, which in turn was reacted with HN03 to produce
N03. Three different temperatures were used in the first
study (278, 298, and 318 K). No discussion of
temperature control was provided for the first study, but
the second study indicated that a water jacket was
utilized to provide temperature control, with no further
discussion. It was determined that the rate constant for
the dimethyl sulfide study showed a slight tendency to
decrease with increasing temperature, but only within the
error limits of their study. The rate that was
determined was (1.00.2) x 10'12 cm3 molecule'1 s'1.
34


Germany and France. Two Labs
A discharge flow mass-spectrometric study of the
reaction between N03 and isoprene was conducted in 1991
(45). The experiments were performed in two different
discharge flow reactors with similar operating
characteristics in two labs. The reactors are made of
Pyrex (2.4 cm i.d.) with side arm tubes and a coaxial
movable inlet, and are coupled to quadrupole mass
spectrometers. The N03 radicals were produced in one side
arm tube using the F-HN03 technique. NO or, in most
cases, 2,3-dimethyl-2-butene, were used as a titrant
before and after each kinetic measurement to determine
the absolute concentration of N03. The analysis of all
species was made by mass spectrometry (MS). In both
systems, the background was efficiently reduced by using
a liquid nitrogen-cooled devise which surrounded the ion
source. N03 was detected at m/e=62. A contribution from
a fragment ion of isoprene was largely reduced by
lowering the electron energy in the ion source.
Experiments were conducted at 298 K, implying a 1 K
temperature control, though no discussion was presented.
Results were k=(7.80.6) x 10'13 cm3 molecule'1 s'1.
35


Relative Rate Studies
Relative rate techniques have produced the majority
of rate data for N03/organics reactions. Using data from
an absolute study, a rate of reaction relative to the
known rate is determined. Concentrations of both
organics are followed, and the ratio kj/k2 is used to
determine the unknown rate constant. The accuracy
depends on how well kj is known, and assumptions that no
unwanted chemistry between the two organics and their
adducts must be made.
There are three areas of considerable uncertainty in
these studies: 1) An aliquot of a N03-N205 mixture is added
to the reaction system. All of the N03 and N205 is
consumed and the reaction system is permitted to come to
equilibrium prior to analyzing a sample of the organics
remaining in the reaction system. Then an additional
aliquot of the N03-N205 mixture is added and the cycle is
repeated. Major losses of N03-N205 may be occurring by
processes other than reactions with the organics. The
reactions are most likely occurring in an inhomogeneous
reaction system. This could lead to side reactions that
would not normally be of importance; 2) The N205
36


concentration at equilibrium in these systems is about
1000 or more times the N03 concentration. In these
studies, a slow reaction with N205 could be interpreted as
a fast reaction with N03; and 3) The N03-N205 equilibrium
is highly temperature dependent. In these relative rate
studies, temperature control is usually stated as 2 K.
If the organic under investigation reacted with N205 to
any significant extent, this lack of temperature control
would make it impossible to obtain any accurate kinetics
information. However, despite these limitations,
relative rate studies are much less expensive and time
consuming to run, and can produce good results.
Statewide Air Pollution Research Center
The majority of kinetic data regarding N03 reactions
with organics has been accomplished at the Statewide Air
Pollution Research Center (SAPRC) of the University of
California at Riverside. Atkinson (46) Atkinson, et
al. (24,47-56), Corchnoy and Atkinson (57), Mac Leod, et
al. (58), Carter, et al. (59), Winer, et al. (60), Pitts,
et al. (23, 61), and Zielinska, et al. (22) have studied
the reactions of N03 with a wide variety of organics.
37


Specific results of these studies can be found in
the original references. Table 2.1 presents rates by
classes of compounds.
Table 2.1: Summary of Rate Constants for N03-0rganics
Reactions [adapted from (37)]
Organic k(298) cm3 molecule'1 s'1
Aldehydes
Alkanes
Aromatics
Phenols
n-Alkenes
Branched Alkenes
Dienes
Terpenes
(5.8-27) x 1CT16
(3.6-13) x 10'17
(3.2-97) x lCT17
(2.5-16) x lO'12
(1.1-97) x 10'16
(3.1-560) x 10'13
(1.0-130) x 10'13
(2.5-14) X 10'12
The reactions were carried out at atmospheric
pressure (735-740 Torr) in about four different systems.
Most studies, however were conducted in the smog chamber
discussed by Winer, et al. (60) and Pitts, et al. (61).
The SAPRC evacuable smog chamber is 5800 L in size
with a surface to volume ratio of 3.4 m"1, with numerous
detection systems. The walls were constructed of Teflon-
lined aluminum, with a temperature control system of heat
exchangers containing ethylene glycol. The temperature
control system was designed to regulate the temperature
38


of the chamber wall to 0.5 K over the 253 to 373 K
range. A YSI Model 71A temperature controller diverts
the circulating fluid through the heat exchangers in
response to a thermocouple signal. The 0.5 K control of
the fluid temperature provided by this system results in
regulation of the chamber air temperature to better than
0.2 K due to the large heat capacity of the chamber.
The walls are insulated with one inch of fiberglass
insulation and two inches of polyurethane foam and are
covered with an aluminum sheath. The non-insulated
window grids are the main sources of temperature
inhomogeneities and are covered with one and one half
inches of polystyrene during experiments that do not
require irradiation. A temperature stratification of up
to 3 K (at the lowest working temperatures) in the
insulated, non-irradiated chamber is reduced to a maximum
of 0.4 K by two 250 L/s stirring fans driven by
magnetically coupled motors mounted on the outside of the
chamber.
Thermocouples are used throughout the chamber to
monitor temperature. Temperature control of a system
this size would be difficult, and repeated injections of
39


the reactants N205/N03 into and subsequent withdrawal for
analytical purposes of organics over a period of hours
could be problematic at best.
Another system was developed recently and utilized
for the study of naphthalene (21). Experiments were
conducted in a 5800 liter evacuable, Teflon-coated
chamber which was equipped with an in-situ multiple
reflection optical system interfaced to a Nicolet 7199
FTIR to monitor reactant concentrations, including N02 and
N205. Temperature control of the chamber to 0.6 K over
the range of 272-297 K was accomplished with an ethylene
glycol refrigeration system. A synthetic air mixture was
used (80% N2 and 20% 02) at 745 Torr pressure. Two sets
of experiments were carried out. The first involved
study of naphthalene, and used thiophene as a reference
material, which reacts with N03, not N205. The second
involved naphthalene, thiophene, and an excess of 2,3-
dimethyl -2 -butene, which reacts very quickly with N03 and
thereby scavenges almost all of it. By utilizing these
two sets of experiments, reaction of naphthalene with N205
could be discerned from reaction with N03.
40


Other Techniques
Morris and Niki conducted a study in the early 1970s
(8) of the reaction of N03 with acetaldehyde and propene,
followed by another study of olefins in 1975 (9) by Japar
and Niki. These studies followed the concentrations of
N205 and N02, and then calculated the N03 concentration
based on the reactions that interrelate the three NOx
species. Such a technique, therefore cannot be termed an
absolute rate study because it does not follow the N03
concentration directly. Likewise, the technique does not
relate the rate of reaction of an unknown species to one
that is known, and therefore cannot be termed a relative
rate study.
Such techniques have some uncertainties that are
similar to relative rate studies, except that only one
organic is being studied at a time. For instance, a
reaction with N205 could not be discerned from one with
N03, and any other reactions would complicate the
interpretation of the results, unless steps are taken to
define the actual reactant. Since the early 1970s the
technique has been developed in Sweden and Italy, and
discussions of these systems follow.
41


Niki and Coworkers
The pioneering work of Morris and Niki and Japar and
Niki in the early 1970s helped set the stage for the
understanding of the importance of the nitrate radical in
the nighttime chemistry of the troposphere. Most
experiments were carried out in a 50 L Pyrex vessel using
400 Torr of 02 and 350 Torr of argon as diluent, at 300 K.
N205 from a dry ice trap at 3-5 mTorr was swept into the
reaction vessel with 02, and the equilibrium between N03
and N205 was used to produce the N03. The concentration of
N03 was controlled by the addition of N02, and if the
decay of the organic was suppressed, this indicated that
N03 was the reactive species, not N205. The rate constants
determined by the two studies (in cm3 molecule'1 s'1) were
1.2 x 10"15 for acetaldehyde, (7.80.8) x 10'15 for 1-
butene, (1.10.1) x 10"13 for isobutene, (1.40.1) x 10'13
for trans-2-butene, (1.80.2) x 10'13 for cis-2-butene,
(5.50.5) x 1012 for 2-methyl-2-butene, and (3.70.5) x
10"11 for 2,3-dimethyl-2-butene. Results for propene and
ethene are on Table 5.1 in Chapter 5.
42


Univ. of Technology
and Univ. of Gotbora. Sweden
One study utilizing N205 as a source of N03 in N2 at
2961 K was conducted by Andersson and Ljungstroem (62).
Infrared spectra were recorded with a Nicolet-MX 1 FTIR
spectrometer coupled to a three mirror White optical
system in an evacuable reactor that was 2 m in length and
0.3 m in diameter. The N205 and N02 concentrations were
monitored and related to N03 concentration by the
interrelated reactions of N205, N02, and N03. A N205 wall
loss was assumed. Four organics were studied, and the
results (in cm3 molecule"1 s"1) were (6.40.3) x 10"15 for
1-butene, (4.40.8) x 10"14 for 1,3-butadiene, and
(1.40.9) x 10"16 for vinyl chloride. The results for
ethene are presented on Table 5.1 in Chapter 5.
Commission of European
Communities. Isora Italy
Nine studies have recently been conducted at the
Joint Research Center, Environmental Institute by a
variety of researchers (27,28,63-69) studies were
conducted in a 480 L Teflon-coated 60 cm diameter
cylindrical chamber equipped with a 81.2 m total beam
43


path length White-type mirror system connected to a
Bruker IFS 113 V FTIR spectrometer. N205 was synthetized
by addition of N02/03, and then the organic was added. N03
could be calculated at any time by following the N203 and
N02 concentrations. Temperature control is indicated as
2 K. The system is used primarily for product
determination via FTIR, gas chromatography (GC)-FTIR, or
GC-MS, though rate constants were obtained for some
reactions.
Purpose of This Research
The purpose of the current work is to develop a
technique that overcomes some of the limitations of the
absolute and relative rate studies. These studies would
provide more reliable information on the kinetics of N03
reactions with organics in a reasonably inexpensive
fashion. Further, the technique needs to define the
temperature dependence and quantify the importance of N205
reactions, allowing the determination of rate constants
or at least bounds for the rate constants of both N03 and
N205 reactions with organics.
The technique described here is neither an absolute
44


nor a relative rate technique. Only one organic is
studied at a time, and the decay of that organic is
monitored by gas chromatography. The concentration
change of the organic being studied is not related to the
concentration change of a reference organic, so it is not
a relative rate technique.
By monitoring the concentration of a reactant other
than N03 in the system, the technique is similar to the
work of Niki, et al. and the other researchers described
above (other techniques). Whereas the N03 concentration
is not monitored directly as is done in absolute rate
studies, it is monitored indirectly by monitoring the
concentration of the organic it is reacting with, and
might best be described as an indirect kinetic technique.
45


CHAPTER 3
EXPERIMENTAL SYSTEM
Reaction Apparatus
and Temperature Control
The experiments were conducted in an approximately
60 liter Teflon film reaction vessel (Figure 3.1)
enclosed in a thermostatically controlled environmental
chamber (Figure 3.2). The National Appliance Model 3518
incubator consists of insulated walls, heater, and fan.
Temperature control was maintained to within better than
0.2 K at temperatures slightly above ambient by a Yellow
Spring Model 72 proportional temperature controller and
monitored by three Type J thermocouples spread out within
the chamber. A centrally located thermistor was used for
feedback to the temperature controller. The three
thermocouples have an independent readout that can be
used either to monitor and record temperatures of the
thermocouples at the start of a kinetics experiment or to
monitor the fluctuation of one thermocouple readout on a
calibrated stripchart recorder during the entire two hour
experiment.
46


FIG
47


FIG. 3.2: Environmental Chamber
48


Reactant Preparation
Known concentrations of the gaseous reactants were
introduced into the reaction vessel by first measuring
the pressure of a pure gas or a mixture of the compound
of interest diluted with nitrogen or helium in a
calibrated Pyrex glass bulb of 1.1 liters. An MKS
Baratron capacitance manometer was fitted to the glass
bulb for measurement of the pressures of the reactants.
The gas measured in this volume was then flushed into the
reaction chamber with a known volume of nitrogen.
Nitrogen flow was monitored using a calibrated Matheson
603 (Model 71314) flow meter. Figure 3.3 shows the
configuration of the N02> organic, and diluent loading
system.
Ozone was introduced into the reaction vessel by
flowing air at a measured flow rate (Matheson 600 flow
meter) through a Monitor Laboratories static discharge
ozone generator for a measured time period (Figure 3.4).
This method of introducing the ozone into the reaction
vessel was calibrated by preparing several ozone in
nitrogen mixtures and then analyzing the amount of ozone
in the vessel with a Bendix Chemiluminescent Ozone
49




FIG. 3.4: Ozone and Oxygen
Loading System


Analyzer calibrated by utilizing a certified Environics
Series 300 Computerized Ozone Analyzer. This latter
instrument has a reproducibility of 1%, concentration
accuracy of 1%, and stability as expressed by a span
drift of <2% of reading in 24 hours and a zero drift
<0.002 ppm per day.
Two vacuum pumps were utilized to control sample
flow and evacuate the bulb and reaction vessel. A Welch
Duo-Seal Model No. 1400 vacuum pump was plumbed to the
inlet of the Pyrex bulb and to the outlets of the organic
and N02 tanks, and was activated with a toggle switch and
used for evacuating the bulb and Teflon tubing to about
one Torr prior to introduction and measuring of the N02 or
organic. This pump was also used to evacuate the tubing
after the requisite volume of N02 or organic was sealed
into the glass bulb. Repeated evacuations and N2 gas
flushing reduced residual organic and N02 in the tubing to
negligible amounts as measured by blank runs, thereby
insuring that known volumes of these reactants were being
introduced into the reaction vessel. The second pump, a
Thomas Industries Inc. diaphragm pump (Model No.
907CA18), ran continuously, and was used to draw sample
52


from the reaction vessel through the gas sampling valve
on the gas chromatograph. This pump also evacuated the
contents of the reaction vessel, thereby facilitating
rapid removal of its contents at the end of a kinetics
experiment and the repeated rinsing of the vessel that
was necessary between kinetics experiments. The Thomas
pump and the gas chromatograph were plugged into a
programmable timer that turned off after two hours at the
completion of the experiment as determined by the lack of
change in organic concentration.
Data Acquisition by
Gas Chromatography
The gas chromatograph (GC) utilized was a Hewlett
Packard 5890A modified with Revision G to accommodate a
Hewlett Packard 3396 Series II integrator. The
integrator was cabled to a standard personal computer
loaded with Chrom Perfect 2 Version 4.06 chromatography
software produced by Justice Innovations, Inc, Palo Alto,
California (Figure 3.5). This software allowed the
storage of raw data, and subsequent integration and re-
integration. Data were compared to calibration runs, and
direct ppm results were recorded along with other
53


FIG. 3.5: GC and Data Acquisition
System


pertinent information. The use of this software allowed
for the enhancement of integration due to the ability to
re-integrate based on the particular characteristics of
an experiment, such as retention time shifts, product
peak growth adjacent to the reactant peak, and in the
case of propene, propane contamination in the original
gas mixture. All data were subsequently stored on floppy
disks for further research purposes. Figure 3.6 is an
example of a chromatogram for an entire seven minute
propene run, and Figure 3.7 shows how the integration was
enhanced by focussing only on the propene peak. Table
A.1 in Appendix A is a report produced by Chrom Perfect.
The GC was equipped with a six port compressed air-
activated gas sampling valve and the Thomas sampling
pump. The sampling pump was configured to draw sample
from the reaction chamber through the gas sampling valve
for two minutes, insuring that a representative sample of
reaction vessel contents was collected. Then the gas
sampling valve switched ports and flushed the gas in the
one milliliter (ml) sample loop onto the column. The air
that controlled the gas sampling valve was diverted by an
electromagnetic switch (mounted on top of the GC) and a
55


FIG. 3.6: Chromatogram of 7 minute Propene Run
PileaB:PROPBNB3.78R Date printed 03-10-1994 Time 18:38:57
1.00 to 7.00 ain. Low Y 0.26000 mv High Y 0.3S000 mv Span 0.09000 nv
FIG. 3.7: Propene Peak and Integration Parameters
FileB:PR0PBHB3.78R Date printed 03-10-1994 Time 18:39:18
3.50 to 6.00 min. Low Y - 0.26458 mv High Y - 0.33305 mv Span - 0.06847 mv
56


solenoid valve (between the GC and the pump) that was
controlled by the GC "purge" control. When the "purge"
was on, the reaction vessel contents flow through the
valve into and through the sample loop. When the "purge"
was off, air was diverted to the other side of the purge
valve, the valve closed and the vessel air flow was
stopped. The trapped sample in the sample loop was then
swept onto the column by the helium carrier gas. The
sampling intervals of five to seven minutes and the
automatic valve were controlled by the computer in the
GC.
A flame ionization detector (FID) was used to detect
both ethene and propene and the reaction products.
Organic compounds are combusted at the temperature of a
hydrogen/air flame, creating charged species. These are
attracted to and captured by a collector, resulting in an
ion current which is then amplified and recorded. The
number of ions produced by the FID is roughly
proportional to the number of carbon atoms in the plasma
of the flame (70) Functional groups, such as carbonyl,
alcohol, halogens, and amines, yield fewer ions or none
at all. In addition, the detector is insensitive to non-
57


solenoid valve (between the GC and the pump) that was
controlled by the GC "purge" control. When the "purge"
was on, the reaction vessel contents flow through the
valve into and through the sample loop. When the "purge"
was off, air was diverted to the other side of the purge
valve, the valve closed and the vessel air flow was
stopped. The trapped sample in the sample loop was then
swept onto the column by the helium carrier gas. The
sampling intervals of five to seven minutes and the
automatic valve were controlled by the computer in the
GC.
A flame ionization detector (FID) was used to detect
both ethene and propene and the reaction products.
Organic compounds are combusted at the temperature of a
hydrogen/air flame, creating charged species. These are
attracted to and captured by a collector, resulting in an
ion current which is then amplified and recorded. The
number of ions produced by the FID is roughly
proportional to the number of carbon atoms in the plasma
of the flame (70). Functional groups, such as carbonyl,
alcohol, halogens, and amines, yield fewer ions or none
at all. In addition, the detector is insensitive to non-
57


combustible gases such as C02, H20, and N0X. The FID
exhibits high sensitivity (10'13 g or pg/ml) large linear
response range (107) and low noise (70) and is rugged
and durable if regularly cleaned and maintained.
However, a FID destroys the sample, and has a low
sensitivity to aromatic compounds, disadvantages that
limit its use in general.
Numerous columns were tested before a suitable one
was found. Initially, a three foot stainless steel
Poropak Q column packed with 80/100 mesh was used. This
column was not suitable because it did not separate
reactants from products, and resulted in slower than
actual rates being determined. A 30 meter fused silica
megabore (0.53mm i.d.) RTX-1 capillary column produced by
Restek Corporation was then used. This column also could
not separate reactants from products due to the volume of
gas instantaneously flushed onto the column by the gas
sampling valve. Next, a Supelco 60/80 (0.5 g/ft)
Carboxen-1000 15' x 1/8" stainless steel (2.1 mm i.d.)
spherical carbon molecular sieve column was tried. This
column also could not be utilized due to its tremendous
separation power. Only ethene eluted from the column
58


within a reasonable time for kinetics data acquisition,
but the severe conditions necessary (maximum temperature
225 C with a maximum helium carrier gas flow rate of
approximately 80 ml/min) created baseline instabilities
that resulted in a signal to noise ratio of one at a
concentration of 1-2 ppm. Finally, a two foot Carboxen-
1000 column (similar to the 15' column described above)
was found to separate products from reactants for both
the propene and ethene studies. Analysis of propene was
conducted at an isothermal oven temperature of 210 C and
a helium carrier gas flow rate of 56 ml/min., while
analysis of ethene was conducted at an isothermal
temperature of 160 C and a carrier gas flow rate of 30
ml/min.
Further investigation of this column, however,
indicated that the column cannot be used for any C4 or
higher hydrocarbons due to the longer than optimal
retention times and broader than desired peaks.
Preliminary work with trans-2-butene with a capillary
column and with the modelling program described later
indicated that the technique described in this paper
should work for this organic also.
59


Reactants
The reactants used include N02/ air, propene, ethene,
and nitrogen. The N02 was supplied by Alphagaz as 995 ppm
N02 with the balance as helium. Propene was supplied by
Scott Specialty Gases at 0.929% with the balance
nitrogen. The ethene was also supplied by Scott at 1040
ppm with the balance nitrogen. The air was breathing
quality and supplied by General Air. Nitrogen was of
standard quality and also supplied by General Air. All
reactants were used without any further purification.
Experimental Procedure
The procedure used in preparing the gas mixtures for
this temperature-controlled technique for studying the
kinetics of N03 and N205 reactions with organics was to
first introduce nitrogen by timing the flow through the
flow meter with a stopwatch into the cleaned reaction
vessel. The vessel was cleaned between experiments by
triple flushing with N2, and then conditioned for five
minutes with high concentrations of an N02/03 mixture (or
in the case of the 03-organic experiments, just 03) A
blank was then run on the GC.
60


The desired concentration of N02 was measured in the
calibrated volume and flushed into the reaction vessel.
Ozone was quantitatively added to the reaction vessel as
described above. Equilibrium was established between N02,
NOj and N205, in 10 minutes for propene and 15 minutes for
ethene, as determined by chemical modelling. There was
always an excess of N02 present in the reaction vessel.
Finally, the requisite volume of the organic was added to
the reaction vessel. Only one organic was used in each
experiment, unlike relative rate studies.
A repetitive sampling sequence using a gas
chromatograph with the two foot packed column and FID was
initiated to measure the time dependence of the
concentration of the organic. The system cycled every
five to seven minutes, depending on the retention time of
the organic being studied. During the first two minutes
of the cycle, a valve was opened which allowed the
sampling pump to draw gas from the reaction vessel and
through the sampling loop on the gas sampling valve of
the GC, insuring that a representative sample of the
atmosphere in the reaction vessel was collected. After
two minutes, the sample was flushed onto the column, and
61


the organic reactants eluted within three to five minutes
of introduction into the column. The decay of the
organic reactants was monitored for a period of
approximately two hours. To control the rate of decay of
the organics it was necessary to adjust the initial
concentrations of N02 and 03 used in the system.
Resulting data were stored on a PC for later
analysis. The system automatically turned off after
about two hours, allowing the analyst to set up and run
an experiment quickly and easily.
Modelling
Experimental data were compared to data generated by
a computer program for modelling complex reaction systems
called "Acuchem/Acuplot" (71). The two executable
programs model spatially homogeneous, isothermal,
multicomponent chemical reaction systems. The output
files were then manually loaded into a spread sheet
program for comparison to the experimental data and for
production of graphics.
62


CHAPTER 4
RESULTS AND DISCUSSION
System Characterization
Characterization of the system entails performing
various tests prior to obtaining actual experimental
results. Characteristics of the system must be well
understood in order to produce meaningful and accurate
results. One such test involved filling the Teflon
reaction vessel and monitoring it over a period of a few
days, to insure that it was "air-tight".
Upon determination of the optimum chromatographic
conditions, blanks were run to determine baseline
conditions. Two point calibration curves were produced
for each organic. Then each organic was reacted to
completion to insure that products did not elute from the
column at the same time as the reactant organic. This
was done for both the 03-propene and N03-propene and
ethene experiments. Propene reacted fast enough for two
ppm to completely react within a couple of days. No peak
was left at the retention time of propene, and two
63


products, one with a retention time slightly less than
propene, were left with what is believed to be propane,
an unreactive contaminant in the propene mixture with a
retention time slightly greater than propene. In the
case of ethene, the reaction with N03 is much slower due
to the instability of the intermediate radical that is
produced. This took over four days to react to
approximately 75% completion, and a product peak was
produced with a retention time near the air/injection
peak. This characterization experiment was then
terminated in the interest of timeliness because during
actual experimental conditions only about 10% of the
ethene reacted, and it was felt that any potential
product other than that which had the retention time near
the air peak could not interfere with results of this
slow reaction.
Another characterization experiment involved filling
the reaction vessel with 30 minutes of N2 at a flow of 110
and 3 minutes of air at a flow of 90, which simulated
actual kinetic experimental conditions, and then
monitoring the chromatographic baseline for the two hour
period equivalent to the kinetic experiments. In this
64


case, no peak grew in, indicating that if a gas such as
water vapor, C02, or air was permeating the Teflon film
from outside to inside (or from higher concentrations to
lower concentrations), no effect on the signal could be
seen.
Prior to each experimental run, a blank was run
using N2 or the conditioning mixture of N2, N02, 03, and
air. This insured that no carry over of reactant was
occurring. However, it appeared that there was a minor
amount of product carry over between runs that was
diluted to negligible quantities upon addition of the N2
gas during the next experiment. This minor product carry
over was determined to not effect the results.
Initial experimental runs were conducted to
determine optimum conditions. These data, used in
conjunction with the initial modelling results, assisted
in the characterization of optimum conditions.
Equilibration time prior to the introduction of the
organic and the initial concentrations of N02 and 03 were
established for use in the actual kinetics experiments.
65


Ozone Wall Loss
An important aspect of the experimental system is
the ozone delivery and wall loss. A five point
calibration curve was used to calibrate the Bendix Ozone
Analyzer (Figure A.l in Appendix A). Three ozone wall
loss experiments were then conducted at various times
through the entire time period necessary to conduct all
of the experimental runs. This was done in order to get
good representation of ozone wall loss. It is assumed
that this wall loss is some first order process, and it
can be used to represent ozone reaction with residuals,
water vapor, and/or permeation out of the Teflon film of
the reaction vessel. The three wall loss experiments
were conducted when the bag was relatively clean, during
kinetics experiments after a normal reaction vessel
rinsing and conditioning, and at the end of experimental
data acquisition. The natural log (In) of the data were
then graphed versus time elapsed (Figure A.2 in Appendix
A), a regression was completed, and the slopes of the
three regressions were then averaged to produce an ozone
wall loss of k=7.67 x 10^ min"1, or 1.28 x 10'5 s'1. This
wall loss rate was then used in all subsequent modelling.
66


Correction Factors
During the N03-ethene experiments, the rates of which
were very slow, a phenomenon was noted in the data that
could not be explained as a result of N03-ethene kinetics.
A series of two hour calibration runs were performed
where 2 ppm of ethene was introduced into the regular
N2/air mixture, and the GC sampling sequence was
initiated. These calibration runs were conducted at two
different temperatures, and after one week of "cleaning"
the Teflon bag with a high concentration of 03 and
repeated flushing with N2.
The phenomenon can best be described as an apparent
slight increase in reactant concentration for 50 minutes
followed by a steady decrease for the remaining 70
minutes. No reactive species were introduced into the
bag, just N2 and air. Temperature change had no effect,
and neither did cleaning the bag with high concentrations
of 03. There appeared to be a slight peak that grew in at
the retention time near the air peak that could not be
explained by permeation through the Teflon film of C02,
H20, NOx, etc. The negligible amount of product carry
over also had no effect, or the data from the experiment
67


after cleaning the reaction vessel would have shown
little change in reactant concentration for this
condition. One possibility that was disproved was the
permeation out of the bag by the organic. At a higher
temperature, a higher wall loss would be expected due to
the increased frequency of collisions of the organic
molecule with the wall. However, there was no
significant difference in the data from the experiments
carried out at 30 C and 45 C.
One explanation might be that some surface chemistry
is occurring in conjunction with a physical phenomenon
such as delayed mixing or a systematic problem with the
sampling. If not thoroughly mixed, the concentration may
increase slightly after introduction into the vessel.
This may explain the increase of 0.1 ppm seen in the
propene calibrations and the 0.02 ppm increase in the
ethene concentrations. GC signal drift also cannot be
ruled out.
The data from the calibrations from about 50 minutes
to two hours, however, show a steady decrease for both
propene and ethene of approximately 0.2 ppm. Surface
chemistry of some sort may be occurring that decreases
68


the concentration of the organic being studied. This
could be investigated by utilizing new Teflon film or
another material entirely to see if the phenomenon is
related to the surface of the reaction vessel.
The approach used in this research, however, was to
use the two hour calibration data to correct the
experimental data. The first data point occurs at four
minutes from introduction of the organic into the
reaction vessel. This point was used as the "true"
concentration, and all subsequent data points were
normalized to the average of the two data points from the
two calibration runs, upon which the difference from the
true concentration was determined. These differences
were then averaged to determine the correction factor for
each data point at a given time. To verify this process,
the calibration runs were then corrected, and the
corrected data for each calibration run were compared to
the normalized data (Figure A.3 in Appendix A). The
variation in the corrected data was within +/-0.03 ppm
for the entire two hours, and within +/-0.015 ppm for the
first 80 minutes. Therefore, the error in the correction
factors was less in the first 80 minutes and increased
69


after that. However, this error is well within the
experimental error of the technique itself.
General Model Results
Model results for each of the experiments and for
determining the wall loss are discussed in the
appropriate sections that follow. The chemical mechanism
used in this modelling is shown in Table 4.1. The model
was utilized to determine equilibration time prior to
introduction of the organic, and to determine appropriate
starting concentrations of N02 and 03. For instance, to
maximize the rate of the N03-ethene reaction, it was
determined that 8 ppm N02 and 4 ppm 03 produced more N03
than the conditions utilized for the propene experiments
of 18.3 ppm N02 and 4 ppm 03. This is because of the
reversible reaction 1.3, which shifts to the left at
lower N02 concentrations.
General characteristics of the pertinent reactions
of the system under study are provided by the model. The
03 concentration decreases rapidly during the first five
minutes or so due to the rapid reaction with N02
producing N03. Following that, N205 is the main source of
70


Table 4.1: Chemical Mechanisms Used to Describe the
Reactions of N03 and N205 with Propene and Ethene
Reaction Rate Coefficients/ Rate Constant cm3 molecule'1 s'1 Ref .
4.1) O3 + NO2 > NO3 + O2 1.2 X 10'13 exp (-2450/T) 72
4.2) N03 + N02 (+M) -> N205 (+M) * 72
4.3) n2o5 -> no2 + no3 72
4.4) 03+ N02 -> NO + 2 02 9.7 x 1019 35
4.5) N03 + NO -> 2 N02 1.5 X 10 " exp (170/T) 72
4.6) no3 + N02 -> NO + N02 + 02 4.5 X 10'14 exp (-1260/T) 72
4.7) N03 + NO3 -> 2N02 + 02 8.5 x lO'13 exp (-2450/T) 73
4.8) O3 + NO > NO2 + O2 2.0 x 1012 exp (-1400/T) 72
4.9) N20j -> WALL 8.33 x 103 s'1 + this work
4.10) 03 -> WALL 1.28 x 10'3 s'1 + this work
4.11) N03 -> NO + 02 2.5 x 10-6 exp ( -6100/T) 1 s'1 73
4.12) 03 + PROPENE -> PRODUCTS 6.5 x 10'13 exp (-1900/T) 72
4.13) 03 + ETHENE -> PRODUCTS 1.2 x 10'14 exp (-2630/T) 72
4.14) NO, + PROPENE -> PRODUCTS 1.81 : ic lO'13 exp (-980/T) this work
4.15) N03 + ETHENE -> PRODUCTS (0 .813-2.03) x 10'16 this work
* Rate determined from Kof K and m contained in Table
2 of the referenced document.
** Rate determined from k^, for Reaction 4.2 = 4.0 x 10'27
exp (10930/T) and k^ = k(fonvard) [k(reverse)]
+ Wall loss rate is the average of the determined bounds
as described in the text. Units are first order and
are in s'1.
71


N03, as it reacts reversibly and also forms N03. N205, on
the other hand, increases in concentration for about 30
minutes, and then decreases. N02 decreases in
concentration for about 30 minutes, and then increases.
This has the effect of decreasing the N03 concentration
further. Between this, the wall loss of N205, and the
other minor reactions of the system, the N03-organic
reaction is essentially completed in about two hours.
An example of the computer model output is included in
as Table A.2 in Appendix A.
Determination of Wall Loss
The experimental data produced by this temperature-
controlled technique were averaged and compared to
modelling results. The model results predict a larger
loss of organic than is observed experimentally. This is
probably due to not accounting for a wall loss for N03
and/or N205 in the model. These losses are assumed to be
some first-order process that removes one of these
reactive species without it reacting with the organic.
Since the kinetics of the N03 reaction with propene have
been studied directly using an absolute technique (18),
72


the rate constant (k) for this reaction is as well known
as it is for any N03 reaction.
An N03 wall loss was incorporated into the mechanism
which provided the best match between the experimental
and model results at 30 C. Figure A.4 in Appendix A
demonstrates the fit of a N03 wall loss at an N03-propene
rate of 7.2 x 10'15 cm3 molecule'1 s'1 (10.62 ppm'1 min'1) .
This same process was repeated assuming that only a N205
wall loss should be used in the model. Determining which
of the wall losses, or a combination of both processes,
correctly describes the system is of crucial importance
in demonstrating the utility of this technique.
The aspect of this technique that allows for
differentiation between N03 and N205 reactions is related
to the large temperature dependence found for the N03 +
N02 <-> N205 equilibrium. The experiments are performed
under conditions in which N205 concentration at
equilibrium may be about 1000 times the N03 concentration.
Since this equilibrium has a very strong temperature
dependence, the N03 concentration increases dramatically
(approximately doubling) for only a 10 degree temperature
change, while the N205 concentration remains essentially
73


unchanged. If the wall loss is due primarily to N03 loss,
then increasing the temperature would increase the N03
concentration, which would increase both the rate of
reaction with propene and the wall loss of N03. If, on
the other hand, the wall loss is due primarily to N205,
increasing the temperature would again increase the N03
concentration. This would increase the N03 reaction with
propene, but the N205 concentration would remain
essentially unchanged, hence the wall loss would not
increase significantly. The comparison at various
temperatures between the experimental results and model
results is much better when the assumption is made that
the wall loss was due to N205, not N03, as demonstrated by
Figures A.5 and A.6 in Appendix A. This suggests that
the N203 wall loss is the only significant loss process
under these experimental conditions. Included on Table
4.1 is the average wall loss rate that was determined by
this study. The wall loss of N205 is discussed further in
the following section regarding temperature dependence.
74


N205 Reaction with Propene
In addition to the obvious temperature effects on
the N02 + N03 <-> N20j equilibrium discussed previously
that creates faster decay rates for propene, the possible
N205-propene reaction was investigated using the model.
All other conditions were kept the same except that
instead of N03 reacting with propene, N205 was set up to
react. Based on the model results, the rate constant for
this reaction was determined to be k=1.36 x 10'18 cm3
molecule'1 s'1 (0.002 ppm"1 min"1) at 30 C and 6.1 x 10' 18 __-3 cm
molecule'1 s'1 (0.009 ppm'1 min'1) at 45 C. This is at least
a factor of 1000 slower than the no3 reaction rate, but
indicates there is an unreasonable temperature dependence
for this kind of bimolecular reaction.
The model was then used to determine what effect, if
any, adding an N205-propene reaction to the N03-propene
reaction would have. At a rate of 1.3 6 x 10"19 cm3
molecule'1 s'1 (0.0002 ppm'1 min'1) a noticeable increase in
the overall propene decay rate indicates that in reality
the N205-modelled propene reaction rate is much less. In
fact, the rate was less than 1.36 x 10'20 cm3 molecule'1 s'1
(2 x 10"5 ppm"1 min'1) which places an upper bound on the
75


N205 reaction with propene.
Propene Reaction Results and Discussion
Prooene and Ozone Reaction Results
In order to test the operation of the system, some
preliminary studies of the reactions of ozone with
propene were performed. The only processes involved in
these studies were the reaction with the organic and the
loss of ozone to the walls of the reaction vessel
discussed previously. The first-order rate constant for
the ozone decay was determined to be 1.2 8 x 10'5 s'1 (7.67
x 10-4 min'1) This rate constant was used in all
subsequent modelling studies.
Figure 4.1 shows a plot of the time decay of the 03-
propene reaction data [natural log (In) of ppm] with the
calculated In ppm from the data regression. The propene
decay is not linear and hence not first order under the
conditions utilized (2 ppm 03 and 1 ppm propene) Ozone
concentrations in great excess of propene would be
necessary to verify this as a first order process, but
the reaction would be too fast to monitor, indicating
again the importance of modelling.
76


FIG. 4.1: OZONE-PROPENE REACTION DATA
LN PPM PROPENE VS. TIME
TIME. MINUTES
LN PPM ---- CALCULATED LN PPM
Comparison of Ozone-Propene
Experimental and Model Data
The experimental data were compared to data
generated by the model under a variety of conditions.
The three factors which were varied in the model to
produce a best-fit were temperature, 03 concentration,
and the 03-propene rate constant.
Figure 4.2 shows the effect of varying temperature.
The effect of a 15 change in temperature noticeably
changes the slope of the decay of propene, but not the
end point of the reaction. This effect is slight,
however, with respect to the variation in the data.
77


FIG. 4.2: OZONE-PROPENE REACTION AT 30C
EFFECT OF TEMP. ON RATE. TIME. MINUTES
a EXPERIMENTAL DATA ----- k-.018 (UT.) P TEMP-35. k-.020
TEMP-45, k.0244
Figure 4.3 shows the effect of reducing the rate
constant about 10%, 20%, and 33%. This has a lesser
effect on the slope, but does produce a better fit with
the experimental data. In fact, the rate that is a third
less than the literature value produces the best fit of
this approach with the experimental data.
The best fit to experimental conditions, however, is
demonstrated by using the literature value for the rate
constant and varying the ozone concentration (the symbol
[] designates concentration of the species inside). This
is shown on Figure 4.4.
78


PPM PROPENE PPM PROPENE
FIG. 4.3: OZONE-PROPENE REACTION AT 30C
EFFECT OF VARYING RATE. a EXPERMENTAL DATA
TIME. MINUTES
k.018 (UT.) A k-.016 A k-.014 X k-.012
FIG. 4.4: OZONE-PROPENE REACTION AT 300
EFFECT OF CHANGE OF OZONE CONCENTRATION
TIME. MINUTES
EXPERIMENTAL DATA ----- k-.018(UT.)[03]-1 O OZONE-0.9PPM
A OZONE-1.1 PPM X 0Z0NE-0.8PPM
79


The less steep slope fits the data better, and
matches the spread in the experimental data exactly. The
static discharge-type ozone generator produces large
quantities of ozone from air. To produce 2 ppm ozone
requires only 15 seconds of activation of the generator
with a toggle switch. Based on these data, it appears
that there is a systematic error pf -0.1 ppm in the
calibration of ozone, which corresponds to 0.75 s, and
that the reproducibility of ozone generation is 0.1 ppm.
Propene and NO? Reaction Results
The temperature-controlled technique for studying
the kinetics of N03 and N205 reactions with organics was
performed by mixing N02 with 03 in nitrogen and allowing
this mixture to equilibrate before adding propene to the
mixture. Figure 4.5 shows the results of the experiments
at the four temperatures. Spread In ppm data are
presented along with the calculated In ppm data from the
data regression at each temperature. The slope of the
calculated data increases with increasing temperature,
and at the higher temperatures the data become much less
linear. This is what is expected for increasing
80


LN PPM PROPENE
FIG. 4.5: N03 PROPENE REACTION DATA
LN PPM VS. TIME
TIME, MINUTES
EXPERIMENTAL DATA + CALC. LN PPM AT 30C O CALC. LN PPM AT 35C
A CALC. LN PPM AT 40C
X CALC. LN PPM AT 45C


temperature conditions and for the N03-propene reaction,
as the N03 concentration approximately doubles for only a
10 degree temperature change, while the N205 concentration
remains essentially unchanged. More N03 results in a
faster rate of reaction with propene, which results in
steeper-sloped lines for propene decay.
Comparison of N03-Propene
Experimental and Model Data
Experimental data were compared to data generated by
the model utilizing a variety of conditions. Data from
the 30 C temperature experiments were used to investigate
varying ozone concentrations and the effects of three
reactions that are felt to be relatively insignificant.
Rate constants for the propene decay and the wall loss
from previous research using this technique were utilized
as a starting point in the determination of best fit of
model data to experimental data. Attention was then
focussed on the data collected at other temperatures. As
will be seen, it became necessary to assume that the N03-
propene reaction is temperature dependent. Further, to
create a best fit at all temperatures, there appears that
there is a temperature dependence to the wall loss rate
82


also. Finally, since the data at the 35 C temperature do
not fit into the N03-propene rate and wall loss continuum
established by the data from the other three
temperatures, they will be discussed separately.
Effect of ozone concentration, temperature, and reactions
4, 7, and 11. Figure 4.6 shows the starting point in the
comparison of experimental data to model data. The model
rate of k=1.2 x 10'17 cm3 molecule'1 s"1 (10.62 ppm'1 min"1) and
wall loss rate of 2.75 x 10-4 s'1 (.0165 min'1) pairing
produce data that fit well with the experimental data.
Figure 4.7 shows the best fit rate constant value
changed by a factor of 2. All data fit well within that
range. Figure 4.7 also shows the effect of the
systematic error in ozone production. The data indicate
that less ozone slows the decay rate of propene down
slightly, and may contribute to the experimental
variation in the data.
83


PPM PROPENE
FIG. 4.6: N03 PROPENE REACTION AT 30C
DATA VS STARTING RATE AND WALL LOSS(WL)
TIME, MINUTES
EXPERIMENTAL DATA
k=10.62, WL=.0165


PPM PROPENE
FIG. 4.7: N03 PROPENE REACTION AT 30<
EFFECT OF [03] & RATE(k)
TIME, MINUTES
EXPERIMENTAL DATA ---- k=10.62, [03]=4PPM O k=5.31 A k=l
[03]=3.8PPM V [03] = 3.9PPM
X


Figure 4.8 shows the effect of temperature variation
of 0.2 C. This small effect is well within the error of
the overall experimental technique. This figure also
shows the effect of equations 4.4, 4.7 and 4.11 found on
Table 4.1. The overall effect of zeroing these three
equations is to speed up the rate of propene decay.
Whereas this effect is minimal, the three equations
enhance the accuracy of the comparison of the model data
to the experimental data, and thereby enhance the overall
accuracy of the technique itself.
The temperature dependence of the N03-propene reaction
rate. The data from 30, 40 and 45 C exhibit a
temperature dependence. In addition, to achieve best
fits at all three temperatures, a temperature dependence
for the wall loss also had to be assumed. With two
variables and only one data set, two approaches to
comparison of model and experimental data were taken.
The experimental data from the best three runs at each
temperature were averaged at each sampling point to
assist in the comparison to the model data.
86


PPM PROPENE
FIG. 4.8: N03 PROPENE REACTION AT 300
EFFECT OF TEMP AND 3 RXNS=0
TIME. MINUTES
EXPERIMENTAL DATA ---- k=10.62, T=30C O k=10.62. T=29.8C
A k= 10.62, T=30.2C
X RXN 4.4.4.7.&4.1 1 =0