United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
EPA-600 3-79-020
March 1979
Research and Development
oEPA
Reactions of Oxy
Radicals in the
Atmosphere
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology.' Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-79-020
March 1979
REACTIONS OF OXY RADICALS IN THE ATMOSPHERE
by
D. G. Hendry, R. A. Kenely
J. E. Davenport, and B. Y. Lan
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
Grant No. 603864
Project Officer
Marcia C. Dodge
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor
does mention of trade names or commercial products constitute endorse-
ment or recommendation for use.
ii
-------
ABSTRACT
This study has been composed of two facets: the atmospheric
chemistry of peroxyacetyl nitrate (PAN) and the products of the reaction
of OH with aromatic compounds which are known to be in the atmosphere.
PAN dissociates rapidly in the atmosphere according to the reaction
CH3C(0)OaN02 —*- CH3C(0)02» + N02
From measurements of a temperature range of 25-39°C, the rate constant
may be expressed by
log k = 16.29 - 26,910/4.576T
The acetylperoxy radical formed can react with N02 to regenerate PAN
CH3C(0)02» + N02 —*- CH3C(0)02N02
or react with NO
CH3C(0)02« + NO —>-CH3C(0)0» + N02
The ratio of rate constants of these two reactions, kv,n/kvin > is 3.0 ± .7
independent of temperature. The rate constants for reactions with N02
and NO are estimated to be 1 x 109 and 3 x 109 M"1 s"1, respectively.
The reactions of OH with simple aromatic hydrocarbons proceeds by
two major reaction channels, abstraction of a hydrogen atom (k , ) in
the benzylic position or addition of OH to the aromatic ring (k ,).
For toluene, the first route primarily yields benzaldehyde, and the
second route yields a mixture of cresols. The values ofk,/(k,+k.)
ab ab ad
for toluene, 1,4-dimethylbenzene, and 1,3,5-trimethylbenzene are 0.15 ±
iii
-------
0.02, 0.15 ± 0.02, and 0.02 + 0.006, respectively. Formation of
m-nitrotoluene, which has been observed in some smog chamber experiments
at high N02 concentrations, results from the OH adduct reacting with
N02 and eliminating H20. At ambient concentrations of N02, this reaction
is unimportant.
The reaction of OH with benzaldehyde, an important product of the
toluene reaction in the atmosphere, results in an exclusive attack at
the aldehydic position. Therefore, the initial stable intermediate is
the benzoylperoxy radical, which can react with either NO or N02. Reaction
of the radical with NO causes loss of C02 and leads directly to the
phenyl radical C6H5*. In our system, this radical produces phenol; how-
ever, in the atmosphere, it should yield ring cleavage products.
This report was submitted in fulfillment of grant number 603864 by
SRI International under the partial sponsorship of the U.S. Environmental
Protection Agency. This report covers a period from June 8, 1975, to
June 7, 1978, and work was completed as of June 7, 1978.
iv
-------
CONTENTS
ABSTRACT iii
FIGURES vi
TABLES vii
1. INTRODUCTION 1
2. CONCLUSIONS AND RECOMMENDATIONS 3
3. GAS PHASE FREE RADICAL REACTIONS OF PEROXYACETYL NITRATE . 5
Introduction 5
Experimental Section 6
Results and Discussion 11
Conclusions 34
4. GAS PHASE HYDROXYL RADICAL REACTIONS. PRODUCTS AND
PATHWAYS FOR THE REACTION OF OH WITH AROMATIC
HYDROCARBONS 39
Introduction 39
Experimental Section 40
Results 42
Discussion 43
5. GAS PHASE HYDROXYL RADICAL REACTIONS. PRODUCTS AND
PATHWAYS FOR THE REACTION OF OH WITH BENZALDEHYDE ... 53
Introduction 53
Experimental Section 54
Results 55
Discussion 60
REFERENCES AND NOTES 63
-------
FIGURES
Number Page
1. U-Tube for Collection of PAN 8
2. Fraction of PAN Remaining ([PAN]t/[PAN]O) Versus Time for
Decomposition of 1.72 x 10"~" M in the Presence of 4.1 x
10-" M NO at 25°C 17
3. Ln(kQks) Versus 1/T (K-1) for Decomposition of PAN in the
Presence of NO 20
RCHO
4. Observed First-Order Rate Constants, ko^s for Decomposition
of PAN at 25°C in the Presence of: CH3CHO With and Without
Added 02; and C2HSCHO With and Without Added 02 21
CHOI
5. Observed First-Order Rate Constants, ko^s 3, for the
Decomposition of PAN at 55°C in CHC13/CC1/, Solutions .... 24
NO 2
6. Observed Rate Constant, ko^s, Versus [N02] for Decomposition
of PAN in the Presence of Added N02 at: 25°C, Slope =
(0.064 ± 0.015) x 10~" M-1 s~a; Intercept = (0.015 ± 0.05) x
ID"" s-1; 34.3°C, Slope = (0.21 ± 0.04) x 10~" M"1 s"1;
Intercept = (0.045 ± 0.15) x 10~" s~l; and 43.9°C, Slope =
(0.55 ± 0.07) x 10~" M-1 s-1; Intercept = (0.030 ± 0.4) x
10-" s-1 26
7. Concentration of N02* (o - o), PAN + PAN (•-•), PAN
(A - A), and PAN* (o- -o) Versus Time for Decomposition of
0.78 x 10-" M PAN in the Presence of 3.15 x 10~" M 1SN02 at
25°C 28
NO N02 _i
8. Reciprocal Corrected Rate Constant (ko^ - k^g) Versus
[N02]/[NO] for Decomposition of PAN in the Presence of Added
N02 and NO at: 25°C, 34.3°C, 39.0°C, and 43.9°C 36
vi
-------
TABLES
Number Page
1. Observed Rate Constants (kQ^s) for PAN Decompositions in the
Absence of Added Reactants at Various Temperatures ..... 9
NO
2. Observed First-Order Rate Constants, k hs for Decomposition
of PAN in the Presence of NO at Various Temperatures ...... 18
3. Selected Infrared Absorption Frequencies (v) for PAN,
13N-Labelled PAN, N02, and 13N02 .............. 29
4. Kinetic Parameters for Decomposition of 0.78 x lO"** M PAN in
the Presence of 3.15 x KT* M 15NO, at 25°C ......... 30
5. Rate Constants and Reaction Conditions for Decomposition of
PAN in the Presence of Added NO and N02 at Various
Temperatures ........................ 35
6. Least Squares Data for Figure 8 and Calculated Values of
ka/kt for Decomposition of PAN in the Presence of Added N0a
and NO at Various Temperatures ............... 37
7. Distribution of Individual Products as a Function of [N0a]
Added for Reaction of OH With Toluene, Plus 9.7 x 10 1
-------
TABLES (Cont.)
Page
14. Wall and Gas Phase Products of the OH-PhCHO Reaction as a
Function of Added N02 57
15. Field lonization Mass Spectral Analysis of Wall Residue for
OH-PhCHO Reaction 58
viii
-------
1. INTRODUCTION
Our understanding of the atmospheric phenomenon of photochemical
smog has evolved gradually since the late 1930s, when an air pollution
problem was recognized in the Los Angeles air basin. The development
of our knowledge can be broken down into three distinct phases. In
the first phase, up to the mid-1950s, a qualitative understanding was
established on how the interaction of hydrocarbons and NO with sunlight
led to smog formation, and what some of the effects of smog on the
environment were.
During the next 15 years (about 1955-1970), research centered on
identifying many of the basic characteristics of smog, including the
NO -03-sunlight equilibrium, hydrocarbon structural effects on ozone
X
formation and eye irritation, and major intermediates contributing to
the effects associated with the overall process.
Finally, since about 1970, we have worked to understand the chemistry
of smog formation at a molecular level. The key concept has been the
realization of the dominant role of OH radical in reactions of hydro-
carbons. The key tool has been the computer, which has allowed us to
combine information on individual reactions to determine whether the
overall concentrations of observed intermediates could be accurately
simulated. Kinetic data, which are absolutely necessary for the computer
calculations, were determined for many of the reactions during this
time. Using computer simulation, it is now possible to determine which
reactions are the most crucial and to subject them to further study.
Thus, the research effort can now be limited to the most crucial areas,
which should lead to an accurate description of photochemical smog
formation that can be used to reliably predict formation of intermediates
such as N02, PAN, and ozone under environmental conditions.
The work reported here considers two important aspects of photo-
chemical smog, both of which are essential to a quantitative understanding
-------
of the chemistry of the atmosphere. The first is the chemistry of
peroxyacyl nitrate (PAN), a compound that was identified in the Los
Angeles atmosphere in early studies. We show that this compound is
not a chemically inert species, as had been thought, but one that can
have a significant effect on the overall chemistry of smog because it
exists in equilibrium with the radical intermediates from which it is
formed.
0 0
SCOO • + N02 T-*- CH3Ci
CH3COO • + N02 T->- CH3COON02 (PAN)
Thus PAN can act as a radical sink or source, depending on the concen-
trations of the peroxyacetyl radical and N02 in the atmosphere, and can
thereby have a large effect on the atmospheric chemistry.
The second important factor considered in this report is the
chemistry of aromatic hydrocarbons with atmospheric reactants. The
report includes an investigation of the products of reactions of OH
with aromatic hydrocarbons using a low conversion flow system in
which OH radicals are produced chemically at low concentrations.
Although data on the rate constants of OH reactions with aromatics
existed previously, these are the first data on the initial products
of these reactions. Since benzaldehyde is a major product of the
toluene reaction, we have also investigated the reaction of benzalde-
hyde with OH.
-------
2. CONCLUSIONS AND RECOMMENDATIONS
The results of this work show that PAN is not chemically stable in
the atmosphere but is in equilibrium with acetylperoxy radicals and N02
according to the reaction
kf
CH3C(0)02N02 -rV CH3C(0)02 • + N02
The dissociation occurs with a lifetime of 30 minutes at 25°C. At high
[N02]/[NO] ratios PAN is regenerated following each dissociation
of a PAN molecule; however even at [N02]/[NO] = 3 a PAN molecule
is regenerated only one out of two dissociations. This is due to the rapid re-
action with NO
CH3C(0)02« + NO—-^- CH3C(0)0» + N02»
*
which has a rate constant three times faster than the reaction with N02.
When PAN does react with NO it will generate radicals which initiate
additional hydrocarbon consumption and NO oxidation. Thus PAN plays an
important role in the overall chemistry and especially has a large effect
on ozone formation.
Our results on the facile reactions of OH radical with aromatic
hydrocarbons show that both oxidation of the ring methyl group and
hydroxylation of the ring occur . Thus both substituted benzaldehydes
and phenols are the major initial products from the aromatic hydrocarbons.
The reaction of benzaldehyde itself with OH is at the aldehydic hydrogen,
leading to peroxybenzoyl nitrate and phenol.
We make two general recommendations with regard to our study.
First, the fact that PAN can play an important role in controlling the
formation of smog raises the question as to whether other peroxynitrates
can also be important. Therefore a study of the chemistry of other
-------
peroxynitrate compounds should be carried out in order that chemical
models that describe photochemical smog formation may be accurate in
this regard. Second, the identification of the initial products of the
reactions of aromatic compounds with OH under atmospheric conditions
is only the first step in understanding the effect of these compounds
in the overall smog formation chemistry. Thus the work must be extended
to determine the fate of various products that are formed at each stage.
-------
3. GAS PHASE FREE RADICAL REACTIONS OF PEROXYACETYL NITRATE
INTRODUCTION
The two principal reactions of peroxy radicals in the troposphere
are reaction with N02 (reaction 1) and with NO (reaction 2).
R00« + N02 > ROON02 (1)
ROO • + NO -^V RO • + N02 (2)
Reaction 2 is recognized as the primary means by which NO is oxidized
to N02 and is thereby crucial in establishing elevated levels of ozone
in the environment.1'2 Reaction 1, which is less well recognized, is
significant in two respects. First, the products of the reaction, peroxy
nitrates (ROON02), are themselves noxious pollutants. The best example
is peroxyacetyl nitrate (CH3C(0)OON02), PAN,3 which is a frequently
observed constituent of photochemical smog,1*4"12 and a known lachrymator13
and phytotoxicant. 6 »lt+»1 5 Second, reaction 1 scavenges peroxy radicals,
trapping them as peroxy nitrates and preventing the radicals from re-
acting with NO. However, the peroxy nitrates act as a radical sink only
to the extent that they do not undergo homolytic decomposition (reactions
3 and 4):
ROON02 —*- ROO* + N02 (3)
ROON02 —*- R0« + N03 (4)
To determine whether the generalized reaction 3 could play a
significant role in the chemistry of polluted urban atmospheres,'we
have begun an investigation of the gas phase free radical reactions of
-------
peroxy nitrates.16 This section presents in detail the results of our
studies concerning the kinetics and mechanisms of the homolysis of PAN.
Also described is the use of PAN as a source of radicals to determine
ki and k2, the rate constants for the reactions of acetylperoxy radicals
with N02 and NO, respectively.
EXPERIMENTAL SECTION
Materials
Peroxyacetic acid was prepared from acetic anhydride, H2SOz,, and
90% H202. As described by Swern,17 the reactants were combined by
stirring at 0°C and then allowed to warm to room temperature overnight.
The resulting solution contained 61% peroxyacetic acid and was used
without further purification.
Peroxyacetyl nitrate (PAN) was prepared as suggested by Louw et
al18 and Stephens3 by the direct nitration of peroxyacetic acid. Thus,
61% peroxyacetic acid, 0.5 g (4.1 x 10~3 mol), in 30 ml reagent grade
pentane was stirred at 0 to -5°C under argon. To this mixture, 30%
S03, 2.3 g (8.8 x 10-3 mol) was added dropwise, followed by 90% HN03,
0.28 g (2 x 10~3 mol). After 15 minutes the reaction mixture was washed
three times with water, then dried with MgSOj,. The ir spectrum of this
solution showed the 1835, 1735 cnr1 peaks characteristics of PAN. On
the basis of the absorbance of the 1735 cur1 peak and the extinction
coefficient at this frequency,3 the yield of the reaction was estimated
to be 39%.
Gas phase samples of pure PAN were obtained by preparative scale
gas liquid partition chromatography (glpc) using the method of Stephens
et al.19 Analysis of PAN purified in this manner showed it to be
contaminated with approximately 0.4% pentane. The last traces of pentane
could be removed by chromatographing the sample a second time. In
kinetic runs, control experiments assured us that the traces of pentane
remaining in once-chromatographed PAN had no detectable effect on
observed reaction rates.
-------
N02 (Matheson) and 1SN02 (Stohler Isotope Chemicals) were purified
to remove NO by combining them with 02 and allowing the mixtures to stand
15 minutes before degassing at -196°C.
NO (Matheson) was passed through Linde 13X mole sieves, degassed at
-196°C, and distilled from a liquid oxygen cooled bulb. Acetaldehyde
and propanal (Aldrich) were purified by fractional distillation under
argon and were degassed before use.
Apparatus
Three different gas phase ir cells were used. Each was constructed
from a 9.0 cm x 1.5 cm i.d., jacketed Pyrex tube, the ends of which were
formed from 0-ring joints. 0-rings and metal brackets were used to
secure NaCl plates to the ends of each vessel. Vacuum stopcocks
(lubricated with a minimum amount of halocarbon grease) and ground-glass
or 0-ring joints on each vessel permitted evacuation of the cell and
admission of reactants.
Cell II differed from the other two in that its interior surfaces
were coated with Teflon. Cells I and III did not differ significantly.
As noted in the Results and Discussion Section, cells I and II were used
to follow some of the decompositions of PAN in the absence of added
reactants. With these exceptions, cell III was used for all the gas
phase experiments reported here. Samples of purified PAN were collected
in the U-tube depicted in the U-tube depicted in Figure 1. By connecting
the 0-ring joint, Jl, to the glpc effluent, it was possible to collect
PAN at -196°C between stopcocks S2_ and :S3 and to collect He in the bulb
(volume = 28.4 ml) between Sl_ and S2. When warmed to ambient temperature,
PAN could be admitted into an evacuated vessel by way of J2. Reactants
in addition to PAN could be added to the system by S4 and J3. Opening
S^L allowed He to enter the vessel as well, resulting in a total pressure
of approximately 500 torr.
Helium was used as diluent gas in most of the reactions. The ex-
ceptions were a few PAN decompositions in the absence of added reactants
to which 1 atm of air was used as diluent (see Table 1).
-------
CHROMATOGRAPH
EFFLUENT
-Q>=®=(
TO
REACTION
VESSEL
J3
REACTANTS
SA-4466-6
FIGURE 1 U-TUBE FOR COLLECTION OF PAN
-------
Table 1. OBSERVED RATE CONSTANTS (kQbs) FOR PAN DECOMPOSITIONS IN THE
ABSENCE OF ADDED REACTANTS AT VARIOUS TEMPERATURES
Temperature
(°C)
25
48.2
57.6
62.5
67.9
72.5
77.2
82.2
[PAN]O
x 104 M
0.809
1.23
1.29
1.41
3.35
2.64
2.83
1.18
0.603
0.711
0.442
0.480
0.390
0.635
0.558
0.780
0.407
0.49
0.660
0.472
0.405
0.405
k u
obs
x 10" s"1
0.015
0.018
0.011
0.010
average =
0.014 ± 0.0032
0.28
0.30
0.18
0.18
0.16
average =
0.22 ± 0.06
0.51
0.77
1.9
1.7
1.8
1.7
average =
1.8 ± 0.1
2.9
5.5
6.8
average =
5.1 ± 1.6
5.9
5.8
12
12
Cellb
III
III
III
III
III
III
II
II
II
I
II
I
II
I
I
II
I
I
I
I
I
I
Diluent
Gas
He
He
He
He
He
He
He
He
He
Air
He
Air
He
Air
Air
He
Air
Air
Air
Air
Air
Air
92.5 0.405 29 I Air
*a
k , = (In 2)/tj5, where t, is observed half-life for decomposition
of pure PAN (see text).
See Experimental Section for description of Cells I through III.
9
-------
Temperature control was maintained to within ± 0.1°C with a Haake
constant temperature circulating bath. Reaction temperatures were
measured with a calibrated thermocouple inserted into the jacket sur-
rounding a reaction vessel.
A Perkin-Elmer Model 467 spectrophotometer was used for all ir
analyses. The spectrophotometer was coupled to an external recorder for
kinetic measurements.
Procedure
All concentrations of gaseous reactants, except for NO, were
determined by ir analysis. Extinction coefficients (e) for PAN given
by Stephens3 are: E183S = 22.4 x 103 M"1 nr1; c173s = 53.0 x 103 M"1
m"1. Values of e for N02 were measured for reaction vessels with and
without added diluent. First, N02 was added to an evacuated ir cell and
the pressure was measured. The concentration of N02 was calculated from
the pressure and the value of the equilibrium constant for the N02/N20i,
equilibrium at 25°C (K = 106 torr).20 Next, the absorbance at 1618 cm"1
was measured with the cell thermostated at 25°C, giving eieia = 6.1 x
103 M'1 nr1. Finally, 500 torr of He diluent was added to the cell and
the absorbance at 1618 cm"1 was recalculated. The value of e obtained
under these conditions was Eieis = 21.7 x 103 M"1 m"1. The 1590 cm"1
extinction coefficients for 1SN02 were assumed to be equal' to the values
measured for N02 at 1618 cm"1. To determine the extinction coefficients
for CH3CHO, we measured amounts of CH3CHO in a gas buret and transferred
them into a reaction vessel with a Toepler pump. The vessel was pressurized
to 500 torr with He, and the absorbances were measured at 2750 and 1770
cm"1. Values of e were found to be: £2730 = 2.84 x 103 M"1 m"1; ei77o
= 9.53 x 103 M"1 m"1. C2HSCHO was assumed to exhibit values of e
identical to those for CH3CHO. Amounts of NO were determined with a gas
buret and transferred to an evacuated reaction vessel with a Toepler
pump.
For the reaction of 15N02 with PAN, kinetics were determined in the
following manner. 15N02 was added to the evacuated cell and its concentration
10
-------
was measured. The cell was thermostated and placed in the spectrophotometer,
PAN and He were added, and the 1900 cm'1 to 1500 cm-1 region of the
spectrum was immediately scanned. This region was then repeatedly
scanned at approximately 10-minute intervals. Spectrophotometer output
was registered on a continuously moving remote recorder and the change
in absorbance of any reactant could be measured as a function of time.
Concentrations of PAN used in the gas phase experiments^typically
ranged from 0.5 to 2.5 x 10-A M. For a 25 ml reaction vessel, it was
therefore never necessary to handle more than approximately 6 x 10"6
moles of the potentially dangerous21 PAN vapor. Moreover, these con-
centrations were well below the vapor pressure of PAN, reported3 to be
15 to 20 torr (8 to 10-" M at 25°C).
Solution phase decompositions were performed using samples of PAN
prepared as described in the Materials section but substituting CC1<, for
pentane as solvent. Samples were purified by column chromatography
(silica get at 0°C) and weighed into nuclear magnetic resonance (nmr)
tubes. CHC13 and t-C<,H9OH (internal standard) were also added and
weighed as needed. The tubes were then degassed, sealed, and placed in
a thermostated bath. Kinetics were measured by withdrawing the samples
from the bath at time intervals and recording the nmr spectra.
RESULTS AND DISCUSSION
The thermal decomposition of peroxynitrates can proceed by two
possible pathways. These are 0-0 homolysis (reaction 4) and 0-N homolysis
(reaction 3), the reverse of the reaction by which peroxynitrates are
formed. In the case of PAN itself, homolytic bond scission (reaction -1)
produces acetylperoxy radicals and N02.
k1
CH3C(0)00-N02 ^^CH3C(0)00« + N02 (-1,1)
Because acetylperoxy radicals are not prone to unimolecular decomposition,
they are relatively long lived, and reaction 1, the reverse of the homolysis,
11
-------
must also be considered. In effect, this requires that 0-N homolysis
involve a dynamic equilibrium between PAN, acetylperoxy radicals, and
N02. Net decomposition of a molecule of PAN could occur only after
destruction of an acetylperoxy radical by a subsequent reaction with a
species other than N02.
CH3C(0)00- + S -^V products -N (5)
For the case of 0-0 homolysis, which is typical of peroxides, the initial
products of the bond cleavage are acetoxy radicals and N03.
CH3C(0)0-ON02-CH3C(0)0. + N03» (6)
Decarboxylation of acetoxy radicals is unimolecular.
CH3C(0)0--^C02 + CH3. (7)
This reaction occurs fast enough22"24 (k7 = 10* s"'1) to preclude bi-
molecular reactions. Thus, the reverse of reaction 5 need not be
considered, and decomposition of PAN by 0-0 homolysis could not involve
an equilibrium between PAN and radical species.
To probe the precise mechanism of PAN homolysis, we have investigated
the rate of thermal decomposition of PAN in the absence and presence of
free radical scavengers. In the absence of scavengers, decomposition of
PAN by reaction 6 requires that the disappearance of PAN obey strictly
first-order kinetics. If 0-N homolysis predominates, decomposition of
pure PAN will proceed by reactions -1, 1, and 8.
2CH3C(0)00» ke> 2CH3» +2C02+ 02. (8)
In this case, the steady-state expression for disappearance of PAN
contains quadratic terms and the observed rate of decomposition need not
be first-order in [PAN].
12
-------
When PAN decomposes in the presence of a radical scavenger, S, the
rate of disappearance of PAN by reaction 6 will be identical to the rate
observed in the absence of the scavenger. That is, the presence of a
scavenger will affect neither reaction 6 nor reaction 7. On the other
hand, if decomposition involves the equilibrium -1,1, scavengers can
intercept acetylperoxy radicals (reaction 5), shifting the equilibrium
to the right and enhancing the observed rate of PAN decomposition.
Decomposition of Pure PAN
The decomposition of pure gaseous PAN was investigated over the
temperature range 25 to 92.5°C. Reproducible rates of decomposition
could be obtained at any temperature or in any of three reaction cells,
provided the vessels had been "seasoned" by repeated PAN decompositions.
Decomposition rates were found to deviate significantly from first-order
kinetics. Semilogarithmic plots of the fraction of PAN remaining
([PAN] /[PAN]-) versus time typically exhibited pronounced upward
curvature. Curvature was evident in all decompositions performed, and
the degree of curvature could not be correlated with temperature or
initial reactant concentration. As discussed above, this type of behavior
is consistent with 0-N but not with 0-0 cleavage being the primary
mechanism for PAN homolysis.
Except for the 25°C experiments, decompositions were followed to
at least 50% of completion and observed half-times (ti ) for decomposition
were obtained directly. At 25°C, semilogarithmic plots were extrapolated
to 50% decomposition to find values for t, . From the half-times, values
of k , , observed rate constant for decomposition of pure PAN, were
calculated from the expression: k , = (In 2)t, . These k , values
(see Table 1) have the dimensions of a first-order rate constant but do
not represent a true first-order process because of the curvature in
t
the logarithmic plot.
From Table 1 it is evident that reproducible rates of PAN decomposi-
tion could be obtained using seasoned reaction vessels made either of
Pyrex (cells I and III) or Teflon-coated Pyrex (cell II). This suggests
13
-------
that the nature of the seasoned vessel surface is not important in
determining the rate of PAN decay.
Moreover, a plot (not shown) of ln(k , ) versus reciprocal absolute
ODS
temperature is linear (slope = 12.36 ± 0.34 x 104, intercept = 27.79 ±
1.0 s"1, ra = 0.984) over the entire 25 to 92.5°C temperature range.
Products of PAN decompositions at various temperatures were studied
by ir and glpc. Analysis by ir showed peaks due to C02 (2320 cm"1),
CH3ON02 (1680, 1660, 1290, 1020, 850 cm-1),25 and inorganic nitrate
(1360 cm"1).26 Analysis by glpc confirmed that CH3ON02 and CH3N02 were
present in approximately equal amounts. This type of product distribution
was seen for all reactions in which He was the diluent gas. With air as
diluent, no CH3N02 was observed, and C02, CH3ON02 and N03 were the major
products.
CH3ON02, CH3N02, and C02 are indicative of formation of CH3 by
either 0-0 cleavage, followed by reaction 7, or 0-N cleavage, followed
by reaction 8. Reactions 9 through 11 explain the formation of CH3N02
and CH3ON02.27-29
CH3 . + N02 >- CH3N02 (9)
CH3. + N02 +- CH30. + NO (10)
CH30« + N02 *- CH3ON02 (11)
If N03 is present from 0-0 cleavage, an additional possible source of
CH3ON02 is
CH3• + N03 —*- CH3ON02 (12)
1
In the presence of large amounts of 02> reactions 9 and 10 are replaced
by
CH3 • + 02 *- CH300» (13)
14
-------
2CH300» *- 2CH30« + Oa (14)
With He as diluent, small amounts of 02 are formed and reactions 9 and
13 could compete, but in the presence of air, only reaction 13 would
occur, eliminating CH3N02 as a reaction product.
Inorganic nitrate was produced on the reaction cell windows in all
PAN decompositions. Surface decomposition of PAN is a possible, but not
very probable, source of N03 , since surface effects on the rate of PAN
decay could not be demonstrated. More likely, N03 results from surface
reactions of CH3ON02 and N02. For CH3ON02 produced by PAN decomposition
at 25°C in a control experiment, the reaction of CH3ON02 with surface
adsorbed moisture
CH3ON02 > CH3OH + H+ + N03 (15)
was shown to occur. Analysis by glpc clearly showed that the CH3ON02
was unstable under the.reaction conditions, decomposing slowly to give
CH3OH. Inorganic nitrate was also produced under these conditions, but
we did not quantitatively measure the amount of N03 formed relative to
the amount of CH3ON02 lost. The possibility of nitrate production from
N02 was directly demonstrated by simply admitting N02 to a reaction cell,
allowing it it stand at 25°C for several hours, and observing an increase
in the absorbance of the 1360 cm"1 peak. Here, nitrate production could
be explained by the following reactions:
2N02 T*~ M-n- (16)
HONO + H+ + N03~ (17)
NaCl *~ NOC1 + Na+ + N03 (18)
15
-------
Decomposition of PAN in the Presence of NO
The effect of radical traps on the rate of PAN decomposition was
first investigated using NO as the scavenger. Decomposition rates were
determined at temperatures ranging from 25 to 39°C and for [NO]/[PAN]
ratios varying as much as ten-fold. Semilogarithmic plots of [PAN] /[PAN]n
versus time (see Figure 2) were linear in the initial stages of the
reaction but exhibited upward curvature as the reaction progressed. The
onset of curvature depended on the [NO]/[PAN] ratio, curvature appearing
sooner at low [NO]/[PAN]. Half-times for PAN decompositions were obtained
from the initial linear regions of the semilogarithmic plots (extrapolating
NO
when necessary) and were used to calculate values for k , the observed
first-order rate constant for decomposition of PAN in the presence of
NO
NO. Experimental conditions and k , values are summarized in Table 2.
Infrared spectra of the products of the decomposition of PAN in the
presence of NO revealed the presence of N02 in addition to C02 and
CH3ON02. CH3ONO, CH3NO, and CH3N02 may also have been present, but
interference from N02 and CH3ON02 absorbances precluded positive identifica-
tion of these species. :
Comparison of the data in Tables 1 and 2 shows that NO has a
pronounced effect on the rate of PAN decomposition. At 25°C the rate
constants for PAN decomposition are 1.4 (± 0.3) x 10~6 s"1 in the
absence of NO and 3.7 (± 0.4) x lO"** s~l in the presence of NO. At the
NO
same time, the values of k , are independent of the initial [NO]/[PAN]
ratio.
Since NO can have no effect on the rate of 0-0 bond scission,
reaction 6 cannot explain the accelerating effect of NO. On the other
hand, the data are entirely consistent with reversible 0-N cleavage as
the primary mechanism for PAN homolysis. To account for the effect of
NO on the kinetics of the PAN decomposition, we apply a steady-state
analysis to reactions -1, 1 and 5. Thus, noting that reactions 5 and 2
are identical if S = NO and R00« = CH3C(0)00«, we obtain:
-d[PAN]/dt = [PAN]k_l{k2[NO]/(k1[N02] + kz[NO])} (19)
16
-------
1.04
0.9
0.8
0.7
0.6
<
o_
<
a.
0.5
0.4
0.3
1000
2000
TIME
3000
seconds
4000
5000
SA-4466-1
FIGURE 2 FRACTION OF PAN REMAINING ([PAN],/[PAN]0) VERSUS TIME FOR
DECOMPOSITION OF 1.72 X 1(T* M PAN IN THE PRESENCE OF
4.1 X KT4 M NO AT 25°C
17
-------
NO
Table 2. OBSERVED FIRST-ORDER RATE CONSTANTS, k FOR DECOMPOSITION
OF PAN IN THE PRESENCE OF NO AT VARIOUS TEMPERATURES
k
Temperature fPAN] obs
(°c) M x iok [NO]/ [PAN] x 10** s~l
25 1.72 2.4 3.96b
1.89 9.2 3.98
1.37 1.9 " 3.21
3.60 5.6 4.11
0.99 0.16 _3._25
average =
3.70 ± 0.39
29.9 1.97 2.5 8.82
1.98 4.8 7.67
1.36 7.2 7.91
average =
8.13 ± 0.50
34.3 1.10 1.5
1.46 16.
1.21 3.6
average =
14.0 ± 1.3
39.0 1.58 5.6 28.8
1.44 0.68 28.5
1.29 8.5 , 29.6
average =
29.0 ± 0.5
kNO
. , determined from linear region of log [PAN] ,./ [PAN] versus
obs ., -,_^ / ____^\ L Jt L Jo
Data of Figure 2.
time plot (see text).
b
18
-------
When k2[NO] » ki[N02], equation 19 reduces to:
-d[PAN]/dt = [PAN]k_x = [PAN]kN° (20)
Here, the rate of PAN decay is first-order in PAN and zero-order in NO,
as is observed in the initial stages of the reactions. However, as the
\
decomposition proceeds, NO is oxidized to N02 by reaction 2, and eventually
the point is reached at which ki[N02] is no longer negligible compared
NO
with k2[NO]. In this case, equation 20 fails and k , is less than k_i
ODS
by the factor k2[N0]/(kj[N02] + k2[NO]).
NO
Figure 3 is an Arrhenius plot of the average k , values taken from
ODS
Table 2. Because all the data in Table 2 correspond to the case where
NO
k2[NO] » ki[NOa], equation 20 holds and k , equals k_x at each temperature.
For this reason, the slope and intercept of Figure 3 can be used to derive
the activation parameters for 0-N homolysis; thus, k , (s"1) equals
(1016'2' ± °'60) esp(26910 ± 900)/G, where 6 equals 2.303 RT in calories
per mole.
Decomposition of PAN in the Presence of Hydrogen Atpa^Donors
To establish the general nature of reaction 5, we followed PAN
decompositions at 25°C in the presence of added hydrogen atom donors.
The intent was to demonstrate an accelerated rate of PAN decomposition
and also to provide direct evidence for the intermediacy of acetylperoxy
radicals by trapping them as peroxyacetic acid.
The two hydrogen atom donors most extensively investigated were
CH3CHO and C2HSCHO. PAN decompositions were followed at 25°C in the
presence of excess aldehyde ([RCHO]/[PAN] ranging from 1.05 to 27) with
either He or oxygen as diluent. Semilogarithmic plots of [PAN] /[PAN]
versus time were linear for at least 70% reaction, and half-times for
RCHO
PAN decay were used to calculate k , , the observed first-order rate
constant for the decomposition of PAN in the presence of aldehyde. Values
RCHO
of k , are plotted versus concentration of aldehyde in Figure 4.
19
-------
-5.5
-6.0
il
JC
NJ
O
-8.0
3.20
3.25
3.30
3.35
'1
SA-4466-5
FIGURE 3 ln(kNOs) VERSUS I/T (K~1) FOR DECOMPOSITION OF PAN IN THE PRESENCE OF NO
Slope = (1.35 ± 0.05) X 104. Intercept = 37.44 ± 1.38.
-------
[ALDEHYDE), M X 104
SA-4466-7
FIGURE 4 OBSERVED FIRST-ORDER RATE CONSTANTS, kgg"0 FOR DECOMPOSITION
OF PAN AT 25°C IN THE PRESENCE OF: CH3CHO WITH (»-») AND
WITHOUT U-A) ADDED 02; AND C2HgCHO WITH (•-•) AND WITHOUT
(0-0) ADDED 02
21
-------
Products of the decompositions in the presence of aldehydes were
shown by ir to be C02, alkyl nitrates, and nitroalkane, which was
observed only in the absence of added oxygen. Peroxyacetic acid was
not observed to be a product of the decompositions, although small
amounts of the peracid might have gone undetected because of interference
from aldehyde absorption bands.
A mechanism that is consistent with these results involves the
equilibrium -1,1 and reaction 5 where S equals RCHO. For this mechanism,
the rate of disappearance of PAN is governed by equation 21.
-d[PAN]/dt = [PANlk.Jlo[RCHO]/(k?CHO[RCHO] H-kxtNOz])} (21)
RCHO
Equation 21 predicts that when ks [RCHO] » ki[N02], PAN decomposition
will follow equation 22
-d[PAN]/dt = [PAN]k_t = [PAN]kR™° (22)
Figure 4 shows that the rate of PAN decomposition becomes zero-order
in [aldehyde], in apparent agreement with equation 22. However, the
RCHO
maximum values obtained for k , (approximately 2 x lO"*1 s~l) are
significantly lower than the value of k_i = 3.7 ± 0.39 x lO"41 s-1
determined for PAN decomposition in the presence of NO. Also, adding
RCHO
02 to the PAN/RCHO system is seen to lower the value of k , relative
obs
to the value obtained with He diluent.
These observations can be rationalized by expanding the mechanism
for the PAN/RCHO system to include reactions 23, 7, and 24.
RC(O)- + N02 *~ RC(0)0- + NO (23)
RC(0)0» *~ R» + C02 (7)
RC(0) • + 02 *~ RC(0)00« (24)
22
-------
If reaction 23 does occur, NO Is produced and reaction 2 must also be
considered. In this case the rate of disappearance of PAN is given by
equation 25.
-d[PAN]/dt = [PAN]k.l{(k3RCH°[RCHO] + kz[NO]/
(k5 [RCHO] + k2[NO] + k1[N02])} (25)
RH
When k2[NO] greatly exceeds k3 [RCHO] but not ki[N02], equation 25
becomes
-d[PAN]/dt = [PAN]k_1{k2[NO]/(k2[NO] + k![N02])} = [PAN]k"u (26)
In this manner, it is possible to explain the zero-order dependence of
RCHO RCHO
k , on [RHCO] and the discrepancy between k , and k_i. When 02 is
added to the system, reaction 24 30 occurs to the exclusion of reaction 23
RCHO
and the rate of PAN decay adheres to equation 21. The value of k3
is so low (k5 - 103 M"1 s"1 for solution phase aldehyde autoxidations
at 0°)31 that concentrations of aldehyde leading to equation 22 could
not be acheived under gas phase conditions. For the particular case of
the PAN/CH3CHO/02 system, reactions 5 and 24 yield acetylperoxy radicals
that can reform PAN by reaction 1. Thus no decrease in [PAN] was
observed during the time when this reaction was studied.
The effect of hydrogen atom donors on the rate of PAN decomposition
was also investigated in solution at 55°C. Here CCli, served as solvent
and CHC13 as the hydrogen atom donor. Nuclear magnetic resonance was
used to monitor the rate of disappearance of 5 to 8 x 10~2 M PAN in the
presence of varying [CHC13]. Semilogarithmic plots of [PAN] /[PAN]
versus time were linear for at least 85% reaction. Observed half-times
were used to find first-order rate constants, k , 3, for decomposition
in the presence of CHC13. These values are plotted versus [CHC13] in
CHC1
Figure 5. For increasing [CHC13], k . 3 is seen to increase, in
accordance with equation 21, where CHC13 is substituted for RCHO.
23
-------
[CHCI3], M
SA-4466-3
CHCI
FIGURE 5 OBSERVED FIRST-ORDER RATE CONSTANTS, k^ 3, FOR THE
DECOMPOSITION OF PAN AT 55°C IN CHCI3/CCI4 SOLUTIONS
Slope = (2.75 ± 0.45) X 10"5 M~1 s"1. Intercept = (1.66 ± 0.84) X ID"5 s~1.
24
-------
Products of the decomposition of PAN in CClj, containing no CHC13
were found (by nmr, ir, and ms) to be C02, CH3C1, CH3ON02, and CH3N02.
For solutions containing CHC13, the products were C02, CH3Cl, CH3N02,
Cm, CH3OH, and CH3C(0)OOH. The presence of the peracid was clearly
indicated by an nmr singlet at 128 cps (downfield from IMS) and ir
absorption bands at 1785 and 1445 cm"1.
These results demonstrate that PAN generated acetylperoxy radicals
and indicate that reactions -1, 1 and 5 are important for the decomposi-
tion of PAN under a variety of conditions.
Decomposition of PAN in the Presence of N02 and 13N02
To further characterize the mechanism of PAN homolysis, we performed
a series of decompositions in the presence of added N02. It was an-
ticipated that N02 would suppress the rate of PAN decomposition by
increasing the importance of reaction 1 relative to reaction -1. Con-
trary to these expectations, N02 in some cases actually enhanced the
rate of PAN decomposition. Semilogarithmic plots of [PAN] /[PAN] versus
time exhibited varying degrees of curvature and, as was done for pure
PAN decompositions, observed half-times were used to calculate k , 2,
the rate constants for PAN decomposition in the presence of N02. Values
of k , 2 obtained at three temperatures and at [N02]/[PAN] ratios ranging
O DS
from 1 to 14 are plotted versus [N02] in Figure 6.
We attribute these results to the presence of NO in the PAN plus
N02 system. The effect of NO would be to accelerate the rate of PAN
decomposition, as described above. A possible source of NO is surface
decomposition of N02 by reactions 16 through 18. These reactions were
previously postulated to explain the formation of N03 on reaction cell
windows. In addition to inorganic nitrate, HONO and NOC1 are products
of reactions 17 and 18, respectively. These products could be converted
to NO by reactions 2732 and 28.33
NOC1 + H20 —*• HONO + HC1 (27)
25
-------
NJ
[NO31. M X 104
SA-4466-4
NO.,
FIGURE 6 OBSERVED RATE CONSTANT, k^2, VERSUS [NO2] FOR DECOMPOSITION OF PAN IN THE PRESENCE
OF ADDED N02AT: 25°C (•-•), SLOPE = (0.064 ± 0.015) X 10"4 M~1 s~1; INTERCEPT = {0.015 ± 0.05)
X 10"4 s'1; 34.3°C (A-A). SLOPE = (0.21 ± 0.04) X 10"4 M'1 s~1; INTERCEPT = (0.045 ± 0.15) X 10"4 s'1;
AND 43.9°C (0-0), SLOPE = (0.55 ± 0.07) X 10"4 M"1 s'1; INTERCEPT = (0.030 ± 0.4) X 10"4 s'1.
-------
2HONO Vall> NO + N02 + H20 (28)
An alternative explanation regarding the PAN/N02 system is that side
reactions such as 17, 18, 27 and 28 are not involved and that the rate
enhancement caused by N02 is an indication of the failure of reactions
-1,1 and 6 to describe accurately the mechanism of PAN homolysis. To
eliminate this possibility, we studied the decomposition of PAN at 25°C
*
in the presence of added N-labeled N02 (N02). Our intent was to
demonstrate that the equilibrium -1, 1 and any surface reactions involving
NO would occur simultaneously. The side reactions would be indicated
by an increase in the rate of disappearance of total PAN relative to the
rate of disappearance of PAN in the absence of N02. The equilibrium
-1,1 would be directly demonstrated by trapping acetylperoxy radicals
with N02 to form 15N-labeled PAN (PAN ).
*
The method used to study the PAN/N02 system was to repeatedly scan
the 1900 to 1500 cm"1 region of the ir spectrum at timed intervals.
This technique is based on the different ir absorption frequencies
exhibited by labeled and unlabeled PAN and N02 (see Table 3). By
measuring the absorbances at 1835, 1734, 1696, and 1590 cm"1 at different
times, it was possible to determine the concentrations of total PAN
* *
(PAN + PAN.) and N02, respectively. Concentrations of N02 could also
be measured using the 1618 cm"1 absorption frequency, but these values
*
were inaccurate because of interference from the strong N02 absorbance
centered at 1590 cm"1.
Results for a typical experiment are shown in Figure 7. From the
figure, it is evident that the concentration of total PAN decreases.
ft
By extrapolation of the line for [PAN] + [PAN ], the half-time for
decomposition of total PAN is found to be 1.73 x 10** s. Tliis equates
to a rate constant of 4.0 x 10~5 s"1, in fair agreement witli the value
of k , \ = 1.9 x 10~5 s"1 predicted from Figure 6 for decomposition of
PAN at 25°C in the presence of 3.15 x 10"'' M unlabeled N02. However,
in addition to the disappearance of total PAN, a much more rapid exchange
reaction is occurring, which causes the simultaneous disappearance of
27
-------
3.1
3.0
x
5
O 2.8
lO
2.6
0.8
•a 0.6
X
5
Z
- 0.4
0.2
O
NO*
PAN + PAN'
1000
2000 3000
TIME — seconds
4000
5000
SA-4466-2
FIGURE 7 CONCENTRATION OF NO* (O-o). PAN + PAN* {•-•), PAN (A-A), AND PAN*
(0--0) VERSUS TIME FOR DECOMPOSITION OF 0.78 X lO^M PAN IN THE
PRESENCE OF 3.15 X lO^M 15N02 AT 25°C
28
-------
Table 3. SELECTED INFRARED ABSORPTION FREQUENCIES (v) FOR PAN,
15N-LABELED PAN (PAN*), N02, AND 1SN02 (NO?)
Compound v, cm
PAN
PAN*
N02
*
N02
1835, 1735
1835, 1696a
1618
1590b
a
Reference 34.
Reference 35.
PAN and N02 and production of PAN . Though it is not shown in the figure,
N02 was produced at a rate parallel to that of N02 disappearance. Clearly,
this exchange can only result from the combination of acetylperoxy radicals
*
(formed by reaction -1) with N02 and N02:
PAM < » rH3r(n)nn« + N02
ki
(29)
* K-j *
PAN "7~»" CH3C(0)00» + N02
Applying the steady-state approximation to the concentration of acetyl-
proxy radicals, the kinetic expression for the exchange reaction is:
+d[PAN*]/dt = -d[N02]/dt
= k_i(([PAN] + [PAN*])([NO*.]/[N02] + [NO*]) - [PAN*]} (30)
Figure 7 shows that at any time, t, a tangent to the curve for N02
disappearance has slope = (-d[NO;!]/dt) . Similarly, it is possible to
* *
directly determine the values [PAN ]t and ([PAN] + [PAN ])t. As noted
previously, it is difficult to measure [N02] accurately. Therefore, to
29
-------
f
find the value for [N02]/([N02] + [N02]) , it is assumed that the N02
*-a
formed at any point in time equals the N02 consumed, which gives:
[N02]/([N02] + [N02])t = |N02]t/[N02]Q
(31)
Substituting equation 31 into equation 30 and rearranging gives:
(-d[N02]/dt)
{([PAN] + [PAN ])t([N02]t/[N02]0) - [PAN ^
(32)
Equation 32 and Figure 7 were used to calculate values for k_i for six
time points during the exchange reaction. Individual values of k_i,
the average of the six values, and the kinetic parameters used in the
determinations are summarized in Table 4.
Table 4. KINETIC PARAMETERS FOR DECOMPOSITION OF 0.78 x lO"" M PAN IN
THE PRESENCE OF 3.15 x ICT* M 15N02 AT 25°Ca
Time, s
600
900
1200
1500
1800
2100
(-d[NO*]/dt)t
M s-1 x 108
2.8
1.9
1.3
0.83
0.58
0.58
([PAN] + [pAN*])t-
f( r*N°2L/L*N09]n)
I r *u u
- LPAN Jj
0.57
0.49
0.42
0.36
0.32
0.28
average =
. b
k-i
s-1 x lO"
4.9
3.9
3.1
2.3
1.8
2.1
3.0 ± 1.1
Date of Figure 7.
k_i calculated from eq 32, see text.
The exchange experiment was repeated a total of nine times, and six
individual values of k_i were found for each experiment. The average
value of all 54 individual rate constants is 4.0 ± 1.0 x 10~" s~l, in
30
-------
,— *• _ —1
excellent agreement with the value of k_i:= 3.7 ± 0.39 x 10 s
obtained from the experiments with NO.36 Because the decompositions in
*
the presence of NO and N02 proceed at the same rate, they must share a
common rate determining step, and hence a common mechanism. Of the
possible mechanisms considered for PAN homolysis only, equilibrium -1,1
*
can satisfactorily explain both the NO and the N02 results. Thus, we
consider the mechanism involving 0-N bond cleavage to be valid. The N02-
enhanced decomposition is attributed to reactions 16 through 18, 27 and
28, which accompany the equilibrium 1,-1 under our reaction conditions.
To test the possibility that 0-0 homolysis (or any other reaction)37
could also accompany 0-N bond scission, we set as an upper limit for the
rate of reaction 6 the rate constant for decomposition of PAN in the
absence of added reactants. At 25°C, therefore, reaction 6 can constitute
no more than 0.4% of the total pathway for PAN homolysis. To explain
the preference for 0-N relative to 0-0 cleavage in PAN homolysis, it is
necessary to consider the changes in the enthalpies and entropies of
reactions -1 and 6. Using established heats of formation when available
and calculated values when necessary,2^ >38-ltl it is estimated that both
reactions -1 and 6 are endothermic by the same amount, 26 ± 2 kcal/mol"1.
This value is in good agreement with the activation energy determined
experimentally for the PAN/NO system. Entropy changes for the two
reactions are estimated to be AS_° = 42 ± 2 and AS° = 30 ± 2 cal deg-1
mol"1 (1 atm standard state). To estimate the effect of the calculated
entropy change on the rate of reaction 6, it is helpful to refer to rate
data for acetyl peroxide and dinitrogen pentoxide homolyses (reactions
33 and 34):
CH3C(0)0-0(0)CH3 *- 2CH3C(0)0- (33)
02N-ON02 *- N02 + N03 (34)
For reactions 33 and 34, the entropy changes and Arrhenius pre-
exponential factors are:42
31
-------
AS33 = 33.4 cal deg'1 mol-1, A33 = 101"'23 s
and
ASa<, = 35.1 cal deg-1 mol-1, A3<, = 10
By analogy, the preexponential factor for reaction 6 is estimated to be
A6 = lO1* s"1. Because the preexponential factor for reaction -1,1 is
observed to be A_i = 1016*2' s"1, entropy effects could indeed dictate
the 250-fold difference in the rate constants for reactions -1,1 and 6.
The calculated value for AS_X is useful in estimating k^. If it
is assumed that the radical-radical combination reaction (reaction 1) has
zero activation energy, then
ki = A! (35)
Also, from transition state theory,24
A_1/A1 = exp(As21/R) (36)
Therefore, AS°i (IMstandard states) and the observed value for A_x can
be used to calculate:
ki = A_1/exp(AS_1/R) = 1.0 x 109 M"1 a~l.
Decomposition of PAN in the Presence of NO and N02
As discussed in the Introduction, the value for the rate constant
ki is important in terms of the atmospheric chemistry of peroxy radicals.
Of equal importance is the value of k2, the rate constant for reaction
of peroxy radicals with NO. To determine a value of k2 for acetylperoxy
radicals, we have measured the rate of decomposition of PAN in the
presence of known concentrations of added NO and N02.
32
-------
In the PAN/NO/N02 system, equation 19 should apply. When NO and
N02 are present in comparable amounts that exceed the amount of PAN, the
disappearance of PAN should follow first-order kinetics. Defining k x
as the observed first-order rate constant for decomposition of PAN in
the presence of added NO and N02, equation 19 becomes
NO
-d[PAN]/dt = [PANlk.akalNOl/UalNO] +k1[N02])} = [PAN]k"J* (37)
However, this treatment neglects the fact that N02 in excess of PAN
enhances the rate of PAN decomposition. To correct for this empirically,
we include a term for the N02-enhanced decomposition in equation 37 to
give :
[PAN]k_a{k2[NO]/(k2[NO] + kjN02])} + [PAN]k* (38)
Taking the reciprocal of equation 38 and rearranging gives:
(kobl ~ "1 = (k-l)~1 + ([N02]/[NO])(k1/k2k_1) (39)
Thus, a plot of (kN°x - k^2)"1 versus [N02]/[NO] should be linear with
obs obs
slope = k!/k2k_! and intercept = (k_i) 1. To test the validity of equation
39, we measured PAN decomposition rates in the presence of excess NO and
N02 at 25, 34.3, 39.0, and 43.9°C. Semilogarithmic plots of [PAN]t/[PAN]0
versus time were typically linear to at least 50% decomposition, although
upward curvature was sometimes noted for cases of low [NO]/[N02]. Linear
NO
regions of the semilogarithmic plots were used to obtain k ,x values,
and these are summarized in Table 5 along with reaction conditions and
values of k , 2 for the experiments. For decompositions at 25, 34.3, and
T\ ^
\Tf\
° 2
43.9°C, k , 2 values were taken directly from Figure 6. For data at
39.0°C, kN02 values were calculated from the relationship
»,,-. obs
(k v2)/[N02] = 0.49 ± .04 M"1 s~l. This relationship was obtained by
ODS wr\
interpolation of a plot (now shown) of log (k , 2/[N02]) at a given
temperature (i.e., the log of the slope of each line in Figure 6) versus
33
-------
reciprocal absolute temperature. For the semilogarithmic plot, slope =
4776 ± 300, intercept = 14.82 ± 1.0 s"1, and correlation coefficient =
0.996.
Figure 8 is a plot of Table 5. The figure exhibits considerable
scatter, probably because of inaccuracies in determining k , 2. Never-
O DS
theless, the slopes and intercepts of the figure can be meaningfully
used to calculate k2/ki ratios for each temperature. Least-squares data
and the calculated ratios are given in Table 6. It is seen that k2/ki
is essentially independent of temperature, which is to be expected for
reactions 1 and 2, both of which should have negligible activation
energies. The average value for the four ratios is k2/ki = 3.02 ± 0.68.
This average is used to calculate k2 from the previously determined value
of ki. Thus,
k2 = 3.02(1.0 x 10* M-1 s-1) = 3.0 x 10' M
which compares favorably with the value of 5.0 x 10* M"1 s"1 reported
by Howard et al.for HOj plus NO.1*3
CONCLUSIONS
The equilibrium -1,1 between PAN, acetylperoxy radical, and N02 has
useful practical applications as well as some interesting environmental
ramifications .
We find PAN to be a convenient thermal source of gas phase acetyl-
peroxy radicals and N02. Low concentrations of PAN are easily prepared
and safely handled, and the rate of 0-N cleavage is rapid even at moderate
temperatures (tt, for homolysis = 31 min at 25°C). By observing the
decomposition of PAN in the presence of a free radical scavenger, it is
possible to measure the rate of reaction of acetylperoxy radicals with
the scavenger relative to the rate of combination of acetylperoxy radicals
with N02 (ki = 1.0 x 10* M"1 s"1). We have used this method to determine
a value for the important reaction of acetylperoxy radicals with NO
(k2 = 3.0 x 109 M"1 s"1). A study of the competition between acetyl-
peroxy radicals, N02, and some hydrogen atom donors suggests that the
34
-------
Table 5. RATE CONSTANTS AND REACTION CONDITIONS FOR DECOMPOSITION OF PAN
IN THE PRESENCE OF ADDED NO AND N02 AT VARIOUS TEMPERATURES
Temp [PAN!
°C MX 104
25
0.730
0.746
1.03
0.645
0.948
0.799
0.572
1.33
—
34.3 0.823
1.01
0.537
1.32
0.830
0.941
0.651
39.0
0.996
0.601
1.83
0.971
43.9
0.488
0.562
0.858
0.922
0.848
[N02]/[NO]
0.83
2.56
3.03
3.54
3.59
3.59
4.08
4.85
0
1.70
2.03
3.45
4.85
5.40
6.53
8.67
0
1.62
2.25
4.75
7.93
0
2.25
2.31
3.09
3.63
7.00
fNOil
I 2J ,
M x 104
0
4.10
5.47
2.79
3.54
4.32
6.03
5.22
3.05
0
3.02
3.83
3.43
6.86
4.70
6.22
5.59
0
3.73
3.88
3.35
5.21
0
2.69
4.47
2.34
3.02
3.38
kN02a
obs
s-1 x lO"
0
0.25
0.33
0.17
0.32
0.26
0.36
0.31
0.18
0
0.51
0.56
0.58
1.17
0.80
1.06
0.95
0
1.68
1.66
1.51
2.34
0
1.48
2.46
1.29
1.66
1..86
kNOxb
obs
s-1 x 10"
3.7°
4.48
2.82
2.49
2.57
2.83
2.32
1.79
1.68
14. Oc
10.5
8.45
7.10
7.86
5.56
6.34
5.39
29.0°
13.6
16.3
13.7
9.11
52. 2a
33.5
28.5
28.1
21.1
14.3
,, NOx , N02i
(k — k )
obs obs
s x 10~3
2.70
2.36
4.01
4.31
4.45
3.90
5.11
6.77
6.68
0.714
1.00
1.28
1.53
1.49
2.10
1.89
2.25
0.345
0.840
0.683
0.820
1.48
0.192
0.310
0.384
0.373
0.514
0.804
3From linear region of semilogarithmic plot of [PAN] /fPAN] versus time
bFrom Figure 6, except for 39.0° data, in which case kQ°2 = 0.49[N02]
c
k °x for [N02]/[NO] = 0 is taken from k_j values of Table II
d
k ,x for [N02]/[NOJ = 0 is taken from k_j value extrapolated from Figure 3.
35
-------
•o
c
8000
7000 —
6000 —
5000 —
0*3
Z o
4000 —
X
3
3000 —
2000
1000 —
4 5
IN02]/[NO]
43.9°
SA-4466-8
*°
FIGURE 8 RECIPROCAL CORRECTED RATE CONSTANT (kx
) VERSUS
[N02]/[NO] FOR DECOMPOSITION OF PAN IN THE PRESENCE OF ADDED
N02 AND NO AT; 25°C (•-•), 34.3°C (O-o), 39.0°C (d-D), AND
43.9°C U-A)
36
-------
Table 6. LEAST SQUARES DATA FOR FIGURE 8 AND CALCULATED VALUES OF
k2/kx FOR DECOMPOSITION OF PAN IN THE PRESENCE OF ADDED N02
AND NO AT VARIOUS TEMPERATURES
Temp
25.0
34.3
39.0
43.9
Slope3
s x 10~3
0.861 ± 0.177
0.178 ± 0.025
0.123 ± 0.029
0.900 ± 0.034
Intercept3
3 x 10 3
1.48 ± 0.58
0.807 ± 0.119
0.425 ± 0.124
0.155 ± 0.009
r2
0.768
0.897
0.860
0.959
\. ^- 1 j
s x ID"3
2.70b
0.714b
0.345b
0.192°
k,/kld
3.14
4.01
2.80
2.13
o
Error limits are standard deviations
k_! taken from Table 2
Q
k_} extrapolated from Figure 3.
k2/kj = (slope x k_j) , see eq 39.
technique can be applied to the reaction of acetylperoxy radicals with
a wide variety of organic substrates.
Environmentally, PAN has until now been considered a product of
photochemical smog only. The equilibrium -1,1, however, suggests that
PAN may also have a significant effect in causing the formation of air
pollution. Under ambient conditions of high N02/N0 (such as in late
afternoon or at night), PAN accumulates and persists in the environment.
However, when the ambient level of NO increases (e.g., when exhaust
emissions increase in early morning) PAN decomposes rapidly, resulting
in the oxidation of NO to N02. Computer simulation of the propylene/NO
X
system indicates that including the equilibrium -1,1 in the model in-
creases both the initial rate of ozone products and the maximum level
of ozone produced. Thus, PAN may play an important role in the early
morning chemistry of polluted atmospheres.
Contrary to suggestions in the literature1*5'1*6 the thermochemistry
of the mechanism of homolysis of peroxy nitrates in general indicates
37
-------
that 0-N cleavage should predominate over 0-0 homolysis for ROON02, where
R is H, alkyl, acyl, and aroyl.
Finally, we compare our values of k2/ki = 3.02 ± 0.68 and kj. =
10(i6.2» ± o.eo) exp(26910 ±900/9) s"1 to values for these rate constants
that have been reported by Cox and Roffey.1*7 From a study of the rate
of oxidation of NO to N02 by PAN, these authors found k2/ki. = 1.85 ±
0.6 and k_x = io(l"-'° * °'6o) exp(24860 ± 760/0) s"1. The two expressions
of k_j. give identical values at 322K, while at 300K the calculated
value of Cox and Roffey is 27% of our calculated value. We believe our
values are to be preferred because our method involves direct measurements
of PAN, whereas Cox and Roffey actually measured changes in NO concen-
tration. The latter method requires knowing the number of NO molecules
oxidized per molecule of PAN decomposed, and there is some uncertainty
in the determination of this quantity using Cox and Roffey's method.
38
-------
4. GAS PHASE HYDROXYL RADICAL REACTIONS. PRODUCTS AND PATHWAYS
FOR THE REACTION OF OH WITH AROMATIC HYDROCARBONS
INTRODUCTION
Aromatic hydrocarbons are important constituents of polluted urban
atmospheres1*8 and the extent to which they contribute to photochemical
smog is of considerable concern.1*9"52 Their participation in atmospheric
chemistry is the result of reaction with OH radical. The kinetics of gas
phase OH-aromatic reactions have been the subject of several recent re-
ports,53"58 and it is clear that initial attack by OH is the major route
for involvement of aromatic hydrocarbons in the chemistry of the tropo-
sphere.59'60 Although the mechanisms of solution phase OH-alkyl benzene
reactions have been extensively investigated,61"68 reactions in the gas
phase, where ambiguities due to solvation or solvolysis are excluded, have
not been studied in depth. Thus products and precise mechanisms of the
gas phase reactions remain undetermined; however, identifying them is essen-
tial for determining the fate of aromatics in the environment and for
developing accurate chemical models of urban airsheds.
Given these considerations, we have undertaken to determine directly
the products of the gas phase reaction of OH with benzene, toluene, p-xylene,
and 1,3,5-trimethylbenzene.
Our results show that two major reaction pathways are available. Shown
for the case of toluene, these are benzylic hydrogen atom abstraction
(reaction 40) and radical addition to the aromatic ring (reaction 41).
OH + C6HSCH3 -*• H20 + C6H3CH2 (40)
(41)
39
-------
For each of the hydrocarbons studied, we have determined the relative
incidence of reaction 40 and reaction 41 by product analyses. Our data
are in close agreement with relative rate constants determined by kinetic
experiments.58 The intramolecular selectivity of ring addition to toluene
by OH is also reported and discussed in terms of formation of an inter-
mediate OH-toluene ir-complex. Finally, we consider the environmental
implications of our findings.
EXPERIMENTAL SECTION
Materials
The benzene and toluene used were Malinckrodt Analytical Reagent
grade. 1,4-Dimethylbenzene (99%) and 1,3,5-trimethylbenzene (99+%, Gold
Label) were from Aldrich Chemical Company. Benzene, toluene, and 1,4-
dimethylbenzene were purified by reaction with concentrated H2SO,», phase
separation, three aqueous washings, drying with MgSO<«, and fractional
distillation. 1,3,5-Trimethylbenzene was fractionally distilled before
use. Argon, H2, and 02 were supplied by Liquid Carbonics and N02 (0.54% in He)
by Linde.
Apparatus
The flow system used in the studies was constructed of 2.5 cm i.d.
Pyrex tubing. 0-ring joints and stopcocks with Viton o-rings were used
throughout. A Scintillonics model HVI5A generator was used to produce
the microwave discharge. A power level of 50 w was typically employed.
Pressures of gaseous reactants were measured with a Validyne model DP7
Transducer. The main pump was an Alcatel direct-drive model.
Glpc analyses were performed with a Hewlett-Packard model 5700A
chromatograph equipped with a flame-ionization detector. A Finnigan Model
3200 tandem GC-MS was used for product identification. Hplc analyses were
performed with a Waters Associates system consisting of two model 6000A
pumps, a model 660 solvent programmer, and a U6K injector. A Schoeffel
Instrument Corporation model GM770 variable wave-length detector was used.
40
-------
Methods
Reaction Conditions
Aromatic hydrocarbons were added by syringe at a rate of 0.17 cm3
min maintained by a syringe pump. Hydrogen atoms were generated by pass-
ing a dilute mixture of H2 in Ar through the microwave discharge. Typical
reactant pressures (torr) were: H2, 0.2; Ar, 5; hydrocarbon, 0.1; N02,
1 to 5 x 10~3; He, 0.2 to 1; 02, 1 to 10. Total pressures thus ranged
from 6 to 15 torr. The linear flow velocity of the system was 1.0 x 103
cm s"1. All reactions were carried out at ambient temperature, 25 ± 3°C.
Product Analysis
Products were sampled by condensation in a cold trap (-78 or -196°C)
or by pulling a fraction of the gas stream through a short glpc column packed
with Chroraosorb G, AW/DMCS, 100/120 mesh (Applied Science). Control ex-
periments demonstrated that the condensation traps collected only about 10%
of the organic material passing through the system. For this reason, mass
balances could not be obtained. Other experiments showed, however, that con-
densable species were trapped nonselectively, i.e., that product ratios in
the condensed material were equivalent to ratios in the gas stream. Control
experiments to test the efficiency of the gas sampling system showed that
products were not quantitatively trapped on the Chromosorb G column, the
more volatile materials being pulled through the column to the sampling
pump. Gas sampling could therefore be used only for qualitative analysis.
Product distributions from both condensed and gas phase sampling methods
were carefully compared to rule out the possibility of interference caused
by side reactions occurring in the cold trap.69
Products were identified by GC-MS and by comparing glpc retention
times with authentic samples. Toluene and its reaction products were
analyzed using a2.5mx0.5mmID stainless steel column packed with 4%
tri-2-cresylphosphate on 100/120 mesh Chromosorb G AW/DMCS and al.8mxl.Qmm
glass column packed with 10% OV-17 on 80/100 mesh Chromosorb W. Benzene and
41
-------
its reaction products were analyzed with the same OV-17 column. The tri-
cresylphosphate column was also used for analysis of 1,4-dimethyl benzene
and 1,3,5-trimethylbenzene reactions. Products of the latter reaction were
also analyzed by hplc using a y-Bondapak Cie column with 30% CH3CH/H20 as
the mobile phase. Glpc and hplc response factors were obtained for all pro-
ducts.
RESULTS
Hydroxyl radicals were generated by the discharge flow method from
hydrogen atoms and N02,
H« + N0a —*- H0» + NO (42)
This method has been used extensively70"76 for determination of OH reaction
kinetics and is known to be reliable. In our system, an aromatic hydro-
carbon and molecular oxygen are added 10 cm downstream from the point of
addition of N02. For the linear flow velocity (1.0 x 103 cm s"1) and con-
centrations of hydrogen atoms, and N02 ([H»] % 6 x 1013 particles cm~3)77
used in our system, reaction 42 (kj2 = 4.8 x 10"11 cm3 molec"1 s"1)78 has
gone to completion at this point. Providing that the amount of [N02] added
is greater than or equal to the amount of [H»], the products observed will
have resulted entirely from reaction of OH with the aromatic hydrocarbon
in the presence of 02, NO0 and N02. When a hydrocarbon concentration
[3 x 1015 molecule cm~3] higher than the concentration of OH
[^ 10*2 particles cm"3]30 is maintained, conversion is low and secondary
OH-aromatic reactions are negligible.
The gas phase products of the reaction of OH with toluene were found
to be benzaldehyde (C6HSCHO), benzyl alcohol (C6HSCH2OH), 3-nitrotoluene
(3-N02C6H<.CH3), isomeric cresols (2-, 3-, 4-HOC6H<,CH3), and 2-methyl-l,4-benzo-
quinone (CH3C6H302). Phenol, benzoic acid, benzyl nitrate, and a-, 2-, and
4-nitrotoluene were searched for but not found. We set 1% of the total re-
action products as an upper limit for the formation of each of these products.
In some runs, small (< 1% of the total products) amounts of an unidentified
product less retentive on glpc than toluene were also observed.
42
-------
Similar results were obtained with hydrocarbons other than toluene.
Reaction of OH with benzene yielded phenol and nitrobenzene. The reaction
with 1,4-dimethylbenzene gave 4-CH3C6H<.CHO, 4-CH3C6H<,CH2OH, 2-HO-l,4-(CH3) 2C6H3,
and 2-N02-l,4-(CH3)aC6H3. Finally, the products of reaction of OH with
1,3,5-trimethylbenzene were 3,5-(CH3)2C6H3CHO, 3,5-(CH3)2C6H3CH2OH, and
2-HO-l,3-5-(CH3)3C6H2.
For each hydrocarbon, product distributions were studied quantitatively
as a function of [02], [N02], and total pressure. These experiments were
studied in greatest detail for the case of toluene. Product distributions
for this reaction at constant [02] and various [N02] are given in Table 7.
Table 8 gives distributions at constant [N02] as a function of [02] and
total pressure.
DISCUSSION
Tables 7 and 8 show that the individual product yields vary with reaction
conditions, but the ratio (C6H3CHO + C6HSCH2OH)/(total products) is independent
of N02, 02, and total pressure. A mechanism that accounts for this and for
the observed products of all the OH-aromatic reactions is given (for the case
of toluene) by reactions 40-51:
C6H3CH2» + 02 •
C6HSCH202« + NO •
2C6HSCH202»
C6HSCH20» + 02 -
C6H5CH20« + NO •
C6H5CH20» + N02 -
I + 02 •
C6H5CH202» (43)
C6HSCH20« + N02 (44)
C6H5CHO + C6HSCH2OH + 02 (45)
C6HSCHO + H02» (46)
C6H5CHO + HNO (47)
C6HSCHO + HN02 (48)
HOC6HACH3 + H02« (49)
I + N02
3-N02C6Hi.CH3 + H20 (50)
-------
H02- + NO-^HO« + N02 (51)
Reactions 43 through 48 are analogous to the known reactions of methyl,79
methylperoxy, 80,81 and methoxy82*83 radicals. Reaction 49 closely re-
sembles the reaction of cyclohexadienyl radicals with 02.8lf Reaction 5085
and similar addition-elimination reactions®6'**7 have been observed in
solution. Reaction 51 has been the subject of numerous recent investiga-
tions.88-91
According to the proposed mechanism, C6H5CHO and C6H5CH2OH result
ultimately from benzylic hydrogen atom abstraction (reaction 40), whereas
3-nitrotoluene and cresols result from ring addition (reaction 41). Thus
the ratio of abstraction products to total products should be a constant
governed by the ratio k<,o/(k<,o + ^n). We have run the HO-toluene reaction
a total of 20 times, and for all runs the average value of (C6H3CHO + C6H5CH2OH)
/(total products) was 0.15 ± 0.02. This is in excellent agreement with the
value of ktt0/(kil0 + k^i) = 0.14 *_,. obtained by Perry et al58 using a
— u • UD
flash photolysis-resonance fluorescence technique. Values of ki,o/(k<,0 + k/,i)
were calculated for the other hydrocarbons used in our study. As Table 9
shows, our values of k<,o/(k40 + k*i) agree in all cases with those of Perry
et al. within the limits of experimental uncertainty.
Over the pressure range of our experiments, we have found no
evidence of a pressure effect. Therefore we believe that the reversible
decomposition of the hot OH-aromatic adducts is not occurring in our
system. This is supported by earlier work of Davis et al.55 For the
case of toluene, these workers found that at 298 K and 5 to 15 torr of
He bath gas, the value of (k<,0 + k^) was about 0.8 of the high pressure
value. Since our results were obtained with a mixture of Ar and Oa as
bath gas rather than He, our value of k/,o/(k4o + k41) is expected to
be essentially at the high pressure limit. Variations in k<,0/(k<.0 +
k/.i) over the pressure range 6 to 15 torr should be well within the +
13% standard deviation observed for our experiments with toluene.
Reaction kA1 for the dimethyl- and trimethylbenzenes should be even
44
-------
Table 7. DISTRIBUTION OF INDIVIDUAL PRODUCTS AS A FUNCTION OF [NOz] ADDED FOR REACTION OF OH WITH TOLUENE PLUS 9.7 x 10" MOLEC CM-3 Oa"
Individual Product as % of Total
INUaJ
molec cm"3
x 10-'*
0.71
1.04
1.39
1.75
C6H3CHO
12.5
12.5
12.4
8.16
C.HjCHjOH
4.0
4.8
4.4
9.52
3-NOaC,
36.
36.
45.
53.
H.CHs
6
9
3
2
2-HOC
37
35
27
19
aH j|CH 3 4— HOC an*, Cn 3
.6
.3
.0
.61
5.
6.
5.
4.
69
03
5
72
3-HOC«H<.CH3
3.
4.
4.
4.
6
04
5
1
Total
• HOCsHiCHj
46.9
45.4
37.0
28.43
CHSC
_
0.
0.
0.
c
_
46
9
69
«H,CHO + C6H,CHaOI
Tofal
16
17
16
17
.46
.3
.7
.7
iotai
- C6H
x
2
5
2
1
products
jCH3
10*
.2
.7
.0
.3
Total pressure - 8.18 torr.
Ul
Table 8. DISTRIBUTION OF INDIVIDUAL PRODUCTS AS A FUNCTION OF [Oa] ADDED AND CF TOTAL PRESSURE FOR REACTION OF OH WITH TOLUENE USING 1.39 x 101* MOLEC CM"3
ADDED NOa
Total Produce
molec cm"3
x 10-16
3.2
6.5
9.7
13.0
14.6
16.2
19.2
Pressure,
torr
6.8
7.8
8.8
9.8
10.3
10.8
11.8
Individual Product as % of Total
C.HjCHO
6.6
7.4
8.5
10.1
11.0
8.5
12.4
C.H,CHaOH
6.2
5.0
3.7
3.3
2.6
3.8
4.4
3-NOaC.H4CH3
44.8
42.0
44.3
49.9
47.6
44.3
45.3
2-HOC,H4CH3
35.2
34.6
30.7
25.8
27.0
29.2
27.0
3-HOC»H»CH,
3.8
6.3
5.5
5.3
5.8
6.5
5.5
4-HOC6H»CHs
3.5
4.5
4.7
5.3
5.9
6.5
4.5
HOCgHfcCH.
42.5
45.4
40.9
36.4
38.7
42.0
37.0
, 2-CH3C,H,Oa
< 0.5
< 0.5
2.5
< 0.5
< 0.5
1.1
0.9
C.HjCHO + C.HjCHjOH
(Total Products)
12.
12.
12.
13.
13.
12.
16.
8
4
2
4
6
3
7
C.H3CH3
x 10*
9.4
6.4
15.7
2.5
2.2
1.8
2.0
-------
Table 9. PRODUCTS AND RATE CONSTANT RATIOS [kjtk,. + ka)] FOR REACTIONS OF OH WITH VARIOUS HYDROCARBONS
IN THE PRESENCE OF N02 AND 02a
Hydrocarbon Products This Work Reference 58
benzene C6HSOH, C6H3N02 < 0.05b 0.05
(0.01 to 0.13)°
toluene CsH3CHO,d C6HsCH2OH,d CH3C6H302e 0.15 ± 0.02 0.16
3-NOaC6H<,CH3, 2,3, and 4-HOC«HACH, (0.11 - 0.23)
1,4-dimethylbenzene 4-CH9C«H^CHO,d 4-CH3C6H<,CH2OHd 0.15 ± 0.02 0.07
2-HO-l,4-(CH3)2C6H3 (0.04 to 0.14)
2-N02-l,4-(CH3)aC6H3
1,3,5-trlmethyl- 3,5-(CH3)2C6H3CHOd 0.021 ± 0.006 0.02
benzene
3,5-(CH3)2C«H3CH2OH 0.01 to 0.06
2-HO-l,3,5-(CH3)3C6H2
Reaction conditions employed were similar in all cases to those reported in Tables 7 and 8 for toluene.
Represents fraction of total reaction proceeding by ring hydrogen atom abstraction.
values in parentheses represent the reported range.
Products derived from benzylic hydrogen atom abstraction.
Tlinor product.
-------
less susceptible to reversibility than it is for toluene, and k/,! occurs
to the exclusion of other reactions for the case of benzene. We there-
fore conclude that the reverse of reaction 41 is unimportant at the
temperature and pressures used in our experiments.
At this point we can calculate k/, i (o) /kz,!, k<, i (m) /ku i, and ki,].(p)/k<,i
(for toluene) i.e., the portions of reaction 41 that proceed by OH
attack at the 2-, 3-, and 4-positions, respectively. To derive these
relative rate constants, we use the measured product ratios for 3-
nitrotoluene and the isomeric cresols and the following relations:
-i fs?1
k4i(m)/ki,i = (3-HOC6HACH3)(total cresols + 3-N02C6HACH3) v '
k4x(o)/k41 = {[2-HOC6Hi,CH3/(2-HOC6Hi,CH3 + 4-HOC6H<,CH3) ] (3-N02C6H<,CH3)
+ 2-HOC6H<.CH3}(total cresols + S-NOaCgH^CHa)"1 (53)
k<.i(p)/k^1 = {[4-HOC6Hi.CH3/(2-HOC6Hi,CH3 + 4-HOC6H,,CH3) ] (3-N02C6H,,CH3)
+ 4-HOC6H,.CH3} (total cresols + 3-N02C6H4CH3)"1 (54)
To derive equation 52, we recall that 2- and 4-N02C6H<,CH3 are not found as
products and that 3-HOC6H<.CH3 is the only observed product derived from OH
attack at the 3-position. For equations 52 and 54, we note that 3-N02C6Hi.CH3
results from attack by OH at either the 2- or 4-position and assume that kn
is independent of the position of OH in the intermediate, _!_. In this case
the total amount of product derived from OH attack ortho to the methyl group
is equal to the observed amount of 2-HOC6H<,CH3 plus the fraction of the
3-N02C6H<,CH3 that was derived from .1 with OH in the 2-position, this fraction
being equated to the product ratio
2-HOC6Ht,CH3/(2-HOC6H<,CH3 + 4-HOC6H,,CH3) .
For toluene, the average values of ki,i(o)ki,i, k^1(m)/k
-------
toluene in solution and for addition to toluene by other highly reactive
species. From Table 10 it is evident that, in reactions with toluene, gas
phase OH is at once a highly reactive and highly selective species. A
similar parallel between reactivity and selectivity has been noted in some
ionic electrophilic aromatic substitution reactions, and Olah91* has suggested
that a ir-complex is involved in these reactions as a precursor to the formation
of a a-adduct:
ArH + X i X.. .ArH ->Ar^H (55)
XX
iT-complex a-adduct
The intermediacy of n-complexes in other radical reactions has been suggested,
!
and, indeed, the formation of an OH-aromatic ir-complex was considered by Perry
et al as a possible explanation for the reversibility of reaction 41 under
the conditions of their experiments. Although it is not conclusive, our
evidence concerning the selectivity of the addition of OH to toluene lends
credence to the postulated participation of ir-complexes in OH-aromatic re-
actions. Regardless of the precise mechanism of the reaction, it is apparent
that ortho-attack by OH is highly preferred for the addition to toluene. This
behavior confirms64'96 the electrophilic nature of the hydroxyl radical. That
ortho attack by OH occurs more readily in the gas phase than in solution is
attributable to steric effects that are significant for solvated OH but not
for the free radical in the gas phase.
Next we wish to estimate the rate constant ratio k50/ki,9, i.e., the re-
lative reactivities of N02 and 02 toward the adduct _!. To do this, we com-
bine and integrate the kinetic expressions for reactions 49 and 50, giving:
v /v - (30-N02C6H,CH3) . [Oa] ,,,,
iC50/K"9 (total HOC6H<.CH3) [N0a] °°'
With our experimental method, the values for (3-N02C6H<,CH3)/(total HOC6H*CH3)
and [02] are readily determined, but the value for [N02] is not. The difficulty
with [N02] is that a fraction of the added N02 consumed in titrating H» through
reaction 42 and that there is no direct measure of [H»] in our system. To
48
-------
Table 10. REACTIVITY AND POSITIONAL SELECTIVITY OF RING-ADDITION TO TOLUENE BY VARIOUS SPECIES
add
100 x Fraction of Ring-Addition References
Species
OH
OH
OH
0(3P)
Phase
gas
soln(PhCH3)
soln(H20)
gas
cm3 molecule"1 sec"1
5.4 x 10-12,
—
4.8 x 10"ia
2.3 x 10-13
2-Position
80.6
59
55
60
3-Position
5.1
6
15
15
4-Position
14.3
35
29.5
18
kadd
19
—
1
93
Products
this work
8
6
92
vo
-------
estimate the [N02] in our system, we have developed a detailed kinetic model
for the OH-toluene reaction and have used [H»] as a variable to obtain a best
fit of the model to the experimental data in a given run. Once an optimum
value for [H»] is chosen, the [N02] present in the experiment can be obtained
by subtracting [H»] from the initial [N0a]. Applying this method to the data
of Table 7, we calculate k50/k<,9 = (4 ± 2) x 103, where the uncertainty is
caused primarily by inaccuracies in estimating [N02].
On the basis of the product distributions and relative rate constants
discussed above, we can assess the implications of the OH-aromatic reactions
with respect to the environment. As was stated initially, the two major path-
ways for the reaction are hydrogen atom abstraction, reaction 40 and ring-
addition, reaction 41. The fate of the benzylic radical formed in reaction
40 will be governed by reactions 43, 44, and 46, and aromatic aldehyde will
be the sole product. This judgment is based on the rate constants for the
analogous reactions of methyl,79 methylperoxy,80»81 and methoxy82»83 radicals,
and the typical value of [02]/[N02] £ 106 under ambient conditions. The
ambient [02]/[N02] ratio also plays an important role in establishing the
ultimate products from reaction 41 because the adduct _I is partitioned between
reaction 49, with 02, to give phenolic products and reaction 50, with N02, to
give nitroaromatics. Using equation 55 and our values of k<,0/(k<,o + k<,i) = 0.15
and kso/k<.9 = 4 x 103 for toluene, we have calculated the yield of 3-N02C6H<,CH3
expected for various concentrations of N02.
3-N02C6H,,CH3/(total products) = {1 - [k^o/Uo + kfcl) ] }(kt !/k10) ([N02]/[02] (57)
These values are given in Table 11. The table shows that 3-N02C6Hi.CH3 can be
a major product for artificially high concentrations of N02, but that under
typical ambient concentrations of N02 (i.e., < 1 ppm), the yield of nitrotoluene
will be less than 1 percent of the total products. Thus we predict that in the
environment, reaction 41 will lead exclusively to phenolic products. Table 12
summarizes the atmospheric product distributions expected from the reaction of
OH with various aromatic hydrocarbons. In all cases, aromatic aldehydes and
50
-------
phenols are the predicted products. Because these species have not been re-
ported as constitutents of polluted urban atmospheres, it seems likely that
they are rapidly removed by as yet undetermined secondary reactions.
A final point concerns recent smog chamber experiments^7»^8 with toluene,
in which 2-, 3-, and 4-nitrotoluene and isomeric nitrocresols are found as
products in addition to the expected benzaldehyde and cresols. The appearance
of nitrotoluenes is readily accounted for by the elevated (£ 1 ppm and greater)
concentrations of N02 and high conversions used in the experiments. That the
2- and 4-isomers of nitrotoluene are observed is perplexing, because we have
shows that the OH attack on the meta- to the methyl group (the reaction that
would lead to the 2- and 4-nitrotoluenes) is of minor importance. Because
ionic nitrations typically give predominately 2- and 4-substitution, and be-
cause we find69 NO to be a powerful nitrating agent in condensing our reaction
X
mixtures, we suggest that the observance of 2- and 4-nitrotoluenes in smog
chamber studies is indicative of heterogeneous reactions that occur in the
aerosol phase or during sampling of the products. Phenolic compounds are
especially susceptible to heterogeneous nitration, and the origin of nitro-
phenols in smog chamber experiments must be interpreted with extreme caution.
51
-------
Table 11. CALCULATED3 YIELD OF 3-NITROTOLUENE (3-N02C6IUCH3/TOTAL
PRODUCTS) FROM OH-TOLUENE REACTION AG A FUNCTION OF N02
CONCENTRATION*1
[N02] . 3-NOaC6H^CH3
10- 13 cm3 mole"1
1.10
3.0
10.0
30.0
100.0
300.0
PPM
0.04
0.12
0.40
1.2
4.0
12.0
(total products)
5.4 x W~k
1.63 x ID'3
5.4 x ID'3
1.6 x ID'2
5.4 x 10~2
1.6 x 10"1
Calculated from equation 57.
b[02] = 0.21 atm.
Table 12. ATMOSPHERIC PRODUCTS FOR THE REACTIONS OF AROMATIC
HYDROCARBONS WITH OH
Q
Hydrocarbon Atmospheric Products
C6H6 C6H5OH, 100%
CfiHsCH3 CsHsCHO, 15/4
2-,3-,4-HOC6H<,CH3, 85%
1,4-(CH3)2C6H<, 4-CH3C6H,,CHO, 15%
2,5-(CH3)2C6H3OH, 85%
1,3,5-(CH3)3C6H3 3,5-(CH3)2C6H3CHO, 2%
2,4,6-(CH3)3C6H2OH, 98%
rt _
Assuming [N02] < 1 ppm, 02 = 2.5 x IQ* ppm.
52
-------
5. GAS PHASE HYDROXYL RADICAL REACTIONS. PRODUCTS AND PATH-
WAYS FOR THE REACTION OF OH WITH BENZALDEHYDE
INTRODUCTION
Single-ring aromatic hydrocarbons comprise a high proportion of the
carbon found in polluted urban atmospheres1*8,99 and are known50"53 to produce
a variety of adverse effects (eye irritation, ozone, oxidant, and peroxy-
nitrate formation, and so forth) associated with photochemical smog. It
is therefore important that the atmospheric chemistry of aromatic hydro-
carbons be understood in detail. Studies of the kinetics of the gas phase
reactions of aromatic hydrocarbons with species such as hydroxyl radical,53"58
oxygen atom,100 ozone101 and peroxy radicals102 leave little doubt that re-
action with OH is by far the most important route for involvement of aromatics
in the chemistry of the troposphere. We previously °3 discussed the mechanism
of the toluene reaction and demonstrated that the major pathways for the re-
action are hydrogen atom abstraction, reaction 58, and addition to the aro-
matic ring, reaction 59.
C6H5CH3 + OH —*- C6HSCH2- + H20 (58)
C6H5CH3 + OH -*- OH (59)
In the ambient, reactions 58 and 59 would ultimately lead to benzaldehyde
and isomeric cresols, respectively. Because neither benzaldehyde nor
cresols are observed to accumulate in the troposphere, it seems likely
that these compounds are themselves rapidly transformed. A recent deter-
mination101* of the rate constant for reaction of OH with benzaldehyde
(kphCHQ = 1.3 x 10~n cm3 molec"1 s"1) suggests that this reaction plays
an important role in determining the fate of benzaldehyde in the atmosphere.
53
-------
Accordingly, we have undertaken to elucidate the products and mechanisms of
the OH-PhCHO reaction. By analogy with the OH-toluene reaction, abstraction
of aldehydic hydrogen (reaction 60), and ring-addition by OH (reaction 61),
must both be considered.
PhCHO + OH -*- PhC(0)« + H20 (60)
CHO
PhCHO'+ OH -+- I (61)
In fact, however, we find that reaction 60 proceeds to the exclusion of re-
action 61. In the following, we detail our experimental observations and
discuss their environmental ramifications.
EXPERIMENTAL SECTION
Hydroxyl radicals were generated in a discharge-flow system by re-
action 62
H + N02 -*- HO + NO (62)
The method used has been described in detail elsewhere70"76»103. All re-
actions were carried out at ambient temperature (25 ± 3°C). Typically,
2 x 10~2 torr H2 in 5 torr Ar was passed through a microwave discharge
and the hydrogen atoms so produced were reacted with 2 to 5 x 10~3 torr
N02. A mixture of 5 to 10 x 10~2 torr PhCHO in 1 to 5 torr of an 02 plus
Ar mixture was admitted 5 cm downstream from the point of addition of N02.
With a linear velocity of 1 x 103 cm-sec"1, reaction 62 (k6a = 4.8 x 10"11
cm3-molec~1 s"1)78 has gone to completion at this point. Because the amount
of PhCHO added was large compared to the hydroxyl radical concentrations
attained ([OH] £ 5 x 10"1* torr77), conversions were low and secondary re-
actions were minimized. Gas phase products were trapped on a solid ad-
sorbent (Tenax-GC)105»106 and then analyzed by glpc. In addition to gas
phase products, a significant amount of solid residue accumulated on the
54
-------
walls of the flow system. This residue was physically removed from the
walls, weighed, and subjected to elemental analysis. Then it was analyzed
by field ionization mass spectral (FIMS) which gives exclusively
the parent peaks in proportion to the composition of the components.
RESULTS
In all reactions studied, the only observable gas phase product was
phenol (PhOH). Benzoic acid, perbenzoic acid, peroxybenzoyl nitrate, and
2- and 3-hydroxybenzaldehyde were looked for but not found. The yield of
PhOH (expressed as a percent of PhCHO) was determined under conditions of
varying N02, 02, and total pressure. These results are summarized in Table 13.
In Table 14, some characteristics of the wall residue and the gas phase
product are presented. Finally, mass spectral data for the wall residue are
present in Table 15.
In Table 13, we see that the yield of PhOH is on the order of 1 to 2%
(consistent with the low conversions expected) for a variety of [N02] and
02 pressures (runs 1-7). However, for runs 8-11, in which Oa was eliminated
from the reaction system, the yield of PhOH decreases by an order of magni-
tude. These data indicate the PhOH is formed by reactions 60 and 63-68.
PhC(0)« + 02 »~PhC(0)02» (63)
PhC(0)02»+ NO »-PhC(0)0» + N02 (64)
PhC(0)0« »-Ph» + C02 (65)
Ph. + 02 >~Ph02» (66)
Ph02« + NO —*- PhO» + N02 (67)
PhO» + wall *- PhOH + other products (68)
This mechanism satisfactorily explains the observed oxygen dependence
of the PhOH yield. An alternative mechanism involving ipso-attack by
55
-------
Table 13. PERCENT YIELD OF PHENOL (100 x PhOH/PhCHO) AS A
FUNCTION OF ADDED N02, 02 AND TOTAL PRESSURE
Run
1
2
3
4
5
6
7
8
9
10
11
NO 2
torr x 103
1.08
2.70
4.05
5.40
2.70
2.70
2.70
1.08
2.70
4.05
5.40
oa
torr
4
4
4
21
4
3
1
0
0
0
0
Total Pressure Ar + 02
torr
10
10
10
10
10
10
10
6
6
6
6
100 x PhOH
PhCHO
2.11
2.21
2.19
1.22
1.89
1.85
1.48
0.091
0.095
0.084
0.019
56
-------
en
-j
Table 14. WALL AND GAS PHASE PRODUCTS OF THE OH-PhCHO REACTION AS A FUNCTION
OF ADDED N02-
Run N02 100 x PhOH Total PhOH^- Wall Residue Wall Analysis, %
PhCHO
torr x 10 3 gm gm C H N 0
12 1.35 2.15 0.31 0.16 49.2 4.74 0.42 45.6
13 3.24 2.0 0.35 0.079 49.4 4.62 1.49 44.5
3.
— 02 = 4 torr, total pressure 10 torr.
- Total PhOH = (PhOH/PhCHO)•(total moles PhCHO added)'(94 gm/mole).
-------
Table 15. FIELD IONIZATION MASS SPECTRAL ANALYSIS OF WALL RESIDUE FOR
OH-PhCHO REACTION^
m/e
Rel Ht
% of Total
Formula
Possible Structure
94
110
122
124
126
138
139
140
152
154
155
168
170
186
202
7.5
70.4
4.4
100
14.5
10.1
24.5
10.1
5.0
12.6
8.2
7.5
4.4
16.0
6.3
2.5
23.0
' lt5
33.0
4.8
3.4
8.1
3.4
1.7
4.2
2.7
2.5
1.5
5.3
2.1
C6H60
C6H602
C7H602
C6H603
C7H603
C6H5N03
C6HA0A
C7HA0A
C7H60A
C6HSNOA
C6HAN20A
Ci2H100
C12H1002
Ci 2Hjo03
PhOH
Ph(OH)2
PhC(0)OH or HOPhCHO
/=
-------
hydroxyl radical (reaction 61;)
H(0)C
PhCHO + OH
PhOH + HC(0)»
(61;)
is rejected on the grounds that the yield of phenol in this case would be
independent of the concentration of added oxygen.
The absence of gas phase products other than PhOH can be explained
by suggesting that such products might be formed only by reactions that
would be conceivable under other conditions, but that are prohibitively
slow under the low total pressures and reactant concentrations used in
our system..
For example, reactions 69-72 are unlikely in our system because of
the low total pressures employed.
+M
PhC(0)00» + N02
PhC(0)0- 4- N02
-»-PhC(0)OON02
+M
-»-PhC(0)ON02
(69)
(70)
PhO»
N0
PhO« + 0
+M
+M
H
NO 2
H
II
-HOPhNOs
Products
(71)
(72)
Similarly, the hydrogen atom abstraction reactions 73 - 75 are slow
as a result of low radical and hydrogen atom donor concentrations.
PhC(0)00» + RH
PhC(0)0» + RH
PhO« + RH
-PhC(0)OOH + R
-PhC(0)OH + R
R
60
59
(73)
(74)
(75)
-------
As an example, for RH = PhCHO at 5 x 10~2 torr (1.6 x 1015 molec cm~3),
the half-life for disappearance of PhC(0)00« by reaction 73 is approximately
20s (k73 = 2 x 10~17 cm3 molec"1 s~1).107At the same time, for NO = 5 x 10~3
torr, the half-life for PhC(0)00« disappearance by reaction 64 is 10~3 s,
assuming k6<, = 3 x 10~12 cm3 molec"1 s"1 (i.e., assuming k6<, equals the
rate constant for reaction of CH3C(0)00« plus NO108).
DISCUSSION
Because no facile gas phase reactions are available to PhO», it is
primarily consumed by heterogeneous processes at the reactor walls. The
extent of wall reactions in the OH-PhCHO system is demonstrated by the
data of Table 14. In run 12, for example, the total yield of PhOH was
0.31 gm, and a total of 0.16 gm of wall residue was recovered. Interest-
ingly, the majority of the wall products are C6 or Ci2 species, i.e.,
species derived ultimately from reactions 60 and 63-68, the only possible
exceptions being compounds with m/e = 122, 138, 152, and 154. Because
these species constitute only 10% of the wall product, which in turn amounts
to only about one-third of the total gas phase plus wall product, we set
3% as the upper limit for addition of OH to PhCHO through reaction 61.
The preponderance of abstraction over addition in the OH-PhCHO re-
action is a dramatic reversal of the trend observed for reaction of OH
with methyl-substituted benzenes. In these reactions,103 ring-addition
was the preferred process, constituting 85 to 98% of the total reaction
for the substrates investigated. This difference is rationalized on the
basis of the electrophilic nature of the hydroxyl radical. Both studies
reported in the previous section and kinetic studies58 show that ring-
addition by OH is enhanced by electron-donating substituents. We expect,
therefore, that the formyl group (an electron withdrawing substituents)
should slow the rate of reaction 59 relative to the rate of addition of
OH to benzene. Even using the OH-benzene rate constant, k^, =1.2 x 10~12
cm3 molec"1 s"1,58 as an upper limit for the rate of reaction 61, the over-
all rate constant for OH plus benzaldehyde, kphrun = 1.3 x 10"^ cm3 molec"1
is sufficiently fast that ring-addition by OH to PhCHO can be only
60
-------
a minor fraction of the total reaction pathway.
Since it has been demonstrated that the OH-PhCHO reaction proceeds
virtually entirely through reaction 60 as the initial step, it remains to
identify the products of the reaction under atmospheric conditions. Formation
of peroxybenzoyl nitrate by reaction 69 is likely under conditions where
N02/N0 is high. As with peroxyacetyl nitrate,108 however, the peroxybenzoyl
nitrate should be in equilibrium with PhC(0)00« and N02, i.e., reaction 69
should be reversible. Because of this, peroxybenzoyl nitrate acts as a re-
servoir of benzoylperoxy radicals, and the latter will be consumed in re-
action 64 when the ambient concentration of NO is high. Benzoyloxy radicals
formed in reaction 64 are expected to yield phenoxy radical through re-
actions 65-67. As a consequence, the fate of PhCHO in the environment will
be largely governed by the fate of the phenoxy radical. Although we cannot
specify the fate of PhO« with certainty, it seems likely that it would react
at atmospheric pressure primarily through reactions 71 and 72, which lead
to ring cleavage. Reaction. 71 gives isomeric nitrophenols as the stable
products. Emphasizing the importance of reaction 71 is the recent observa-
tion of nitrophenols as products from PhC(0)» plus 02 plus NO in the chlorine
X
atom initiated oxidation of PhCHO.109 As to reaction 72, it is likely to be
reversible even at atmospheric pressure, owing to the weak C-02 bond strength
in the intermediate II. We estimate DH°(C--02) = 10 kcal/mole. To the ex-
^^
tent to which reaction 72 is not reversible, the adduct II will react with
NO as in reaction 76.
II + NO *- rO' + N02 (76)
^a^
III
Suggestions for subsequent reactions of III are speculative, but
oxidative ring-opening to yield low-molecular-weight carbonyl compounds
is a possibility.
61
-------
An important distinction to be made is that reaction 71 terminates
radical chains and removes N02 from the atmosphere, whereas reactions 72
and 76 propagate radical chains and affect the important oxidation of
NO to N02.
The partitioning of phenoxy radical between reactions 71 and 72
therefore plays a potentially important role in the formation of photo-
chemical smog. For this reason, further studies on the atmospheric re-
actions of PhO» would be of interest.
62
-------
REFERENCES AND NOTES
1. J. N. Pitts and B. J. Finlayson, Angew. Chem. (Int. Ed.), 14, 1
(1975).
2. J. A. Kerr, J. G. Calvert, and K. L. Demerjian, Chem. Brit., J3, 252
(1970).
3. E. R. Stephens, in "Advances in Environmental Sciences," Vol. I,
J. N. Pitts and R. L. Metalf, Eds. (Wiley-Interscience, New York,
1969) p. 119.
4. E. F. Darley, K. A. Kettner, and E. A. Stephens, Anal. Chem., J35_
589 (1963).
5. H. Mayrsohn and C. Brooks, presented at Western Regional Meeting
of the American Chemical Society, (1965).
6. 0. C. Taylor, J. Air Poll. Cont. Assoc., 19 (5), 347 (1969).
7. D. T. Tingey and A. C. Hill, Utah Acad. Proc.. 44 (1), 387 (1967).
8. "Air Quality Criteria for Photochemical Oxidants" Nat. Air Poll.
Cont. Assoc. Admin. Pubic. Ap-63, pp 3-10 (1970).
9. W. A. Lonnemann, J. J. Bufalini, and R. L. Seila, Environ. Sci. Tech.,
10 (4), 374 (1976).
10. S. A. Penkett, F. J. Sandalls, and J. E. Lovelock, Atmos. Environ., _9
139 (1975).
11. H. Nieboer and J. vanHam, TNO Nieuws, 5, 170 (1972).
12. H. Nieboer and J. vanHam, Atmos. Environ., 10 (7), 115 (1976).
13. E. R. Stephens, E. F. Darley, 0. C. Taylor, and W. E. Scott, Int. J.
Air Water Poll., 4_ (1,2), 79 (1961); Proc. Am. Petrol. Inst. Sec.
Ill, _40 325 (1960).
14. E. F. Darley, C. W. Nichols, and J. T. Middleton, Calif. Dept. Agr.
Bull, 55 (11) 1966.
15. E. F. Darley, W. M. Dugger, J. B. Mudd, L. Ordin, 0. C. Taylor, and
E. R. Stephens, Arch. Environ. Health, J3 761 (1963).
63
-------
16. D. G. Hendry and R. A. Kenley, J. Am. Chem. Soc., 99, 3198 (1977).
17. D. Swern, in "Organic Peroxides," Vol. I, D. Swern, Ed. (Wiley-'Inter-
science, New York, 1970, p. 480).
18. R. Louw, G. J. Sluis, and H.P.W. Vermoren, J. Am. Chem. Soc., 97.
(15), 4396 (1975).
19. E. R. Stephens, F. R. Burleson, and E. A. Cardiff, J. Air Poll. Cont.
Assoc., 15 (3), 87 (1965).
20. P. Gray and A. D. Yoffe, Chem. Rev., _55_, 1069 (1955).
21. E. R. Stephens, F. R. Burleson, and K. M. Holtzclaw, J. Air Poll.
Cont. Assoc.. 12 (4) 261 (1969).
22. W. J. Braun, L. Rajenbach, and F. R. Eirich, J. Phys. Chem., 66,
1959 (1962).
23. K. U. Ingold, in "Free Radicals," Vol. I, J. K. Kochi, Ed. (Wiley-
Interscience, New York, 1973 p. 1.)
24. S. W. Benson, "Thermochemical Kinetics — Methods for the Estimation
of Thermochemical Data and Rate Parameters" (J. Wiley and Sons,
New York, 1968).
25. L. J. Bellamy, "The Infrared Spectra of Complex Molecules," (Wiley-
New York, 1958).
26. K. Nakamoto, "Infrared Spectra of Inorganic and Coordination Com-
pounds," (Wiley-Interscience, New York, 1963, p. 94).
27. E. R. Allen, and K. W. Bagley, Berg. Bunsengesell., 72, 227 (1968).
28. A. E. Pedler and F. H. Pollard, Trans. Far. Soc.. 53, 44 (195 ).
29. L. Phillips and R. Shaw, 10th Snip. (Int.) on Combust., 453 (1961).
30. J. A. Howard, in "Free Radicals," Vol. II; J. K. Kochi, Ed. (Wiley-
Interscience, New York, 1977, p. 1).
31. J. A. Howard in "Advances in Free Radical Chemistry," Vol. IV,
G. A. Williams, Ed. (Academic Press, New York, 1972).
32. F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry,
2nd Edition (Wiley, New York, 1966) p. 355.
33. D. Gray, E. Lissi, and J. Heicklen, J. Phys. Chem., J76, 1919 (1972).
34. E. L. Varetti and G. C. Pimentel, Spectrochim. Acta, 30A, 1069
(1974).
64
-------
35. E. T. Arakawa and A. H. Nielsen, J. Mol. Spectry, 2, 413 (1958).
36. C. T. Pate, R. Atkinson, and J. N. Pitts, Environ. Sci. Eng., All,
19 (1976). These aurthors report k_l = 2.8 (± 0.8) x 10~H s * at
23 ± 1 C using PAN and NO in much lower concentrations
(PAN £ 10-7 M, NO ^ 10-fy) .
37. Stephens and co-workers (see references 3 and 13) originally pro-
posed a six-center cyclic elmination of CH30N02 and CC>2 as the
mechanism for the decomposition of PAN. As with reaction 6, the
rate for the cyclic elimination can be no faster than the decomposi-
tion of pure PAN in the absence of added reactants.
38. E. S. Domalski, Environ. Sci. and Technology, 5, 443 (1971).
39. S. W. Benson, F. R. Cruickshank, D. M. Golden, G. R. Haugen,
H. E. O'Neal, A. S. Rogers, R. Shaw, and R. Walsh, Chem. Rev., 69,
279 (1969).
40. S. W. Benson and R. Shaw, in "Organic Peroxides," D. Swern, Ed.,
(Wiley-Interscience 1970, p. 105).
41. "JANAF Thtermochemical Tables," 2nd Edition, U.S. Department of
Commerce, National Bureau of Standards, NSRDS-NBS 37, Washington,
D.C., (1971).
42. S. W. Benson and H. E. O'Neal, "Kinetic Data on Gas Phase Unimolecular
Reactions," U.S. Department of Commerce, National Bureau of Standards,
NSRDS-NBS 21, Washington, D.C. (1970).
43. C. J. Howard and K. M. Evenson, Trans. Amer. Geophys. Union, 58,
464 (1977).
44. D. G. Hendry, unpublished results.
45. E.F.J. Duynstee, J.L.J.D. Hennekens, J.G.H.M. Hausmans, W. Van Raayen,
and W. Voskuil, Rec. Travl. Chem. Pays-Bas, 92, 1272 (1973).
46. E. M. Kuramshin and V. D. Komissarov, Izv. Akad, Navk. SSSR, Ser. Khim,
1937 (1976).
47. R. A. Cox and M. J. Roffey, Enviro. Sci. Tech., JL1, 900 (1977).
48. W. A. Lonneman, S. L. Kopczinski, P. E. Darley, and F. D. Sutter-
field, Environ. Sci. Tech.. j8, 229 (1974).
49. J. M. Heuss and W. A. Glasson, Environ. Sci. Technol., 2^, 1109 (1968).
50. A. P. Altshuller, S. L. Kopczinski, D. Wilson, W. Lonneman, and
F. D. Sutterfield, J. Air Poll. Contr. Assoc.. 19, 291 (1969).
65
-------
51. A. P. Altschuller, S. L. Kopczinski, W. A. Lonneman, F. D. Sutter-
field, and D. L. Wilson, Environ. Sci. Technol.. _4, 44 (1970).
52. J. M. Heuss, G. J. Nebel, and B. A. D'Alleva, Environ. Sci. Technol.,
8, 641 (1974).
53. F. D. Morris and H. Niki, J. Phys. Chem., _75, 3641 (1971).
54. D. A. Hansen, R. Atkinson, and J. N. Pitts, J. Phys. Chem., 79,
1763 (1975).
55. D. D. Davis, W. Bollinger, S. Fischer, J. Phys. Chem.. ?9, 293 (1975).
56. G. J. Doyle, A. C. Lloyd, K. R. Darnall, A. M. Winer, and J. N. Pitts,
Environ. Sci. Technol.. 9_, 237 (1975).
57. A. C. Lloyd, K. R. Darnall, A. M. Winer, and J. N. Pitts, J. Phys.
Chem.. 80, 789 (1976).
58. R. A. Perry, R. Atkinson, and J. N. Pitts, J. Phys. Chem., 81, 296
(1977).
59. K. R. Darnall, A. C. Lloyd, A. M. Winer, and J. N. Pitts, Environ.
Sci. Technol., 10, 692 (1976).
60. K. R. Darnall, G. J. Doyle, A. M. Winer, and J. N. Pitts, Environ.
Sci. Technol., 12, 100 (1978).
61. L. M. Dorfman, I. A. Traub, and D. A. Barter, J. Chem. Phys.. 41,
2954 (1964).
62. K. Sehested, H. Corfitzen, H. C. Christensen, and E. J. Hart, J. Phys.
Chem., 2i» 310 (1975).
63. H. C. Christensen, K. Sehested, and E. J. Hart, J. Phys. Chem., 77,
983 (1973).
64. M. Anbar, D. Meyerstein, and P. Neta, J. Phys. Chem., 70, 2660 (1966).
65. J.R.L. Smith and R.O.C. Norman, J. Chem. Soc., 1963, 2897 (1963).
66. C.R.E. Jefcoate, J.R.L. Smith, and R.O.C. Norman, J. Chem. Soc. (B),
1969, 1013 (1969).
67. R. Tomat and A. Rigo, J. Appl. Electrochem., j), 257 (1976).
68. Y. Ogata and K. Tomizawa, J. Org. Chem., 43, 261 (1978).
66
-------
69. In some instances products indicative of ionic (as opposed to free
radical) reactions were found in the cold traps. In general, these
side reactions could be eliminated by sampling at -78°C rather than
-196°C and by working at low pressures ( < 5 x 10~3 torr) of added
N02.
70. F. Kaufman and F. P. Del Greco, J. Chem. Phys.. 15, 1895 (1961).
71. F. P. Del Greco and F. Kaufman, Disc. Far. Soc.. 33, 128 (1962).
72. L. F. Phillips and H. I. Schiff, J. Chem. Phys.. _37, 1233 (1962).
73. J. E. Breen and G. P. Glass, Int. J. Chem. Kin., J3, 145 (1970).
74. J. E. Breen and G. P. Glass, J. Chem. Phys., 52, 1082 (1972).
75. A. Pastrana and R. W. Carr, Int. J. Chem. Kin., ^, 587 (1974).
76. W. E. Wilson, J. Phys. Chem. Ref. Data. _!, 535 (1972).
77. Concentrations estimated from a detailed chemical model of the re-
action system; see text.
78. CIAP Monograph 1, DOT-TST-75-51, page 5-232 (1975).
79. D. A. Parkes, Int. J. Chem. Kin., 9^ 451 (1977).
80. R. Simonaitis and J. Heicklen, J. Phys. Chem.. 78, 2417 (1974).
81. C. T. Pate, B. J. Finlayson, and J. N. Pitts, J. Amer. Chem. Soc..
9.6, 6554 (1974).
82. J. R. Barker, S. W. Benson, and D. M. Golden, Int. J. Chem. Kin.,
2, 31 (1977).
83. L. Batt, R. T. Milne, and R. D. McCulloch, Int. J. Chem. Kin., jj,
567 (1977).
84. D. G. Hendry and D. Schuetzle, J. Amer. Chem. Soc.. 97, 7123 (1975).
85. E. Halfpenny and P. L. Robinson, J. Chem. Soc.. 939 (1952).
86. T. Suehiro, M. Hirai, and T. Kaneko, Bull. Chem. Soc. Jap., 44,
1407 (1971).
87. M. E. Kurz, R. L. Fozdar, and S. S. Schultz, J. Org. Chem., 39,
3336 (1974).
88. R. Simonaitis and J. Heicklen, J. Phys. Chem.. 78, 653 (1974).
67
-------
89. R. Simonaitis and J. Heicklen, J. Phys. Chem.. 80, 1 (1976).
90. C. J. Howard and K. M. Evenson, Trans. Am. Geophys. Union, 58, 464
(1977).
91. R. A. Graham, A. M. Winer, and J. N. Pitts, Chem. Phys. Lett., 51.
215 (1977).
92. E. Grovenstein and A. J. Mosher, J. Am. Chem. Soc., j?2_, 3810 (1970)
93. I. Mani and M. C. Sauer, Adv. Chem. Ser.. 82^, 142 (1968).
94. G. A. Olah, Accts. Chem. Res., 4_, 240 (1971).
95. J. C. Martin, in Free Radicals, Vol. II, J. Kochi, ed., p. 516,
(Wiley-Interscience, New York, 1973).
96. R.O.C. Norman and G. K. Radda, Proc. Chem. Soc. 1962, 138 (1962).
97. R. J. O'Brien, P. J. Green, and R. A. Doty, "Interaction of Oxides
of Nitrogen with Aromatic Hydrocarbons," 175th National Meeting
of the American Chemical Society, March 1978.
98. D. R. Fitz, D. Grosjean, K. Van Cauwenbergher, and J. N. Pitts,
"Photooxidation Products of Toluene-NO Mixtures Under Simulated
^f
Atmospheric Conditions," ibid.
99. J. G. Calvert, Environ. Sci. Tech., W, 256 (1976).
100. R. Atkinson and J. N. Pitts, J. Phys. Chem., 7j), 295 (1975).
101. T. W. Nakagawa, L. J. Andrews, and R. M. Keefer, J. Am. Chem. Soc.,
£2, 269 (1960).
102. D. G. Hendry, T. Mill, L. Piszkiewicz, J. A. Howard, and H. K.
Eigenmann, J. Phys. Chem. Ref. Data, J3, 937 (1974).
103. R. A. Kenley, J. E. Davenport, D. G. Hendry, J. Phys. Chem.. 82.
1095 (1978).
104. H. Niki, P. D. Maker, C. M. Savage, and L. P. Breitenbach, J. Phys.
Chem.. 82, 132 (1978).
105. E. D. Pellizari, J. E. Bunch, and B. H. Carpenter, Environ, Sci.
Tech., ^, 552 (1975).
106. E. D. Pellizari, B. H. Carpenter, and J. E. Bunch, Environ. Sci.
Tech., 9, 556 (1975).
68
-------
107 J. A. Howard, in Advances in Free-Radical Chemistry, Vol. IV, p. 99
(Academic, New York, 1972).
108. R. A. Kenley and D. G. Hendry, J. Am. Chem. Soc.. J9£, 3198 (1977).
109. H. Niki, P. F. Maker, C. M. Savage, and L. P. Breitenbach, "Fourier
Transform IR Studies of Gaseous and Particulate Nitrogenous Com-
pounds of Atmospheric Interest," 175th National Meeting of the
American Chemical Society, March 1978.
69
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA -600/3-79-020
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
REACTIONS OF OXY RADICALS IN THE ATMOSPHERE
5. REPORT DATE
March 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
D.G. Hendry, R.A. Kenley, J.E. Davenport, and
B.Y. Lan
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
SRI International
333 Ravenswood Ave.
Menlo Park, California 94025
10. PROGRAM ELEMENT NO.
1AA603 AC-21 (FY-78)
11. CONTRACT/GRANT NO.
Grant No. 603864
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory-RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final 6/75 - 6/78
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Results are presented of a research program concerned with the study of
selected reactions of importance in atmospheric chemistry. The decomposition of
peroxyacetyl nitrate (PAN) was studied over the temperature range 25-39 C. The
rate constant was determined to be log k = 16.29 - 26,910/4.576 T. The reactions
of acetylperoxy radicals with NO and NO- were investigated. The ratio of the rate
constants for these reactions was determined to be k/k = 3.0+0.7.
The products of the reaction of OH with various aromatic compounds were also
determined. The investigation showed that the reaction of OH with simple aromatic
hydrocarbons proceeds by two major pathways, abstraction of a hydrogen atom in
the benzylic position or addition of OH to the aromatic ring. Ratios of the rate
of abstraction versus addition were determined for toluene, 1,4-dimethylbenzene
and 1,3,5-trimethylbenzene.
Results of a study to elucidate the products and mechanism of the reaction
of OH with benzaldehyde are also presented. Research showed that this reaction
proceeds exclusively by abstraction of the aldehydic hydrogen.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
* Air pollution
* Photochemical reactions
* Reaction kinetics
* Oxidation
* Aromatic hydrocarbons
* Peroxy organic compounds
13B
07E
07D
07B
07C
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
78
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
70
------- |