EPA-600/3-77-111
October 1977
Ecological Research Series
JN«STt6ATO OF mmWl H Wm
PERTURBED
IN THE
«e s.
j.
-------
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-600/3-77-111
October 1977
INVESTIGATIONS OF IMPORTANT HYDROXYL RADICAL REACTIONS
IN THE PERTURBED TROPOSPHERE
by
D. D. Davis
Engineering Experiment Station
Georgia Institute of Technology
Atlanta, Georgia 30332
804629
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
U. S. ZNY." ' '
EJHSSfl, N. ,.
-------
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 endorsement or
recommendation for use.
n
-------
ABSTRACT
The flash-photolysis resonance fluorescence technique has been utilized
to study the reaction kinetics of hydroxyl radicals with ten aromatic and six
olefinic hydrocarbons at 298 K and several diluent gas pressures. The aromatic
compounds that were studied include benzene, toluene, ethylbenzene, n-propyl-
benzene, isopropylbenzene, hexafluorobenzene, n-propyl pentafluorobenzene,
and o-, m-, and p-xylenes; and the olefins include ethylene, acetylene, pro-
pylene, 1-butene, cis-2-butene, and tetramethylethylene. Based on our exten-
sive data on OH-substituted aromatic hydrocarbon reactions, it has been
inferred that addition of hydroxyl radicals to the aromatic ring is the
dominant reaction in these systems.In the case of OH-olefin reactions,addition
of OH to the double bond seems to be a prominent path for the heavier unsatu-
rates. From these rate constant data the lifetimes of all these hydrocarbons
in the lower troposphere has been calculated. Utilizing the technique of
laser flash photolysis, time-of-flight mass spectrometry, attempts were made
to understand the mechanisms involved in the reactions of OH with substituted
aromatic hydrocarbons.
m
-------
CONTENTS
Abstract iii
Figures vi
Tables vii
1. Introduction 1
2. A Kinetics Study of the Reaction of the OH Free
Radical with Aromatic Compounds. 1. Absolute
Rate Constants for Reaction with Benzene and
Toluene at 300 K 3
Introduction 3
Experimental Section 4
Results and Discussion 5
3. A Kinetics Study of the Reaction of OH Radicals
with Two C2 Hydrocarbons: CpH^ and C2H2 11
Introduction " 11
Experimental Section 14
Results and Discussion 17
A. OH + C^ > Products 17
B. OH + C2H2 * Products 21
Discussion of Previous Work 28
A. OH + C2Hlt > Products 28
B. OH + C2H2 > Products 30
4. A Kinetics Study of the Reaction of OH Radicals
with Aromatics and Olefins 33
Introduction 33
Experimental Section 35
Results and Discussion 38
Substituted Aromatics 51
OH + Propylene >- Products 56
OH + 1-Butene, c^-2-Butene, and
Tetramethylethylene 60
Atmospheric Implications 61
5. Identification of OH-Hydrocarbon Reaction Products 65
Overview 65
Experimental Approach 67
Experimental Results 69
References . . . 75
-------
FIGURES
Number Page
2-1 A 1/k vs. 1/p Lindemann plot showing the pressure
dependence of the reaction of OH with benzene
and toluene 8
3-1 A plot of log k2 (the bimolecular rate constant) vs.
log P(He) for the reaction of OH with ethylene 22
3-2 Kinetics of the OH + C2H2 reaction. A plot of the
pseudo-first order rate constant as a function
of both acetylene pressure and flash energy 25
5-1 A schematic drawing of the Laser Photolysis - T.O.F.
Mass Spectrometer System 68
-------
TABLES
Number Page
2-1
2-II
3-1
3-II
4-1
4-II
Rate Data for the Reaction of OH with Benzene at 298 K ....
Rate Data for the Reaction of OH with Toluene at 298 K ....
Rate Data for the Reaction of OH with Ethyl ene at 298 K. . . .
Rate Data for the Reaction of OH with Acetylene at 298 K . . .
Rate Data for the Reaction of OH with Ethyl enzene at 298 K . .
Rate Data for the Reaction of OH with N-Propyl benzene
at 298 K
. 6
. 7
. 18
. 23
. 39
40
4-111 Rate Data for the Reaction of OH with Isopropylbenzene
at 298 K 41
4-IV Rate Data for the Reaction of OH with N-Propyl
Pentafluorobenzene at 298 K 42
4-V Rate Data for the Reaction of OH with Hexafluorobenzene
at 298 K 43
4-VI Rate Data for the Reaction of OH with 0-Xylene at 298 K 44
4-VII Rate Data for the Reaction of OH with M-Xylene at 298 K 45
4-VI11 Rate Data for the Reaction of OH with P-Xylene at 298 K 46
4-IX Rate Data for the Reaction of OH with Propylene at 298 K .... 47
4-X Rate Data for the Reaction of OH with 1-Butene at 298 K 48
4-XI Rate Data for the Reaction of OH with Cis-2-Butene
at 298 K 49
4-XII Rate Data for the Reaction of OH with Tetramethylethylene
at 298 K 50
4-XIII Summary of Rate Data for OH + Aromatics 52
vii
-------
TABLES (Continued)
Number Page
4-XIV Summary of Rate Data for OH + 01efin 57
4-XV Lifetimes of Hydrocarbons in the Lower Troposphere 63
vm
-------
SECTION 1
INTRODUCTION
In the last ten years, atmospheric chemistry has increased in both
breadth and significance. During this time period, there has been a growing
sense of awareness on the part of the scientific community and the general
public that our planet's atmosphere, for all its enormity, is still rather
fragile. The atmosphere is not a passive reservoir, capable of storing end-
less amounts of waste products from increasing industrialization. Trace gases
and aerosols injected into the atmosphere by man interact chemically and
physically with natural constituents. Some toxic species are produced; some
essential or productive species are depleted; and the concentrations of those
trace constituents which are primarily responsible for maintaining the radia-
tive and precipitation budgets of the atmosphere are being altered.
A complex web of competing and successive photochemical, free radical,
and molecular reaction rates determine, in combination with poorly understood
molecular diffusion considerations, the nature of the atmosphere. An assess-
ment of the rates of formation or degradation of atmospheric species necessi-
tates a knowledge of the molecular concentrations, reaction rate constants,
and mechanisms of reaction of processes involving both the species of interest
and all influencing sources and sinks.
Hydroxyl radical reactions are among the most important laboratory
1
-------
studies presently being conducted in atmospheric chemistry. Among the key
processes now receiving considerable attention are those involving the reac-
tions of hydroxyl radicals with organic pollutants. The extremely reactive
OH radical is thought to be the principal means of removal for most airborne
toxic chemicals (some undoubtedly carcinogenic) released by a wide assortment
of man's industrial activities.
The spectrum of organic pollutants of interest here includes: chemical
industry solvents, such as benzene, toluene, tetrahydrofuran, xylenes, etc.;
chlorinated solvents used in metal-working plants, dry cleaning processes and
paints; and saturated and unsaturated hydrocarbons emitted from internal com-
bustion engines. The photochemical smog cycle is, in fact, a special case
within a localized air mass of the degradation cycle described above. Thus,
our efforts which have been directed towards obtaining rate constants and in-
vestigating the mechanisms of the OH-hydrocarbon reactions has local, regional,
and global importance in terms of air management.
We have described in the following sections the results of our investiga-
tions on OH-hydrocarbon reactions. They are:
Section 2: "A Kinetics Study of the. Reaction of the OH Free Radical
with Aromatic Compounds. 1. Absolute Rate Constants for
Reaction with Benzene and Toluene at 300°K"
Section 3: "A Kinetics Study of the Reaction of OH Radicals with Two
C2 Hydrocarbons: C2Hit and C2H2"
Section 4: "A Kinetics Study of the Reaction of OH Radicals with
Aromatics and Olefins"
Section 5: "Identification of OH-Hydrocarbon Reaction Products"
-------
SECTION 2
A KINETICS STUDY OF THE REACTION OF THE OH FREE RADICAL
WITH AROMATIC COMPOUNDS. 1. ABSOLUTE RATE CONSTANTS
FOR REACTION WITH BENZENE AND TOLUENE AT 300°K
INTRODUCTION
Whereas considerable work has been reported on the reaction of OH with
paraffinic and olefinic hydrocarbons,1"8 virtually no systematic study has
been carried out on aromatics. We wish to report, therefore, a recent study
completed in our laboratory involving the reaction of OH with two aromatic
compounds, benzene and toluene.
OH + K) E-* Products (1)
OH + l( )1 Products (2)
To the best of our knowledge, the absolute rate constants given represent the
first absolute rate constants reported for these organic species. The impor-
tance of these new measurements (in addition to their obvious fundamental
significance to kinetics) lies in the fact that OH-aromatic reactions are of
major concern in the combustion of nonleaded gasoline and in the formation of
photochemical smog. For example, recent examinations of the emissions from
cars running on regular nonleaded gasoline have shown that over 20% of the
hydrocarbons emitted were aromatics.9
-------
EXPERIMENTAL SECTION
In this study the reaction of OH with benzene and toluene was followed
by monitoring the concentration of OH as a function of time. The detection
technique for OH was that of resonance fluorescence which has been discussed
in detail in previous publications.10'11 The photolysis of H20 (in the region
above the CaF2 cut off at 1250 A) as a source of OH has also been described
in earlier work.11
In all experiments reported here, gas mixtures were made up using an all-
glass gas handling system. The toluene and benzene used in this study were
from Fischer Scientific Co. and had a purity level of 99.96% or better. All
low-pressure measurements of toluene, benzene, and H20 were made using a MKS
Baratron. High-pressure measurements (10-800 Torr) were made with a two-turn
Bourdon gauge (Wallace and Tiernan Type FA145). The precision to which gas
mixtures could be prepared, with the exception of H20, was estimated to be
or better.
-------
RESULTS AND DISCUSSION
The results from experiments performed at various total helium pressures
are shown in Tables I and II. Of considerable interest here is the observed
pressure dependence for the reaction of OH with both benzene and toluene; also
there is the fact that for both compounds rather large k values were measured
at 100 Torr He pressure, kx = 1.59 ± 0.12 x 10"12 and k2 = 6.11 ± 0.40 x 10'12
cm3 molecule"1 sec"1. In each case, the bimolecular rate constants reported
were obtained from the slope of a plot of the pseudo-first order rate constant
k vs. the aromatic concentration. The nonzero value for k at zero reactant
pressure represents the loss rate of OH due to diffusion out of the sampling
region.10*11 A close examination of the pressure dependence shown in Tables I
and II for benzene and toluene indicates that over the pressure range studied
in this work the reported rate constants are in the pressure fall off region
for each reaction (e.g., the rate constant is not a true third-order rate
constant nor is it a true bimolecular rate). The rate constant reported,
therefore, is calculated in the form of a bimolecular rate constant at each
total gas pressure employed. To better estimate the high-pressure limiting k
value for these processes, a Lindemann plot (1/k vs. 1/p) is given for both
aromatics in Figure 1. This figure clearly shows that benzene has a greater
pressure dependency than toluene as might be expected; but, somewhat surpris-
ingly, it also indicates that a very significant fraction of the total re-
action of toluene with OH proceeds by the addition of OH to the aromatic
ring. The evidence here is the fact that the change in the effective
5
-------
TABLE I. Rate Data for the Reaction of OH with
Benzene at 298 Ktt
[Reactant]
m Torr
0
2
4
8
0
0.5
3
5
5
5
5
8
0
3
6
9
[Diluent]
Torr
3 (He)
3
3
3
20 (He)
20
20
20
20
20
20
20
100 (He)
100
100
100
Flash
Energy ,J
88
88
88
88
88
88
88
88
88fa
45
180
88
88
88
88C
88
k',s-l
210
267
322
430
68
87
205
295
330
290
330
420
36
190
350
500
3 -1 -1
k. ., cm molecule s
(0.849 ± 0.08) x 10"12
(1.36 ± 0.09) x 10"12
(1.59 ± 0.12) x 10"12
aln all experiments, CaF2 window was used and the pressure of H20, the
photolyte, was 100 m Torr.
^150 flashes/gas filling (for all other experiments only 30 flashes/filling
was used).
C200 m Torr of H20
-------
TABLE II. Rate Data for the Reaction of OH with
Toluene at 298 Ka
[Reactant]
m Torr
0
1
1.5
2
3
0
0.5
1
1.5
1.5
1.5
1.5
2
3
4.5
0
1
2
3
[Diluent]
Torr
3 (He)
3
3
3
3
20 (He)
20
20
20
20
20
20
20
20
20
100 (He)
100
100
100
Flash
Energy ,J
88
88
88
88
88
88
88
88
88
88b
45
180
88
88
88
88
88
88
88
k'.s-1
210
325
390
450
565
68
170
240
330
310
290
350
390
555
790
36
250
425
630
3 -1 -1
k.. , cm molecule s
(3.60 ± 0.26) x 10"12
(5.00 ± 0.18) x 10"12
(6.11 ± 0.40) x 10"12
corresponding footnotes to Table I,
-------
pHe(Torr)
100
20
10
12-
IC"
o
*. 8
o
Benzene
Toluene
10
20
30
103/P
Figure 2-1. A 1/k vs 1/p Lindemann plot showing the pressure
dependence of the reaction of OH with benzene and toluene.
-------
bimolecular rate constant from 3 Torr to 100 Torr is nearly a factor of 2.
Since only the addition reaction would show a pressure dependence, it is con-
cluded that at least half of the total reaction is additions. Because of the
weak benzyl carbon-hydrogen bond in toluene it is to be expected that some
abstraction is also occurring although our data only indicate that the impor-
tance of this process is probably less than 50% of the total reaction. The
dependence of the bimolecular rate constant on the total pressure for both
reactions (1) and (2) can be explained on the basis of OH adding directly to
the aromatic ring. A possible explanation for the observed difference in the
pressure dependency for the two reactions is the larger number of degrees of
freedom available in the case of toluene for stabilization of the transition
complex. Figure 1 also indicates that toluene is more reactive than benzene
by at least a factor of 4. This can be explained by the higher efficiency of
the addition process for reaction (2), but also important, as indicated above,
is the fact that reaction (2) can proceed by abstraction as well as by addi-
tion. The second process would involve abstracting a hydrogen atom from the
methyl group on toluene.
There has been much speculation as to whether OH only abstracts the a
hydrogens of branched aromatics or whether addition to the ring in possible.
This study indicates that the addition process is very important for toluene
and will therefore have a very significant effect on the product distribution
resulting from reaction (2).
Concerning the possible role of aromatics in smog formation, an examina-
tion of Figure 1 would indicate that at near atomospheric pressure (for the
case of M = He) the respective rate constants for processes 1 and 2 would
probably be very close to their 100 Torr values, 1.59 x 10'12 and 6.11 x 10"12
-------
cm3 molecule'1 sec"1. Using an estimated OH steady-state concentration for
the atmosphere of 5 x 106 molecules/cc would therefore give a 1/e lifetime
for benzene and toluene of ^36 and 10 hr under daylight conditions. (This
would probably be even shorter if rate constants for M = N2 were to be used.)
These simple calculations, along with available concentration data, corroborate
the findings that aromatic compounds contribute to the formation of photochemi-
cal smog in areas involving heavy automobile traffic.
10
-------
SECTION 3
A KINETICS STUDY OF THE REACTION OF OH RADICALS WITH
TWO C2 HYDROCARBONS: C2H4 AND C2H2
INTRODUCTION
Hydroxyl radical reactions have been the subject of considerable interest
in recent years due to their importance in combustion processes and in the
chemistry of the atmosphere. Ethylene is thought to be a minor, but signifi-
cant constituent of the effluents emitted from automobile exhausts,12 and its
presence in urban atmospheres has been shown to increase the rate of conver-
13
si on of NO to N02, which is the precursor for the formation of ozone. The
reaction of hydroxyl radicals with ethylene is now thought to be an important
step in this process.
CzfV + OH » Products (e.g., C2H50) . (1)
The reaction of hydroxyl radicals with acetylene is of importance in the de-
gradation of acetylene to C02, both in the atmosphere and in diffusion con-
trolled flames.
C2H2 + OH 1- Products (e.g., C2H20 + H) . (2)
As a result of the importance of reactions (1) and (2) there have been
numerous studies of these chemical systems.3-5,14-18 jn ^ne case Of reaction
(1), the values reported by different workers for the rate constant at ambient
temperatures (^300°K) are at variance by as much as a factor of 3.3-5,14-15
Most of these previous studies, which were performed in both flow and static
11
-------
systems using a wide variety of detection techniques, were carried out at
total pressures lower than 20 Torr. In none of these investigations was a
systematic variation in pressure performed.
Morris, et al.3 qualitatively observed mass spectrometric adduct peaks
in their study of reaction (1), and proposed the primary process to be
OH + C2H!f » C2H50 . (la)
Bradley, et al.5 also reported observing the adduct peak, but in addition
observed peaks indicating that ethanol also plays a role in the overall
mechanism.
Of the studies on reaction (2) documented in the literature, only four
were performed at room temperature15"18; and the rate constants reported
varied by as much as a factor of 5. Those studies carried out in flow sys-
tems16'18 used low initial stoichiometries ([C2H2]0/[OH]0, ranging from 2.3
to 120), which resulted in stoichiometric corrections being required for most
of the rate constant data. However, only one study17 actually measured the
stoichiometric correction factor, n, while other workers reported either k2/n
or used the single value of n which had been previously published. A more
recent study performed,15 using the flash-photolysis resonance absorption
technique, employed moderately high initial stoichiometries, e.g., [C2H2]o/
[OHJo^lO2. That study reported that corrections to the rate data due to
secondary processes were not required.
Breen and Glass17 used mass spectrometric product analysis to obtain a
value for the stoichiometry of reaction (2). From this it was argued that
the following two channels were consistent with the data:
12
-------
OH + C2H2 > C2HO + H2 (2a)
OH + C2H2 » C2H + H20 , (2b)
while the following was not:
OH + C2H2 » CH3 + CO . (2c)
However, evidence obtained from a crossed molecular beam experiment (Gehring,
et a!.)19 supported reaction (2c) as being the dominant primary process.
Kanofsky, et al.,20 on the other hand, proposed that the primary process was
not (2a), (2b), or (2c), but rather
OH + C2H2 » C2H20 + H . (2d)
13
-------
EXPERIMENTAL SECTION
The flash photolysis-resonance fluorescence technique has been described
in great detail in previous publications11'21 and only a summary description
will therefore be presented in this text.
As in an earlier study involving reactions of the hydroxyl radical,11'21
photolysis of H20 was used as the source of OH:
H2o + to; X>llt0nm > H + OH (X2H) .
In the present study, reaction mixtures consisted of 50-300 mTorr of H20, 20-
500 Torr of He, and either 1-7 mTorr of C^ or 2-30 mTorr of C2H2. These
mixtures were photolyzed using a N2 spark flash lamp equipped with either a
CaF2 or Suprasil quartz window. CaF2 was used in the study of reaction (1)
almost exclusively to reduce the photolysis of C^. Similarly in the case
of reaction (2), Suprasil windows were normally used to minimize photolysis
of C2H2. Based upon the known absorption spectrum of H20 and previously con-
ducted actinometry on the flash lamp using ethylene as the actinic gas, it
was determined that within the spectral range 105-200 nm on the order of
1.5-36x 1011 OH radicals/cm3 were typically produced per flash in the re-
action cell (the precise flash energy, the spectral bandwidth, and the H20
concentration defined the actual concentration).
Excitation of OH was accomplished via the use of an OH resonance lamp.
This lamp primarily produced the emission characteristic of the (AE2+:v'=0) ->
(X2II:v"=0) transition of OH. A small fraction of the OH, produced by the
14
-------
photolysis of H20, was continuously excited by the emission from this lamp.
Fluorescence from excited OH was measured using a photomultiplier tube located
at right angles to the lamp. The intensity of this fluorescent emission was
found to be directly proportional to the amount of OH present in the mixture.
The fluorescence signal from the photomultiplier tube was then stored as a
function of time in a multichannel analyzer, operated in the multiscaling
mode.
Since the C2Hit and C2H2 pressures were adjusted to make reactions (1) and
(2) kinetically pseudo-first order with respect to OH, the observed rate of
OH disappearance was always exponential. Because the OH concentrations
utilized in the two studies were low ([OH]0<3.6 x 1012 molecule cm~3), the ob-
served signal levels were also low. Thus, multiple flashes on a single gas
mixture were required to produce a single smooth kinetic decay curve. How-
ever, the number of flashes per gas mixture was always restricted to minimize
the decomposition of C2H2 or C2Hlf to less than 3%. Thus, several fillings
of an identical gas mixture were used for the development of a single experi-
mental decay curve. The initial hydroxyl radical concentration was kept low
in order to ensure that the bimolecular disproportionate reaction of
hydroxyl radicals could be neglected, and to obtain high initial stoichio-
metries, typically ([C2H2]0/[OH]0>500), so that the importance of secondary
reactions could be minimized. Even with high initial stoichiometries, how-
ever, secondary reactions were not always eliminated (to be discussed later).
Since the observed kinetics were pseudo-first order, the first order
rate constants could be obtained from a plot of the logarithm of the count
rate in each channel of the analyzer (after the background was subtracted)
versus time. The slope of the line in each case was established by a least
15
-------
squares treatment of the data. This treatment was extended out to two and in
some cases three 1/e times of the decay. Bimolecular rate constants were ob-
tained from a least squares treatment of the slope of the line obtained from
a plot of the pseudo-first order rate constants versus C2H2 or C2Ht( pressure.
Each bimolecular rate constant, therefore, represents an average of more than
ten individual experiments.
Gas pressures of less than 3 Torr were measured using an MKS Baratron
pressure gauge which was periodically checked against a dibutyl phthalate
manometer. The high pressure measurements (10-800 Torr) were made with a
two-turn Bourdon gauge (Wallace and Tierman type FA-145). The precision to
which gas mixtures could be made, with the exception of H20, was estimated to
be 3% or better. The H20 pressure could not be metered so precisely due to
adsorption effects on the surfaces of the reaction cell.
The C2H2 and C2H4 used in these experiments were from Matheson Co. and
had stated purities of 99.6% and 99.98%. Both gases were degassed in liquid
nitrogen prior to use. Matheson "Gold Label Ultra-High Purity" helium was
used without further purification. The H20 used in all experiments was dis-
tilled and then degassed using liquid N2 prior to its introduction into the
gas handling system.
16
-------
RESULTS AND DISCUSSION
A. OH + C2H4 > Products
The results for reaction (1) are presented in Table I. It can be seen
that a wide variation in experimental conditions was performed in an effort
to show that kinetic complications were not affecting the observed rate con-
stants. These variations included using H20 pressures of 50 to 300 m Torr,
flash energies of 45-^500 J, and ethylene pressures of 1-7 m Torr. Under
these widely differing conditions the pseudo-first order rate constants were
found to be invariant (for any given C2H4 pressure) within the assigned ex-
perimental uncertainties of the measurements. The indications are, there-
fore, that OH reactions of the following type were not significant:
C2tti> + kv > fragments, (3)
OH + fragments > products, (4)
OH + OH + M » H202 + M, (5)
OH + OH > H20 + 0, (6)
H + OH + M > H20 + M, (7)
0 + OH » 02 + H. (8)
In addition, the reaction of OH with the products of (1) could not be impor-
tant. In all the above cases, the rate of removal of OH would depend upon
the square power of the flash intensity since the radical concentration in
these rate expressions depends directly upon the flash energy. If these re-
actions had been significant, then there would have been an observed
17
-------
TABLE I. Rate Data for the Reaction of OH with
Ethylene at 298 K
[Reactant]
m Torr
0
1
2
3
0
1
1
2
2
3
5
5
0
5
0
1
2
3
0
2
0
1
2
2
2
3
5
[Diluent]
Torr
3 (N2)
3
3
3
i
3 (He)
3
3
3
3
3
3
3
5 (He)
5
6 (He)
6
6
6
10 (He)
10
20 (He)
20
20
20
20
20
20
Flash
Energy ,J
320
320
500
500
88
88
500
88
500
500
88
500
88
320
88
88
88
88
88
88
88
88
45
88
500
88
88
k',*-1
83
220
325
420
205
290
295
370
365
430
525
575
140
585
120
230
330
445
105
340
70
215
350
365
380
500
700
3 -1 -1
k. . , cm molecule s
(3.64 ± 0.20) x 10"12
(2.24 ± 0.22) x 10"12
(2.79 ± 0.37) x 10"12
(3.32 ± 0.43) x 10"12
(3.63 ± 0.50) x 10"12
(4.06 ± 0.38) x 10"12
18
-------
TABLE I. (Continued)
[Reactant]
m Torr
0
1
2
3
5
0
0.5
1
1.5
2
[Diluent]
Torr
100 (He)
100
100
100
100
300 (He)
300
300
300
300
Flash
Energy, J
88
88
88
88
320
88
88
88
88
88
KM'1
40
204
350
510
650
28
140
210
267
362
3 -11
k. . , cm molecule s
(4.72 ± 0.60) x 10~12
(5.33 ± 0.65) x 10"12
The pressure of H^O, the photolyte, was 100 m Torr.
19
-------
dependence of kj upon the flash intensity. Finally, since the products of
ethylene photolysis consist chiefly of H2 and C2H2,22'23 experiments were
performed in which the number of flashes per gas filling was varied. The
range of this variation was 10 to 100. Reactions of possible importance here
would include the following:
OH + H2 > H20 + H (9)
OH + C2H2 1- products. (2)
The results of this test were that within the experimental error of the
measurements (±6%) no variation in the observed first-order rate constants
was noted. Even though secondary reactions (2)-(8) were found to be in-
significant under our experimental conditions, a CaF2 window (A>120 nm) was
used over the flash lamp to limit the extent to which C^ was photolyzed.
A possible source of systematic error in the above measurements could
involve rotationally or vibrationally excited OH species produced by the
flash pulse. Previous experiments11'21 utilizing 307 nm and 345 nm inter-
ference filters over the resonance lamps indicated that vibrationally ex-
cited OH, if present, was not directly detected by the system. If there was
a significant initial population of OH in the u"=l level, which was subse-
quently quenched into the v"=0 level, then an apparent underestimate in the
pseudo-first order rate constants would be noted due to the slow production
of ground state OH. However, previous experiments11'21 utilizing 40 m Torr
of CO, 20-100 Torr of He or 20 Torr of N2, and 50-300 m Torr of H20 for the
study of reaction (10)
OH + CO > C02 + H (10)
showed no evidence for quenching of excited OH outside of the experimental
uncertainty. If there had been a significant population of excited OH, then
20
-------
the observed rate constant would have varied with total pressure due to
quenching, and the decay plots would not be logarithmic as observed. Thus,
at the pressures utilized in this study, excited OH does not appear to have
been a problem in defining a value for either kx or k2.
The only experimental parameter which caused deviations in ki outside
the experimental uncertainty were variations in total pressure. Within the
experimental range of pressure 3-300 Torr of He, a variation of a factor of
3 in the experimentally determined value of Iq was observed. Experiments
utilizing 3 Torr of N2 were also performed. The He pressure data are pre-
sented in Figure 1 as the logarithm of the bimolecular rate constant versus
the logarithm of the total pressure. The observed results indicate that
reaction (1) is intermediate between second and third order in the pressure
range examined in this study.
The present results indicate that within the pressure range 3-300 Torr
of He, the bimolecular rate constant varies from 2.24 to 5.33 x 10"12 cm3
molecule"1 s"1. The estimated uncertainty in all first-order rate constants
was judged to be ±6%. The uncertainties in the bimolecular rate constants
represent two standard deviations of the mean, as determined from a linear
least-squares-curve fitting technique. In general, these uncertainties
ranged in value from 10 to 13%.
B. OH + C2H2 > Products
Table II summarizes the data used to compute the bimolecular rate con-
stant for reaction (2) at 300°K. As in the case of reaction (1), wide varia-
tions in experimental conditions were performed. The H20 pressure was re-
duced from 300 m Torr by a factor of 6, the total pressure (He) was varied
21
-------
. 1
r ^^^i ^~*
1 1 CO '.
r-~ co
cr> s- r^
< I QJ CT»
"O " C < I "O
c o>- - c
*O t- fO ftj
CD 3 C
C C C <4-
O £ « i i -*: E >-
_i I r-*. "O r-. - - CD o 5- -C
H O~H CT1 QJCTlOOM (O OJU
" 1 r i -> i 1 1~~ n: >>co
^ s C/) ^* ^ Ot "O O
.) C "31 »
S_ ^., fO >> S-
CD CO *^ ci^ ~O CD
c: -r- -i- P _c r c ^:
CD i- 3 -r- eo CO
S- O «-> E S- !-
CD S 00 CO CQ U_
DO »© O
HH o
.
I «fV I
I ^ 1
HH
^HH
b^
1 1 f 1 * 1 I
Lf) O LO O LD O LO
tr> in «d-
i i
^
10
^
>
1C
Q
1 ^
i *Q I
O U
CM r-
O
-o
CO
CO
>
*~~**
-l->
c
2
CO CD
c: c
»O 0 CD
O 0
i-H >)
CD *"
4-* ^->
ro CD
s_
f~
S» i '
(O !-
^ 5
3
^-x OH
S_ CD O
S_ r
0 O M-
H- E 0
JQ C
O O) O
TM S_ CD !-
3 J^ +->
co -MO
CO (O
CD CD
i_ CM S-
Q. -^
CD
-O CD CD.C
1-1 :n o -(->
S-
M- O
O M-
"^° +-> ^~-
*« ^^
CL
Ol
-00 .JO
CO
cr»
iZ
T
H
y
22
-------
TABLE II. Rate Data for the Reaction of OH with
Acetylene at 298 K
[Reactant]
m Torr
0
2
10
10
10
20
20
20
30
0
5
10
15
0
2
5
5
10
10
20
20
20
20
30
0
5
10
10
[Diluent]
Torr
20 (He)
20
20
20
20
20
20
20
20
50 (He)
50
50
50
100 (He)
100
100
100
100
100
100
100
100
100
100
500 (He)
500
500
500
H20
m Torr
100
100
100
100
100
100
150
150
150
300
300
300
300
300
300
300
300
300
50
300
300
300
300
300
300
300
300
300
Flash
Energy ,0
88
88
88
88
45
88
88
88
88
45
45
45
45
88
45
20
45
20
45
20
45
320
500
45
45
45
20
45
k',s kbi,cm molecule" s
45
58
105
97
103
155
155
160
205 (1.68 ± 0.12) x 10"13
43
69
99
125 (1.73 ± 0.13) x 10"13
25
33
45
48
72
76
135
125
225
340
185 (1.61 ± 0.17) x 10"13
22
48
72
71 (1.55 ± 0.18) x 10"13
23
-------
by a factor of 25 (20-500 Torr), and the acetylene pressure ranged from 2 to
30 m Torr. For a constant acetylene pressure, no significant variations of
the pseudo-first order rate constant outside of the experimental error were
observed for the above changes. Variation of the flash intensity by a factor
of ^25 (20-^500 J) did show significant perturbations in the first-order rate
constants.
Figure 2 shows the variation in the pseudo-first order rate constant as
a function of acetylene pressure for several flash energies. The dependence
of the pseudo-first order rate constant with flash intensity was thoroughly
investigated at a fixed total pressure (100 Torr) and fixed acetylene pressure
(20 m Torr). For flash energies of 320 J the observed first order rate con-
stants were nearly a factor of 2 faster than those obtained at energies <45 J.
The data collected using flash energies >45 J were prone to scatter and the
individual logarithmic decay plots were not always strictly linear, indicat-
ing the importance of secondary processes. These high flash energy experi-
ments were particularly sensitive to changes in the transmission characteris-
tics of the flash lamp window, as this governed the photon flux entering the
cell. Changes in transmission properties were due to material being deposited
upon the window from the electrodes, and to F-center formation in the window.
When the flash energy was varied below 88 J there was no perceptible deviation
in the observed rate constants outside of the bounds of the experimental error.
There are numerous possible explanations for the above observations, some
of which will now be explored in greater detail. The experimental data sug-
gest that the flash decomposition product of acetylene, C2H, should not be
important as its concentration can be calculated to be ^ ± [OH]0 (this is
based upon the relevant absorption cross-section data of H20 24 and C2H2,25»26
24
-------
0)
c/)
400-
350-
300-
250-
200-
150-
OH + CH
22
Products
I
5
T
10
cm3'molec"1'Sec~1
30
[C2H2], (mTorr)
Figure 3-2. Kinetics of the OH + C2H2 reaction. A plot of the
pseudo-first order rate constant as a function of both
acetylene pressure and flash energy.
25
-------
and the concentrations of these species in the reaction cell).
C2H2 + kv > C2H + H (11)
Therefore, even if C2H reacted on every collision with OH radicals, this
should not cause a significant increase in the net rate of removal of OH.
The H atoms formed in the photolysis of H20 are initially produced with a
concentration equal to that of the hydroxyl radicals, and can react with
acetylene to form CzHa:
H + C2H2 * C2H3 . (12)
The rate constant data for reaction (12) published by different workers are
not in good agreement27"30; and consequently, the equivalent bimolecular rate
constant at 100 Torr total pressure is not well established. It could be as
high as 10~12 cm3 molecule'1 sec'1. The product, C2H3, is expected to be
highly reactive towards OH radicals.
A further process which could be important in the removal of OH radicals
is that involving OH radicals reacting with the primary product of reaction
(2). For example,
OH + C2H20 » C2HO + H20 . (13)
Again the rate constant for this process could be expected to be rapid. The
overall rate of removal of OH can now be expressed as:
-d[OH]/dt = k2[OH][C2H2] + k'[OH][C2H] + k"[OH][C2H3] +k"1 [OH][C2H20] .
It can be seen that all processes, except for (2), depend upon the square
power of the flash intensity. It has been shown by these authors, using the
above kinetic scheme, that when [OH]0 = (1.5-3.5) x 1011 radical cm'3 there
appears to be no significant perturbation due to secondary processes; whereas,
when [OH]0 = (2.3-3.6) x 1012 radical cm'3, the experimental data and computer
26
-------
model calculations show the presence of significant secondary processes.
Other reactions of possible significance could involve electronically
excited C2H2 »
C2H* + OH > Products (14)
C2H* + C2H2 " MJ , etc. (15)
These species are thought to be long-lived enough to participate in reactions
with OH.31 Also, since excited C2H* is formed at long wavelengths where the
light flux from the flash lamp is more intense, these reactions could have
been of significant importance to the observed flash dependence of the first-
order rate constants.
At flash energies < 45 J and for total pressures ranging from 20 to 500
Torr of He, the bimolecular rate constant k2 was found to be 1.65±0.15x10~13
cm3 molecule-1 sec"1 independent of the total pressure.
27
-------
DISCUSSION OF PREVIOUS WORK
A. OH + C2Hit - * products
There have been six room-temperature studies of Reaction (1). 3-5,1^-16
The earliest study was performed by Wilson and Westenberg16using a discharge-
flow system coupled with ESR detection of OH. They reported a value for ki
of 5xlO~12/n cm3 molecule"1 .sec"1; where n is a stoichiometric correction
factor which represents the number of OH radicals that react for each initial
C2H1+ that reacts (i.e., n >1 due to OH reacting with species produced in the
primary and subsequent processes). Considering the low pressures utilized in
flow systems, the value of n which is required in order to bring their
results into agreement with the present study would have to be approximately
2.
Greiner, 4 utilizing kinetic absorption spectroscopy, reported a value
for k j of 5x10 ~12 cm 3 molecule "Isec " . No mention was made of the total
pressure used in his system. The experimental data was corrected in order to
allow for secondary reactions caused by reaction products. However, it was
assumed that the photolysis products of ethyl ene would not cause any
appreciable increase in the rate of removal of OH radicals. This assumption
appears to be valid in that the photolysis products of £2^^ are probably
C2H2 and \\2, and these both react slower with OH radicals than does C2Hi+.
Also, if the total pressure in his system was 100 torr, as was used in
1 2 32
previous studies, ' ' then the results obtained are in substantial agree-
28
-------
ment with those obtained in the present study at 100 torr total pressure of
He.
Morris, Stedman, and Niki3 used a discharge-flow system at 1 torr total
pressure of He coupled with mass spectrometric detection of OH under both
ethylene-rich and OH-rich conditions. Their value of k1=1.8xlO"12cm3
molecule'1.sec"1, obtained by monitoring C2Hit in the presence of an excess
concentration of OH radicals, is in substantial agreement with our experi-
mental value of ki at 3 torr total He pressure. This study did not require
a stoichiometric correction factor, as the primary products of Reaction (1)
would preferentially react with the excess concentration of OH radicals and
not with the CzH^ which was being monitored. The value reported for the
ethylene-rich system, where OH radicals were monitored, was 2.5/nxlO~12cm3
molecule-1.sec-1. No value was reported for n.
Stuhl,14 utilized the resonance fluorescence technique to monitor
hydroxyl radicals in the presence of an excess concentration (>100) of C2Hi+
and obtained a value for kx of 3±lxlO~12cm3 molecule'1.sec"1at 300°K and
20 torr total pressure of He. This value is slightly lower than that
obtained in the present study at the same total pressure, k,=4.1±0.2xlO-12cm3
molecule"1.sec"1. However, it can be seen that the two values are in agree-
ment within the expressed uncertainties. Secondary reactions were shown to
be unimportant under their experimental conditions. No mention of any
pressure dependence was noted by Stuhl.
Smith and Zellner15 used the flash photolysis-resonance absorption tech-
nique to study Reaction (1) in a static cell at either, (a) 10-20 torr total
pressure of He or, (b) 10 torr of H2 and 10 torr of N20. High initial con-
centrations of hydroxyl radicals were used ([OH]
-------
the authors reported that this did not lead to the presence of complicating
secondary reactions. They found a value for kx of 5.2xlO~12 cm3 molecule"1.
sec~ which was invariant with total pressure or diluent gas. Although the
value obtained in the present study with 20 torr of He (k1=4.2±0.2xlO~12 cm3
molecule"1.sec'1) is within 25% of their value, it is felt that the two
studies are not in particularly good agreement as N20 would be expected to be
a more efficient third body than He.
Bradley et a_1__.5 studied Reaction (1) in a discharge flow system at 3
torr total pressure (He), using ESR detection of hydroxyl radicals. Low
initial stoichiometries, ([C2H[+]0/[OH]0=5.5=13.1) were used, which resulted
in a stoichiometric correction factor, n, being required to allow for
secondary removal of hydroxyl radicals. A mass spectrometric end product
analysis yielded a value of 2.6 for n, which when combined with the experi-
mental value of kj/n, produced a value of 1.67±0.5xlO~12 cm3 molecule'1.sec'1
for kj. This value is slightly lower than that obtained in the present study
at 3 torr of He, but within the reported experimental uncertainties of the
two studies.
In summary, from an examination of all the previous data on kl it is
seen that when the pressure dependence of Reaction (1) is considered, the
previously published results are in reasonably close agreement.
B. OH + CzH2 v products
Of the room temperature studies of Reaction (2) that have been reported,
the ealiest work was done by Wilson and Westenberg16 utilizing ESR detection
of OH generated by the reaction H + N02 -> OH + NO in a fast flow system.
They obtained the result that k2=1.0xlO-12/n cms molecule-1.sec-1, where n is
30
-------
a stoichiometric correction factor. For there to be substantial agreement
between the present work and that of Wilson and Westenberg, a stoichiometric
correction factor of 5-6 would have to be employed. This correction appears
to be too large and does not fully account for the difference between the two
measurements.
A measurement by Breen and Glass 17is in better agreement with the
present work. They used a discharge-flow system coupled with ESR detection
of OH which was again produced from the H/N02 reaction. A stoichiometry
correction factor of 2 was required for OH consumption via secondary
reactions. This was obtained from a mass spectrometric analysis of the end
products of the reaction. The rate of consumption of OH radicals in the
presence of an excess concentration of CoH9([C9H~] /[OH] = 3.7-57.0} was
^ ^ ^ o o
monitored via ESR detection, and the decay curve was modeled with an assumed
kinetic scheme to calculate k2(eff)=k2/n (n = stoichiometric correction
factor). From the values of k2(eff) and n, a value of 1.9±0.6xlO~13cm3
molecule"1.sec^was obtained. This value is in good agreement with that
obtained from the present study.
A study of Reaction (2) was performed by Smith and Zellner utilizing the
flash photolysis-resonance absorption technique. They obtained a bimolecular
rate constant of 3.7x10 ~13cm3 molecule"1.sec"1, a result which is approxi-
mately a factor of 5 -higher than that obtained in the present work. High
flash energies on the order of 180-500 J were used, producing initial
hydroxyl radical and hydrogen atom concentrations of <3xl013 radical cnr3.
This could have resulted in secondary processes becoming a highly significant
perturbation in their system even though their initial [C2H2] /[OH] ratio
was typically 100 (estimated from the experimental data given in their paper).
31
-------
Thus, the stoichiometric correction factor could have been quite large. This
line of argument presupposes that the secondary reactions of C2H3 (from
reaction of H and C2H2) and C2H20 with OH radicals are rapid.
Recently Pastrana and Carr18 studied Reaction (2) in a flow system
utilizing resonance line absorption detection of OH. Acetylene to OH ratios
of (a) 2.3-13.2 and (b) 14-125 were used in two separate studies. The values
obtained for the bimolecular rate constant k2 were (2.9±0.3)/nxlO~13 and
(2.1±0.6)/nxlO~13 cm3 molecule'1.sec"1, respectively, where n is the stoi-
chiometric correction factor. They assumed that n had a value of 2.1 (as
measured by Breen and Glass17) in series (a), yielding a bimolecular rate
constant of (1.4±0.3)xlO"13 cm3 molecule'1.sec"1. In series (b) they
assumed that experimental conditions were such that secondary processes would
be of less significance and that n would have a value close to unity.
Although secondary reactions may not be so important in series (b), it is
quite conceivable that n could be significantly greater than unity. The
overall value reported for k2 was (2.0±0.6)xlO"13cm3 molecule'1.sec'1. It
can be seen that this value is in good agreement with that obtained in the
present study.
In summary, the present study has shown .that there is no pressure
dependency for the rate of Reaction (2). However, the rate of removal of
hydroxyl radicals is extremely sensitive to secondary reactions. The species
most likely to participate in secondary removal of OH are: (a) the primary
product of Reaction (2), e.g., C2H20 and (b) the C2H3 radical formed in the
H + acetylene reaction. The value reported for k2 in the present study was
obtained using low flash energies and therefore should be free of any signi-
ficant perturbation due to secondary processes.
32
-------
SECTION 4
A KINETICS STUDY OF THE REACTION OF OH RADICALS WITH
AROMATICS AND OLEFINS
INTRODUCTION
The hydroxyl radical is known to be one of the most reactive species in
the atmosphere.33 In particular, its reaction with hydrocarbons in the tropo-
sphere is believed to be the key initiating step in the oxidation of these
compounds, thereby leading to the formation of ozone via a complex chemical
degradation cycle involving NOX species33 Thus, since recent measurements
have indicated the presence of a considerable amount of aromatic and olefinic
hydrocarbons in the troposphere^it is apparent that an understanding of the
reactivity of OH with hydrocarbons is essential to a realistic evaluation of
the ozone budget in both the perturbed and the unperturbed troposphere. In
addition to its key importance in the field of tropospheric chemistry, OH-
hydrocarbon reactions are also of crucial importance in combustion systems35
There have been numerous rate constant measurements for the reaction of
hydroxyl radicals with a variety of hydrocarbons36 Whereas reasonably
good agreement exists for the reaction rate constants for OH plus saturated
hydrocarbons, very little agreement can be found for those rate constants
measured for OH-olefin processes. In the case of aromatic species, there is
a major absence of data, especially involving substituted aromatic hydro-
carbons. For this reason, we have initiated a new study to determine the
33
-------
rate constants for the reactions of the hydroxyl radical with both olefinic
and aromatic hydrocarbons. The data for olefin reactions were presented in an
ACS meeting.50 Those reactions investigated include:
OH + o-xylene
OH + m-xylene
OH + p-xylene
ethyl benzene
n-propylbenzene
isopropylbenzene
hexafluorobenzene
n-propyl hexafluorobenzene
propylene
1-butene
c^&-E-butene
tetramethylethylene
products
products
products
products
products
products
products
products
products
products
products
-* products
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
34
-------
EXPERIMENTAL SECTION
The experimental details and operating principles of the flash
photolysis-resonance fluorescence techniques have been fully described in the
literature;10'11'21 hence, only a brief summary will be presented in this text,
The reaction vessel used in this work was made of quartz and was equip-
ped with either a calcium fluoride or a quartz window through which the
photolyzing beam could enter the cell. As in earlier studies involving OH
radicals, H20 was photolyzed using a N2 spark discharge lamp to produce a
typical ground state OH concentration of 3 x 1011 cnr3:
H20 + hv » H + OH(X2n)
Excitation of OH was accomplished via the use of an OH resonance lamp.
This lamp primarily produced the emission characteristic of the (AZ2+:u'=0) >
(X2n:v"=0) transition of OH. A small fraction of the OH, produced by the
photolysis of HaO, was continuously excited by the emission from this lamp.
Fluorescence from excited OH was measured using a photomultiplier tube lo-
cated at right angles to the lamp. The intensity of this fluorescent emis-
sion was found to be directly proportional to the amount of OH present in
the mixture. The fluorescence signal from the photomultiplier tube was then
stored as a function of time in a multichannel analyzer, operated in the
multiseal ing mode.
Since the hydrocarbon pressures were in large excess relative to the OH
concentration, the observed kinetic decays for reactions (1)-(12) were
35
-------
pseudo-first order with respect to OH. Under typical operating conditions,
the hydroxyl radical concentration was kept sufficiently low that the initial
stoichiometry of the system (i.e., [hydrocarbons]0/[OH]0) was always greater
than 50. Because of the low initial OH concentration, multiple flashes (<60)
on each of many identical gas mixtures were required to produce a single
smooth kinetic decay curve. The pseudo-first order rate constants were then
obtained from a least square analysis of a plot of the logarithm of the count
rate in each channel of the analyzer (after the background was subtracted) vs.
time. The bimolecular rate constants were obtained from a least squares
treatment of the slope of the line obtained from a plot of the pseudo-first
order rate constants vs. hydrocarbon concentrations.
Gas pressures of less than 3 Torr were measured using an MKS Baratron
pressure gauge which was periodically checked against a dibutyl phthalate
manometer. The high pressure measurements (10-800 Torr) were made with a two-
turn Bourdon gauge (Wallace and Tierman type FA-145). The precision to which
gas mixtures could be made, with the exception of H20, was determined to be
<3% or better. The H20 pressure could not be metered so precisely because of
absorption effects on the surfaces of the reaction cell. In the latter case,
the precision was between 10-20%.
The purity of each of the hydrocarbons used in this study was as follows:
propene (Matheson, 99.7%), 1-butene (Matheson, 99.9%), cis-2-butene (Matheson,
99.9%), tetramethylethylene (Aldrich Chem. Co., >99%), ethyl benzene (Baker-
Ultrex, 99.99%), n-propyl benzene (Baker-UHrex, 99.9%), isopropyl benzene
(Baker-Ultrex, 99.97%, o-xylene (Baker-Ultrex, 99.99%), m-xylene (Baker, >96%),
p-xylene (Baker-Ultrex, 99.99%), hexafluorobenzene (PCR, >99.7%), n-propyl
hexafluorobenzene (PCR, >94%, the main impurity is isopropyl hexafluorobenzene).
36
-------
All hydrocarbons were subjected to multiple freeze-pump-thaw cycles before
use. The olefins were checked for their purity using a mass spectrometer.
Matheson gold label high-purity helium and argon were used without further
purification. The H20 used in this study was double distilled and before
use was subjected to two freeze-pump-thaw cycles.
37
-------
RESULTS AND DISCUSSION
The results from this investigation have been summarized in Tables I-XII.
For most of the reactions studied, a wide range of experimental conditions was
covered in order to verify that kinetic complications due to secondary pro-
cesses were not affecting the observed OH decay rates. These variations in-
cluded changing the photo-flash energy, the total pressure of the system, the
hydrocarbon reactant gas pressure,and the number of flashes to which a given
gas mixture was subjected. Since the initial concentrations of both OH and
photofragments from the hydrocarbon reactant would necessarily increase linear-
ly with an increasing photolyzing photon flux(i.e.,the flash energy), the
decay of OH via reaction with photofragments would depend on the square of
the flash energy. If this path had been a significant fraction of the
measured pseudo-first order rate constant, k', changes in the flash energy
of a factor of four or more would have had a significant effect on the value
of k1. As can be seen in Tables IV, V, VIII, X, and XI, this obviously was
not the case in our study. (Even though flash variations were not performed
on every compound investigated, the similarity of these compounds to those
exposed to flash variations would strongly point to the absence of anomalous
behavior for these other compounds.) Moreover, a calculation of the contri-
bution of this type of radical-radical secondary reaction, even with the
assumption that it occurs at gas kinetic collision rates, indicates that the
process is unimportant under our experimental conditions. (In this calcula-
tion the absorption cross section for the hydrocarbons is assumed to be
38
-------
TABLE I. Rate Data for the Reaction of OH with
Ethyl benzene at 298 K
[Reactant]
m Torr
0
0.23
0.46
1.00
0
0.25
0.25
0.50
0.75
0.75
1.00
1.50
0.50
0.50
1.00
1.50
2.00
[Diluent]
Torr
3 (He)
3
3
3
20 (He)
20
20
20
20
20
20
20
200 (He)
200
200
200
200
Flash
Energy ,J
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
Flashes/
Filling
60
60
60
60
60
60
60
60
60
60
60
60
30
30
30
30
30
k'.s-1
186
250
301
430
55
125
128
183
252
240
280
406
163
160
275
419
543
3 -11
k. . , cm molecule s
bi
(7.50 ± 0.38) x 10"12
(7.06 ± 0.26) x 10"12
(7.95 ± 0.28) x 10"12
The pressure of H^O, the photolyte, was 200 m Torr.
39
-------
TABLE II. Rate Data for the Reaction of OH with
n-propylbenzene at 298 K
[Reactant]
m Torr
0
0.25
0.50
0.75
1.00
1.25
0
0.25
0.25
0.50
0.50
0.75
0.75
1.00
1.25
1.25
1.50
[Diluent]
Torr
20 (He)
20
20
20
20
20
200 (He)
200
200
200
200
200
200
200
200
200
200
Flash
Energy ,0
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
Flashes/
Filling
30
60
60
60
60
60
30
60
60
60
60
60
60
60
60
60
60
kV1
105
157
209
256
292
375
30
85
102
135
132
175
195
243
270
275
320
k.. , cm molecule" s"
(6.40 ± 0.36) x 10"12
(5.86 ± 0.16) x 10"12
of HpO, the photolyte, was 200 m Torr.
The pressure
40
-------
TABLE in. Rate Data for the Reaction of OH with
isopropylbenzene at 298 K
[Reactant]
m Torr
0
0.25
0.50
0.50
0.75
1.00
1.00
1.25
1.50
[Diluent]
Torr
200 (He)
200
200
200
200
200
200
200
200
Flash
Energy ,J
88
88
88
88
88
88
88
88
88
Flashes/
Filling
30
60
30
100
60
60
60
60
60
k',3-1
34
99
164
170
206
275
305
374
396
3 11
k. . , cm molecule s
(7.79 ± 0.40) x 10"12
of HJD, the photolyte, was 200 m Torr.
The pressure
41
-------
TABLE IV. Rate Data for the Reaction of OH with
n-propyl pentafluorobenzene at 298 K
;Reactant]
m Torr
0
0.5
1.0
1.5
0
0.5
1.0
1.0
1.0
1.5
2.0
2.5
3.0
0
0.5
1.0
1.5
1.5
1.5
2.0
2.5
[Diluent]
Torr
3 (He)
3
3
3
20 (He)
20
20
20
20
20
20
20
20
200 (He)
200
200
200
200
200
200
200
Flash
Energy ,J
88
88
88
88
88
88
88
40
245
88
88
88
88
88
88
88
88
40
245
88
88
Flashes/
Filling
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
k'.s'1
208
255
316
323
80
139
180
185
181
246
315
330
357
45
85
160
184
170
195
248
290
3 11
k. . , cm molecule s
(2.52 ± 0.54) x 10"12
(3.01 ± 0.76) x 10"12
(3.06 ± 0.24) x 10"12
The pressure of H,,0, the photolyte, was 200 m Torr.
42
-------
TABLE V. Rate Data for the Reaction of OH with
Hexafluorobenzene at 298 K
[Reactant]
m Torr
0
5
10
15
15
20
25
30
40
15
0
5
10
10
15
20
30
10
[Diluent]
Torr
20 (He)
20
20
20
20
20
20
20
20
20 (He)
200 (He)
200
200
200
200
200
200
200 (He)
Flash
Energy ,J
88
88
88
88
40
88
88
88
88
245
88
88
88
40
88
88
88
245
Flashes/
Filling
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
k'.s-1
70
109
164
179
184
211
248
319
346
230
34
75
117
108
152
196
240
135
3 1-1
k..p cm molecule s
(2.21 ± 0.20) x 10"13
(2.19 ± 0.16) x 10"13
The pressure of H20, the photolyte, was 200 m Torr.
43
-------
TABLE VI. Rate Data for the Reaction of OH with
o-Xylene at 298 K
[Reactant]
m Torr
0
0.5
1.0
1.5
0
0.5
0.5
0.5
0.5
1.0
0
0.5
1.0
[Diluent]
Torr
20 (He)
20
20
20
20 (Ar)
20
20
20
20
20
200 (He)
200
200
Flash
Energy ,J
88
88
88
88
88
88
88
88
30
88
88
88
88
Flashes/
Fi 1 1 i ng
30
30
30
30
30
30
100
200
30
30
30
30
30
K-.S-1
48
266
476
680
33
250
250
256
242
451
48
255
455
k. . , cm molecule s
(1.29 ± 0.01) x 10"11
(1.30 ± 0.03) x 10"11
(1.24 ± 0.01) x 10"11
The pressure of HpO, the photolyte, was 200 m Torr.
44
-------
TABLE VII. Rate Data for the Reaction of OH with
m-Xylene at 298 K
[Reactant]
m Torr
0
0.333
0.667
1.0
1.5
0
0.5
0.75
1.0
1.5
0
0.5
1.0
1.5
0
0.25
0.50
0.75
1.0
1.5
0
0.50
0.75
1.0
1.5
[Diluent]
Torr
3 (Ar)
3
3
3
3
20 (Ar)
20
20
20
20
20 (He)
20
20
20
200 (Ar)
200
200
200
200
200
200 (He)
200
200
200
200
Flash
Energy ,J
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
88
Flashes/
Filling
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
k'.s-1
81
256
400
629
830
41
365
539
658
999
41
406
761
1095
20
194
350
479
742
1000
50
424
534
787
1048
k. . , cm molecule s
(1.56 ± 0.14) x 10"11
(1.94 ± 0.08) x 10"11
(2.14 ± 0.02) x 10"11
(2.03 ± 0.19) x 10"11
(2.06 ± 0.13) x 10"11
The pressure of H20, the photolyte, was 200 m Torr.
45
-------
TABLE VIII. Rate Data for the Reaction of OH with
p-Xylene at 298 K
[Reactant]
m Torr
0.188
0.25
0.50
0.75
0
0.25
0.50
0.50
0.50
0.75
1.00
1.25
0
0.25
0.50
0.75
1.00
1.25
1.50
[Diluent]
Torr
3 (Ar)
3
3
3
20 (He)
20
20
20
20
20
20
20
200 (He)
200
200
200
200
200
200
Flash
Energy, J
88
88
88
88
88
88
88
88
30
88
88
88
88
88
88
88
88
88
88
Flashes/
Filling
60
60
60
60
60
60
60
25
60
60
60
60
60
60
60
60
60
60
60
k',5-1
312
345
395
482
89
162
259
287
253
315
464
520
92
193
289
392
452
545
591
3 -1-1
k.. , cm molecule s
D1
(8.8 ± 1.2) x 10"12
(1.01 ± 0.10) x 10"11
(1.05 ± 0.06) x 10"11
The pressure of H^O, the photolyte, was 200 m Torr.
46
-------
TABLE IX. Rate Data for the Reaction of OH with
Propylene at 298 K
[Reactant]
m Torr
0
0.25
0.50
0.75
1.00
1.25
1.50
0
0.5
0.75
1.00
[Diluent]
Torr
20 (He)
20
20
20
20
20
20
200
200
200
200
Flash
Energy ,0
88
88
88
88
88
88
88
88
88
88
88
Flashes/
Filling
50
50
50
50
50
50
50
50
50
50
50
k'.s-1
55
274
446
622
926
1060
1316
33
412
619
910
3 -1-1
k. . , cm molecule s
v<&
" 7
(2.56 ± .12) x 10"11
(2.63 ± .12) x 10"11
The pressure of H^O, the photolyte, was 200 m Torr.
47
-------
TABLE X. Rate Data for the Reaction of OH with
1-Butene at 298 K
[Reactant]
m Torr
0
.2
.4
.6
0
.2
.4
.4
.4
.4
.4
.4
.6
.8
[Diluent]
Torr
3
3
3
3
20
20
20
20
20
20
20
20
20
20
Flash
Energy ,J
88
88
88
88
88
88
88
45
180
281
88
88
45
88
k'.s-1
210
380
590
790
77
280
447
465
480
410
430*
405**
640
840
3 1-1
k. - , cm molecule s
(2.96 ± .19) x 10"11
(2.94 ± .14) x 10"11
**
20 flashes/filling
200 flashes/filling
The pressure of H90, the photolyte, was 200 m Torr the number of flashes/
filling was 50. 2
48
-------
TABLE XI. Rate Data for the Reaction of OH with
(U4-2-Butene at 298 K
[Reactant]
m Torr
0
.2
0
.1
.2
.2
.2
.4
.4
.6
IDiluent]
Torr
3
3
20
20
20
20
20
20
20
20
Flash
Energy ,J
88
88
88
88
88
45
180
88
180
88
k'.s-1
210
490
68
190
350
360
380
620
620
890
3 -1 -1
k. . , cm molecule s
(4.32 ± .41) x 10"11
(4.26 ± .25) x 10"11
The pressure of H^O, the photolyte, was 200 m Torr, the number of flashes/
filling was 20.
49
-------
TABLE XII. Rate Data for the Reaction of OH with
Tetramethylethylene at 298 K
[Reactant]
m Torr
0
.1
.2
.3
[Diluent]
Torr
20
20
20
20
Flash
Energy, J
88
88
88
88
k'.s-1
68
250
430
620
3 -1 -1
k. . , cm molecule s
(5.69 ± .13) x 10"11
The pressure of H?0, the photolyte, was 200 m Torr, the number of flashes/
filling was 20.
50
-------
< 4 x 10~16 cm2). A similar type calculation shows that the products of the
OH-hydrocarbon reactions are also not of any significance.
Variation of the number of flashes on a given gas mixture was carried out
to determine if: (a) final stable reaction products could be influencing the
measured k1 values; and (b) reactants were being depleted during multiple
flashing. The results (see Tables I, II, III, VI, and VIII) show that neither
of the above possibilities were significant. In the case of tetramethyl-
ethylene, however, no such variations were carried out. However, the number
of flashes per filling were minimal.
Finally, concerning other possible radical-radical reactions involving
OH, H, and 0, viz:
OH + OH >- H20 + 0
0 + OH » 02 + H
we have previously shown that these processes are unimportant under the con-
ditions employed in this study.
Substituted Aromatics
In Table XIII we have listed the rate constants measured in this work for
three xylenes, three monosubstituted benzenes, and two monosubstituted penta-
fluoro benzenes at different pressures. Also included in Table XIII are the
results of several other investigators which have appeared in the litera-
ture. As can be seen in this table, we have covered a wide range of diluent
gas pressures (3 Torr to 200 Torr) and have also used two different diluent
gases, argon and helium.
The rate constant for the OH-o-xylene system, kl5 was found to be pressure
independent between 20 and 200 Torr, indicating that kj was near or at its
51
-------
o
§
O
S-
o
fO
4->
(O
-P
A3
a:
(O
3
00
~~
,_,
1
t/1
QJ
13
O
CD
i
m
E
u
CsJ
0
X
CD
CO
O
O
o:
Q
3:
O
f
o
CD
^
LjJ
ce
CTi
CO
U-
«
r-
U-
LU
,1
UD
""""'
LL-.
LJJ
ce.
CO
ro
U-
C£
CM
o
1
o
UJ
O
O
13
4-
fO
X
QJ
z
03
^
- OJ
OJ 31
31
s- s-
0 r
o
O O
CM CM
CO r-H
O <*D
.
\D LD
OJ
C
CD
NJ
CD
CL.
O
CL
*C
r^
«D
QJ
S-
h-
O
CD
CM
CD
CD
.
f^1
CD
QJ
NI
CD
cx
o
fc.
cx
o
(/I
- QJ
'oT^ ^
31 S-
S- i-
S- 0 f
O H-
1 O
O O
ro CM CM
LO r-- CM
CM r-H y^
. .
CM ro m
o
s-
o
c
0-
r OJ
>> c
CX QJ
0 M
O- QJ
C
f - QJ
^ . QJ 31
QJ 3;
1C t.
S- S-.
S- 5- O
S~ 0 h-
O I
F 0
0 O
ro CM '
CO O> CM
O CD "
cn CO O
CM r-H <^J-
00 r-H
l£> O r-H
COLO l£)
CD
CD
13
o
h-
52
-------
high pressure limit. For the compounds m-xylene and p-xylene, lower total
pressures were employed (i.e., 3 Torr); in this case, the measured rate
constants at 3 Torr of argon were slightly less than those measured at 20 and
200 Torr. This observed change in k with pressure would definitely appear to
be outside the experimental error limits of the measurements. Thus, the
indications are that the high pressure limit for the OH-xylene addition re-
action is in the vicinity of 3 Torr. In the case of ethyl benzene, on the
other hand, the 3 Torr rate constant is,within experimental error»the same as
that at 20 and 200 Torr of helium. The same trend holds for the other four
compounds studied, namely, hexafluorobenzene, n-propyl benzene, isopropyl-
benzene and n-propyl pentafluorobenzene.
On the basis of the rate data presented, one would conclude that the re-
actions of OH with substituted aromatics studied in this investigation reach
the high pressure limit at or near 3 Torr. This observation of a reduced high
pressure limit for both the disubstituted and the higher molecular weight mono-
substituted aromatics relative to benzene and toluene is consistent with that
expected from the unimolecular theory of reaction rates which predicts a lower
value for the high pressure limit of a third order reaction when the number of
vibrational degrees of freedom of the transition state complex increases.
The rate data obtained by Perry, et al_. ,37 Hansen, ert a]_. >38 and Lloyd,
ejt a^L ,39 for g-xylene,m-xylene, and p-xylene agree, within experimental
error,with those obtained in this work. These first two studies were per-
formed using flash photolysis-resonance fluorescence technique, while the
third investigation utilized an environmental chamber. The rate constant
obtained by Morris and Niki40 for a mixture of the three xylenes is, within
experimental error, also the same as that obtained in this study. In the
53
-------
experiment by Morris and Niki a discharge flow-mass spectrometer system was
used. For the compounds ethyl benzene, n-propylbenzene and isopropylbenzene,
there is only one set of data reported in the literature. These results were
obtained by Lloyd, et_ al_.3,9 by measuring the rate of disappearance of these
aromatics relative to that of n-butane in an environmental chamber.
In our study, we have investigated a series of substituted aromatic
hydrocarbons including those where the hydrogen atoms on the aromatic ring
have been altered. As can be seen from Table XIII, the replacement of all
the hydrogen atoms in the benzene ring with fluorine atoms reduces the re-
action rate of this aromatic with OH by a factor of 6. (The rate constant for
the OH-benzene reaction is 1.59 x 10~12 cm3 molecule"1 s"1 as compared to
0.2 x 10*12 cm3 molecule"1 s"1 for OH-hexafluorobenzene.) We attribute this
reduction in reactivity solely to the depletion of the electron density on
the aromatic ring which would affect the extent of electrophilic addition of
the OH radical.
As shown in our earlier study (see Section 2),a comparison of the reac-
tivity of hydroxyl radicals with benzene and toluene reveals that the addition
of a methyl group to the aromatic ring increases the rate constant of the OH-
aromatic reaction by a factor of M. Moreover, the rate constant for the
reaction of OH with toluene increases by a factor of 2 between 3 and 100 Torr.
Based on these observations, these authors concluded that the addition of OH
to the ring was the predominant reaction and the abstraction was less than 50%.
A further increase in the chain length of the substituent on the aromatic ring
does not change its reaction rate with hydroxyl radicals to a significant ex-
tent. This point is illustrated here in Table XIII, where it is seen that the
rate constant for the reactions of OH with ethyl benzene, n-propylbenzene, and
54
-------
isopropylbenzene are essentially the same. This invariance of the rate con-
stant as a function of the chain length of the substituent indicates that the
abstraction of a hydrogen atom from the side chain is not a significant
fraction of the total reaction of these aromatics with OH. This conclusion
is in agreement with the known electron donating characteristics of these
substituents to aromatic rings, as reflected in their similar Harriett constant
values!*1 To further test this hypothesis, we have measured the rate constant
for the reaction of hydroxyl radical with n-propyl pentafluorobenzene. If the
abstraction of a hydrogen atom from the side chain was a very significant
part of the reaction, the fluorination of the aromatic ring would not decrease
the rate constant for the reaction significantly. However, we found that n-
propyl pentafluorobenzene reacted with OH about half as fast as did n-propyl
benzene. This decrease in rate constant would put a maximum limit of 50% on
the abstraction route. (Hexafluorobenzene itself reacts with OH about 1/6 as
fast as does benzene.) It should be noted that n-propyl pentafluorobenzene reacts
as fast as 3 x 10"12; this could be partially explained by assuming that the
propyl side chain donates a significant amount of electron density back to
the aromatic ring.
The above proposition that addition is the main route for the reaction
of OH radicals with substituted benzenes is further substantiated when one
examines the relative reactivity of the three xylenes with OH. It is well
known that the electron donating ability of a substituent to an aromatic ring
is very sensitive to the position of the substitution. In the case of o-,
p-, and m-xylene the number of abstractable hydrogen atoms is the same and,
in all three cases, the C-H bond energies on the methyl group are equal. Yet
m-xylene was found to react almost 1.5 times faster than both o- and p- . i
55
-------
xylene.Hence, as noted in our ealier work,we must conclude that the addition
of OH to the ring takes place because of the electrophilic nature of the OH.
Our conclusions, presented above, are in good agreement with those of
Perry, et al37 They measured the temperature dependencies for these reactions
of OH with benzene,toluene, and o-,m-,and p-xylene. These authors concluded
that at lower temperatures, namely betwen 296K and 325K, the OH reaction
ocurred both via the addition and abstraction mode. These conclusions were
based on the signs of the activation energies in the two temperature regimes
(viz, 296-325K and 380-473K) and an interpretation of the non-exponential
behavior of OH consumption in the intermediate temperature range of 325 to
380K. Based on their rate data, Perry, et al. have estimated the branching
ratios for the two reaction pathways (k addition/k abstraction) at 298K to be
5, 4, 24, and 14 for toluene, o-xylene, m-xylene, and p-xylene, respectively.
These numbers are within the range estimated in our earlier study (Sec. 2).
The above authors attempted to further check their inference by studying
the OH-perdeuterotoluene reaction, where they observed significant kinetic
isotope effect only at 432K. Perry, et al., however, have not tried to cor-
relate the values of the rate constants for different aromatic hydrocarbons
at room temperatures with the substitution of the aromatic ring and the re-
sulting change in the electron density on the aromatic ring.
OH + Propylene * Products
Our measured value of the rate constant for the reaction of OH with
propylene is independent of pressure between 20 and 200 Torr of He, with an
average value of 2.6 x 10"n cm3 molecule"1 s^1 for kg. Table XIV summarizes
the results on the OH-propylene system obtained in the present work as well
56
-------
O)
O
I
03
-U
CM CM
CD
C
QJ
cx
e
Q_
-;
0
LO
T 1
*fr
t.
1_
O
t
o
LO
CO
+1
OO
LO
CO
CD re
1- S-
i- 0
O h-
h-
o
CO CM
t I 1 t
CTi CTt
CM CM
c:
QJ
CD
r-i
«J3
S-
a:
S-
0
t
o
LO
LO
+\
CO
LO
- . QJ
tu rr
re
s-
S- S-
s- o
O f
o
00 CM
-H LO
*& CM
OJ 1JD
00 CM
^f- ^;l-
c
QJ
zs
CD
1
^
°
ro
LO
QJ
re
s-
s-
o
h-
o
CM
CO
T-H
«JD
LO"
,
OJ
4-3
OJ
QJ
E
QJ
1
QJ
^-J
C
QJ
OJ
ZJ
VI
"3
t/1
!c
4-J
O
C
OJ
z:
O"
QJ
to
^
t/1 QJ
j"i
c
ts s-
t/i 2
c o
O f
o
QJ
QJ JZ
M -M
S- S-
0
LO U-
jr QJ
-Q
S- -F-
O to
4- c:
0
d.
-1 to
1 QJ
o s-
t 1
LO
X QJ
S-
(J
-o «-
OJ
i~ (O
ro -»->
i i
o
QJ
4-> M-
JZ
4-) tO
O QJ
O. JZ
CU 4->
i-
c:
C 0
OJ
CU
cx
CM S-
^ o
to
iff -O
QJ fC
u
QJ fO
S_ JZ
Q) 4->
4-
QJ "O
ce: c:
zs
cr o
*
57
-------
as those from six other studies. 3,5,114,40,42,43 Not reported in this table are
the values of 1.1 x 10"11 and 1.34 x 10'11 cm3 molecule"1 s"1 obtained by
Simonaitis and Heicklen,1^ and Gorse and Volman,8 respectively. Both these
investigations were competitive kinetics studies where kg was measured rela-
tive to the OH+CO reaction.lt is apparent that both these values are approxi-
mately a factor of 2 lower than the results obtained in this study. It is
possible that the reason for this discrepancy lies in a higher value for
the rate constant of the OH + CO reaction at higher pressures.
Our value of 2.6 x 10"11 for k9 agrees extremely well with that obtained
by Atkinson and Pitts1}2 who also utilized the technique of flash-photolysis
resonance fluorescence. However, Stuhlllf employed the same technique of flash-
photolysis resonance fluorescence, and obtained a value of 1.45 ± .22 x 10"11
cm3 molecule"1 s"1. We believe that this lower value could be due to the loss
of propylene to metallic walls of the reaction vessel. Previous investiga-
tions in our laboratory utilizing an aluminum reactor have clearly shown ab-
sorption of reactants on walls to be a serious problem.
The last four columns in Table XIV are results obtained in low pressure
flow systems. These low pressure results are lower than our value of kg by
factors of 1.5-5. The reaction of OH with propylene has been assumed to occur
through the abstraction of a hydrogen atom. We believe, however, that under
high pressure conditions, these reactions proceed mainly by addition of OH to
the double bond. Our results indicate that the addition reaction has reached
its high pressure limit even below 2OTorr.However,we cannot explain the large
discrepancy between the low pressure flow tube data and the high pressure
resonance fluorescence data solely on the basis of the addition reaction. It
is our contention that part of this disparity is due to complicating chemistry
58
-------
in the flow tube which would affect the estimated stoichiometric correction
factors as well as to the fact that the addition reaction is still in the
fallout region.
In the work by Slagle, et al \5 where these authors carried out the direct
identification of reaction channels in the reactions of hydroxyl radicals with
propylene, the results suggest that even in the case of this olefin the ab-
straction of an allylic hydrogen by the OH radical is predominant. This, of
course, would be incompatible with our conclusion. More recently these
authors'+e have suggested , and we concur, that the explanation for this con-
tradictory observation by Slagle, et al , can be found in a more detailed
examination of their experimental conditions. In the latter work, the
authors crossed a "molecular beam" of OH with a propylene beam and detected
the products of the reaction utilizing photoionization mass spectrometry.
The pressure in the reaction region was approximately 10~2 to 10~3 Torr. If
we were to assume a reasonable value of 105 cm s'1 for the velocity of the
molecular beam,it can be estimated that the time spent by the product in the
"high pressure region" ( <10~2 Torr ) is approximately a few micro-
seconds. Under these conditions, the product of the addition
reaction would have only a few microseconds during which it could
undergo collision with the diluent gas (MO collisions). Hence, it can be
concluded that unless the energy rich addition product of the OH-propylene
reaction has a lifetime of several hundred microseconds, and would not need
more than a few collisions for stabilization, the addition channel would
never be detected.Thus, even if the reaction channel for abstraction of an
allylic hydrogen by OH was 100 times less than that for addition to thedouble-
bond, the experimental conditions of Slagel,et al were such that the abstraction
59
-------
would be the dominant observable reaction path. Finally, we would like to
point out that the results obtained by Cvetanavic, et_ aJ_^J in which low in-
tensity photolysis experiments were carried out at high pressures, indicate
that addition of OH to the double bond of propylene is the main route for
this reaction.
OH + 1-Butene. cu-2-Butene, and Tetramethylethylene
The results for these three reactions are also included in Table XIV.
k2
Atkinson and Pitts utilized the technique of flash photolysis-resonance
fluorescence as in the case of the present work. As noted earlier, Morris
40 43
and Niki and Pastrana and Carr utilized flow techniques. As can be seen
from Table XIV,the results obtained by Atkinson and Pitts for the reactions
of OH+1-Butene and OH+cXi-2-Butene agree with those obtained by us.
However, the rate constant for the reaction of OH + 1-Butene
obtained by Pastrana and Carr is approximately a factor of
two lower than that obtained in this work. We believe that the
explanation for this discrepancy lies, partially, in the lower pressures
utilized by Pastrana and Carr. As discussed above for the OH-propylene
system, we believe that at pressures of 1 Torr the OH-olefin reaction has
not reached the high pressure limit. As in the case of OH + 1-Butene, the
rate constants for the OH-cxi-2-Butene and OH-tetramethylethylene reactions
obtained by Morris and Niki are very much higher than those obtained by us.
This discrepancy would be further enhanced if we were to utilize our value
for k relative to which k , k and k12 were measured. There is, however,
a possibility that a significant perturbation in this system due to secondary
reactions is responsible for the disagreement with our data.
60
-------
Atmospheric Implications
In the introduction, the reaction of the hydroxyl radical with hydro-
carbons was indicated to be the initial step in the oxidation of hydrocarbons
leading to the formation of ozone through a complex sequence of reactions in-
volving NOw species. The reactions which follow the initial step are believed
to be relatively fast. Hence, the effective time before which these hydro-
carbons can generate ozone is dictated by the lifetime of the hydrocarbon with
respect to its reaction with OH.* This would mean that the confinement of the
ozone problem to the hydrocarbon emission region would be governed by the rate
of chemical degradation relative to transport. The latter variable, the mass
transport, would depend on exact meteorological conditions at, and at least
one to two days after, the time of release. The chemical degradation rate
can be calculated using available rate constant data and atmospheric modelled
and/or measured OH concentrations.
In Table XV we have listed our calculated hydrocarbon lifetimes, under
varying conditions, for all the compounds studied during this program. In
this case, the destruction rate of a given hydrocarbon is given by
= k[OH]ave[RH] (13)
and Te(lifetime) = (k[OH]-Je( (14)
where [OH] is the average OH concentration which has been computed for
several different time periods. Thus, the two parameters required to
This statement is only strictly true for the case of saturated and aromatic
hydrocarbons. In the case of highly reactive olefinic hydrocarbons, reac-
tion with ozone is also an important degradation mechanism.
61
-------
calculate the lifetime are the average steady state OH concentration and the
bimolecular rate constant for a given OH-hydrocarbon reaction. The bimolecu-
lar rate constant, k, depends on the temperature while the steady state OH
concentration varies with the altitude, latitude, season, and the time of day.
However, since the value of k for most compounds examined here would vary
little over a nominal seasonal temperature change of 0-30 C, we have used a
300K value in all our calculations. OH concentrations were taken both from the
quasi-equilibrium modelling calculations of Chang, Wuebbles, and Davis,1*8 and
from the direct OH measurements reported by Davis et al^at low altitudes
(2.1 km). The latter measurements were carried out at approximately high noon
at 37° N latitude under summertime conditions (OH = 9 x 106 cm"1). Chang,et
al have calculated steady state OH concentrations as a function of latitude,
altitude, season, and time of day; they have also evaluated diurnal averaging
factors for different latitudes, altitudes, and seasons of the year.
As seen from Table XV, we have calculated the lifetimes of 14 hydro-
carbons under widely different conditions. For those lifetimes which would
be longer than the diurnal solar cycle (12 hours in summer and 8 hours in
winter), we have used diurnally averaged OH concentrations. From the calcu-
lated lifetimes, which vary from less than one hour to as long as a few
months, we can classify those hydrocarbons studied into three categories: (1)
*
those which have lifetimes always less than 3 to 5 hours during
summer and winter; (2) those which have lifetimes less than 3 to 5 hours in
summer but greater than 7 to 9 hours in winter; and (3) those which have life-
times greater than 7 to 9 hours during summer and winter.
The significance of such a classification is obvious-lifetimes oflessthan
3 to 5 hours would imply that these hydrocarbons would be degraded essential -
The number represents the hydrocarbon lifetime required to degrade ^60% of
a 12-hour day's emission (at constant rate) within the same solar day.
62
-------
i.
O)
o
CL
O
1
OJ
s_
to
o
o
T3
O
CO
X
UJ
_i
CO
QJ C
E O)
^ E
V- E E
- QJ -- -
*~ ^ OJ Ol
OJ M- M-
.C OJ 'r- -r-
I- -Q ' I
X X X X
o o o o
CM CM CO CvJ
--< C\J --H
c: c: c ci
OJ OJ QJ CJ
u u o u
o o o o
63
-------
ly the same day they are released, thereby making them a local or
regional problem depending on exact meteorological conditions. Those
hydrocarbons which have lifetimes greater than 7-9 hours would most likely
be transported away from the release point, making them a regional or con-
tinental problem.
The olefins-propylene, 1-butene, cis-2-butene, and tetramethylethylene-
fall into the first category. Especially during summertime conditions, these
olefins would be degraded within a few hours. The three xylenes, toluene,
ethyl benzene, the two propyl benzenes, and ethylene have lifetimes shorter
than the solar day during summer months; therefore, these compounds would
effect local areas. However, during winter months, they could be transported
long distances. The three xylenes could have either local or regional effects,
depending on exact meteorological conditions. Those compounds in the third
category, benzene and acetylene, would have very long lifetimes and would be
transported long distances.
The results from our calculations once again point out that control
strategies for hydrocarbons should be based on detailed calculations which
reflect the reactivity of a given hydrocarbon towards OH, the OH concentra-
tion level on a local and regional scale, and meteorological conditions. Thus ,
considerably different strategies could be expected for different sources
having different hydrocarbon emissions. Of course, if the 03 production is
the point of attention, the NOX levels in the local environment must be con-
sidered. Also, the long range transportation of hydrocarbons does not
necessarily lead to 03 production since NOw is essential for the ozone forma-
tion sequence to be effective.
64
-------
SECTION 5
IDENTIFICATION OF OH-HYDROCARBON REACTION PRODUCTS
OVERVIEW
During the last year of our three year effort involving studies of OH-
hydrocarbon reactions, two different laser systems were successfully inter-
faced with a T.O.F. Bendix mass spectrometer for purposes of investigating
the reaction products from OH-hydrocarbon processes. Two different modes of
operation for this system were investigated; although neither mode was
completely successful, the work did proceed sufficiently far as to make quite
clear the potential of this new methodology and the new elements needed to
fully realize this potential.
The two basic components in the Georgia Tech system are (1) a T.O.F. mass
spectrometer and (2) a high energy pulsed laser. The time-of-flight mass
spectrometer operates by separating, in real time, the ions formed in the ion
source into groups characterized by their mass to charge ratio. The ions
arrive at the magnetic electron multiplier detector in groups separated in
time. This process of ionization, separation, and detection is repeated every
100 ys. Hence, this machine has the potential of acquiring 10,000 mass spectra
in one second. As indicated above, there are two modes in which we operated
the T.O.F. during the present investigation: (a) spectrum acquisition
and (b) single ion monitoring. In the first mode of operation, the output
of the magnetic electron multiplier (MEM) was constantly gated into the
65
-------
scope anode--the output of this anode being the mass spectrum. The alterna-
tive mode of operation was that of diverting the output of the MEM correspond-
ing to one particular mass into the anode. This provided the abundance of a
single ion as a function of time with a time resolution of 100 ys.
The second major component in the Georgia Tech system, the photolysis
source, consisted of either a Nd-Yag laser or a tunable dye laser-
The fundamental output of the Nd-Yag System (ILS-NT-572) was quadrupled to
o
give 2650 A radiation. This wavelength was suitable to photolyze 03 to give
Q(1D) which, via the 0(1D)-H2 reaction, generated an instantaneous source of
OH radicals. Since the first laser system had a repetition rate of 10 pps,
the possibility of generating high steady state concentrations of OH reaction
products was explored. The second laser, a frequency doubled tunable dye
laser, had wavelength output in the spectral region of 2500-3500 A. Single
shot outputs from this system typically ranged from 25-100 mJ, thus making
this system suitable for examining the time history of a given reaction
product.
66
-------
EXPERIMENTAL APPROACH
The basic experimental setup is shown in Figure 1. In this system, a
quartz reaction cell which was equipped with a molecular leak let a constant
fraction of the cell contents effuse into the ionizing region of the T.O.F.
mass spectrometer. The mass spectrometer, therefore, acted as a detector
which could identify product constituents and measure their relative amounts
in the effusion mixture.
For the single ion monitoring mode, the electronic processing of the
signal was considerably different than for the steady state operation (mode
#1). In the former case, the output of the mass spec anode was monitored by
feeding an amplified signal into a multi-channel analyzer (MCA). The MCA was
then triggered to start the ion monitoring by a photodiode which was activated
by photolyzing laser pulse. Thus, the MCA was set up to record the intensity
of a given ion as a function of time. In an effort to further improve on the
signal to noise ratio of this system, data from several laser shots was aver-
aged on the MCA.
67
-------
I/O
_I_
CD
1 CD
"O -
TRIGGER
-^k
-0
i
"
LU
<.
<_
I
r
~L
c
r
L
\
X
c
<
!
i
f^
CL
r
j-
r
o:
c:
x
"L
s
y
/
1 1
t
a;
=c
j_
/o
s:
n _
T-
I
L
_t
j-
V
^w
"i <"">
r
Q±
>
|
-4
^f,t,
S K
«E^_ rj
1
1
1
I
1
|
I
(
| »
i
4
1
1 i 1
A-i'r^
1 Hy3
i " ^mi
^ '
i n1"1
i
j
*
*.
e
c CD
O 4-^
r- CO
O» (O c/1
_Q 4_J ,
3 oo s- i- s-
1 OJ O) C (O O) S- O)
_c T- _i c co O E
rO-i-ECn-r-3: lOJ i-
QJi 3(O4->O OL +J
^ U_ Q_ >- fl) O "O GJ d) 1 O
i-O(OC-oa;>>s- a>
C£3 S-OOEOOI-r-O-MQQJ Q.
CO roa)O)3-i-Qi3C:rO-O t/1
r-o-Q-T--a =ccrsa>s-
U >iO4-)4JCLI(l)JDO CO
, _ . ._ (U CO CO "O -4-* 5- S- Cl Q- ro -r- rO
i COCOOOrOfOOOCcO S
CD IE: 2: 2: z a. crcroo oo i >
< u.
i i i i i i i i i i i
o
j | ^/) tjj (T^> t_3 |^g ^t CD \ c£ *
^^ \,A fi c^ r\_ (v* cy ty) t/^ c^ ^?* i
oo co >- cr i i
s: 2: i i
T> co
z: !-
CO
O
O
O-
... S-
O)
CO
CO ro
> O- _l
oo
s: ro r C
, -M ro
en <_3 > O
^ Q- CL C 4J i C -i-
O E -t- to ro O -*->
3 i >> c: i- ro
1 ex LO JD co <^ 4-> E
i -i- o C7i i a> jc
ro S- Q C OJ i O
OJ -f- CJ3 t- C UJ OO
4-> 4-> >,^- C
rO CEUT3T3CCO eC
CD QJ O C C -r- J^ -r-
QjS-S-O)
QJ QJ i QJ 4-> 3 T" c£5 ! <1J
-O-OM-M-OCT 4->c -<
OOM-M-QJQJcOCi CO 1
CCrO-i-r S-(OO3(O LT)
CD
1 1 1 t 1 1 1 t 1 1 S~
Q
-------
EXPERIMENTAL RESULTS
As of this final report, the T.O.F. mass spec-laser photolysis system has
been only marginally successful as an operational data collecting system. Even
so, we believe that this aspect of the project has been successful in that the
operational parameters required to make the system highly successful have now
been clearly defined.
The principal chemical system which was examined with the present experi-
mental arrangement was the reaction system ,OH-Toluene. In these experiments,
OH radicals were produced by the UV photolysis of 03 in the presence of H2.
The 0(1D) product resulting from the photolysis of 03 reacted with H2 accord-
ing to the scheme:
0(!D) + H2 » OH + H
H + 03 » OH + 02
Tests carried out with ozone-toluene mixtures showed no evidence of a measur-
able chemical reaction within the time duration of our experiments (i.e., a
few minutes). Experimental conditions were also adjusted such that the dom-
inant reaction of OH was with toluene rather than H2 or 03 (both are slow
processes). To date, we have seen H20 as a reaction product from the OH-
toluene system; but we have not been able to get the sensitivity of the mass
spec-laser system high enough to carry out quantitative experiments. The
original idea was that of trying to quantitatively assess the fraction of
the total OH-toluene reaction process which was abstraction and that which
69
-------
was addition to the ring. This was to be carried out by comparing the amount
of H20 formed (a product of abstraction only) in the OH-toluene system with
that in a second system where only abstraction could occur with H20 forma-
tion (e.g., a reactive saturated hydrocarbon). As stated above, however, a
lack of sensitivity precluded such experiments being carried out. In like
manner, the low sensitivity of the T.O.F. mass spec also made it impossible
to identify any ring addition products produced from the OH-toluene reaction.
In the case of the high rep rate experiments where steady state concen-
trations of products were to be examined, the low ionization efficiency of
the T.O.F.ion source coupled with an oversized molecular leak inthernass spec
reaction cell resulted in less than optimum results. In addition, the use of
ozone as a source of OH produced further complications since reactive frag-
ments formed from the initial reaction step of OH with toluene could very
Ukely react with 03 .We believe the solution to the above set of problems is
3-fold: (l)the replacement of the original reaction cell with one containing a
smaller molecular leak (permitting a higher build-up product concentration
in the reaction cell); (2) the elimination of 03 as a source of OH; and (3)
the generation of high concentrations of OH per unit time from the laser.
With regard to the single ion monitoring mode, where in some cases the
time history of a species may be examined, the difficulty here
was again the inherent inefficiency of the T.O.F. ion source plus an addition-
al problem involving the integration circuits on the output of the MEM. In
the present configuration, the T.O.F. electrometer has a time constant
sufficiently long that it is signal averaging even when the reaction products
from the laser pulse have long been removed from the ionizing region of the
mass spec. In effect, this has caused a factor of 20 to 30 loss in sensitivity.
70
-------
To resolve the two general problems outlined above; namely, low sensi-
tivity and complex secondary chemistry due to the use of 03 as a source of OH,
we have come up with the following recommendations: (1) switch from 03-H2
mixtures to H20 as a source of OH; (2) increase the OH concentration per laser
shot by utilizing an ArCl Eximer laser (^1700 A); (3) modify the sampling
electronics to permit monitoring of only those mass spec cycles which are
synchronized with the laser firing and (4) in the case of the steady state
monitoring experiments, employ a reaction cell with a reduced effusion rate.
The switch from 03-H2 mixtures to H20 would prevent the primary reaction
products from OH-aromatic reactions from reacting with 03 to yield non-
representative stable secondary reaction products. It also would make possible
the study of OH-olefin reaction products since olefin-03 mixtures are far too
reactive to permit any assessment of the reaction products from OH reactions.
A crucial new element required to take advantage of H20 as an OH source in our
studies is the ArCl Eximer laser (point (2) given above). With an output
0
energy of ^25 mJ and a wavelength of ^1700 A, this vacuum UV laser source
would not only permit the use of H20 as a reliable source of OH, but its high
energy output coupled with the high cross section for absorption by H20 at
1700 A would also result in an increase in the OH concentration per laser
pulse of at least one order of magnitude. The addition of new sampling elec-
tronics (i.e., a Biomation 8100 transient recorder) would also improve the
sensitivity of the experiment by another order of magnitude by ensuring that
only those cycles of the T.O.F. mass spec were processed which correlated with
reaction products actually being in the ion source of the mass spectrometer.
With a Biomation transient recorder the following sequence of events would
occur:
71
-------
1. The laser would fire and this photolysis flash would initiate the
OH-hydrocarbon reaction.
2. The reaction products along with the reactants and the diluent gas
would effuse into the path of the ionizing electron beam.
3. The Biomation transient recorder would be used to capture (in 2048
memory locations) the scope anode output wave form generated by a
single cycle of the mass spectrometer when the concentration of the
products is high.
4. The captured wave form (the mass spectrum) would be transferred to
a Northern Scientific signal averager within 25 milliseconds after
the termination of the sweep.
5. Within the next five milliseconds the Biomation transient recorder
would be cleared and "armed" to accept a new wave form.
6. The next laser flash would trigger the same sequence of events except
that the new wave form obtained would be added to the previous wave
form that is stored in the Northern Scientific signal averager.
Through this sequence of events, we could signal average as many as 1200
times/second.
The necessary requirements for obtaining a pertinent wave form would be
that the laser has fired, the reaction products have effused into the ion
source, and the mass spectrometer has just started a new cycle. Hence, the
triggering scheme must involve a trigger pulse going to the Biomation trans-
ient recorder only after the above events have taken place. For our experi-
ments, we would utilize a triggering scheme as shown below:
72
-------
Mass
Spectrometer
Pulse
Constant
Delay
Laser
X TVinnfir
Constant
Voltage
Generator
t
Variable
Delay
t
Photo
Diode
AND \ Pulse
GATE! to the
"* J biomation
The master pulse from the mass spectrometer clock would be suitably de-
layed (maybe as long as 90 microseconds) and fed into the laser. The laser
would fire within a few microseconds after being triggered. The photodiode
would detect the laser flash and send a signal to the constant voltage gener-
ator box through a variable delay. Once the pulse reaches this box it would
set a flag in the And-Gate, thereby partially arming the gate. As soon as
the mass spectrometer starts a new cycle, the master pulse would feed a signal
into the And-Gate thereby enabling the gate to produce the trigger for the
Biomation recorder. By changing the delay between the emergence of the
photodiode pulse and the time at which the constant generator sets up a flag
in the gate, a mass spectrum characteristic of the reaction products at
different reaction times could be obtained. By suitably gating the Biomation
73
-------
output, we could also monitor the time profile of a single ion intensity.
This kind of study could yield kinetic information.
74
-------
LIST OF REFERENCES
1. N. R. Greiner, J. Chem. Phys., 46_, 2795 (1967).
2. N. R. Greiner, J. Chem. Phys., 53_, 1070 (1970).
3. E. D. Morris, Jr., D. H. Stedman, and H. Niki, J. Amer. Chem. Soc., 93,
3570 (1971).
4. N. R. Greiner, J. Chem. Phys., 53., 1284 (1970).
5. J. N. Bradley, W. Hack, K. Hoyerman, and H. Gg. Wagner, J. Chem. Soc.,
Faraday Trans. 1, 69 (11), 1889 (1973).
6. I.W.M. Smith and R. Zellner, J. Chem. Soc., Faraday Trans. 1, 69,
1617 (1973).
7. N. R. Greiner, J. Chem. Phys., 46_, 3389 (1967).
8. R. A. Gorse and D. H. Volman, J. Photochem., 3^ 115 (1974).
9. J. H. Seinfeld, T. A. Hecht, and P. M. Roth, EPA Report No. EPA-R4-73-031,
Research Triangle Park.
10. D. D. Davis, R. Klemm, and M. Pilling, Int. J. Chem. Kinet., IV, 367
(1967).
11. D. D. Davis, R. Schiff, and S. Fischer, J. Chem. Phys., 6U 2213 (1974).
12. W. A. Lonneman, S. L. Kopczynski, P. E. Darley, and F. D. Sutterfield,
Environ. Sci. Techno!. {J, 229 (1974).
13. W. A. Glasson and C. S. Tuesday, Environ. Sci. Techno!. 4., 916 (1970).
14. F. Stuhol, Ber. Bunsen ges. Phys. Chem. 9_, 674 (1973).
15. I.W.M. Smith and R. Zellner, J. Chem. Soc. Faraday Trans. II. 69,
1617 (1973). ~
16. W. E. Wilson and A. A. Westenberg, llth Symp. on Combustion, Berkeley,
CA, 914 (1966).
17. J. E. Breen and G. P. Glass, Int. J. Chem. Kinet., 3., 145 (1970).
18. A. Pastrana and R. C. Carr, Int. J. Chem. Kinet., 6, 587 (1974).
75
-------
19. M. Gehring, K. Hoyermarm, H. Gg. Wagner, and J. Wolfrum, Z. Naturforsch.
A 2_5, 675 (1970).
20. J. R. Kanofsky, D. Lucas, F. Fruss, and D. Gutman, J. Phys. Chem., 78,
311 (1974).
21. (a) D. D. Davis, R. Huie, J. Herron, W. Braun, and M. Kurylo, J. Chem.
Phys., 56, 4868 (1972).
(b) D. D. Davis and R. B. Klemm, Int. J. Chem. Kinet., £, 367 (1972).
22. M. C. Sauer and L. M. Dorfman, J. Chem. Phys., _35_, 437 (1961).
23. Adv. Photochem., 3., 228 (1964).
24. K. Watanabe and M. Zelikoff, J. Opt. Soc. Am., 43_, 753 (1953).
25. K. Watanabe, J. Chem. Phys., 4C>, 558 (1964).
26. Adv. Photochem., 3., 226 (1964).
27. K. Hoyermann, H. Gg. Wagner, and J. Wolfrum, Ber. Bunsen ges. Phys. Chem..
72, 1004 (1968).
28. J. V. Michael and H. Niki, J. Chem. Phys., 46_, 4969 (1967).
29. G. G. Volpi and F. Zocchi, 0. Chem. Phys., 44, 4010 (1966).
30. J. V. Michael and R. E. Weston, J. Chem. Phys., 45, 3632 (1966).
31. M. Zelikoff and L. M. Aschenbrand, J. Chem. Phys., 24_, 1034 (1956).
32. N. R. Greiner, J. Chem. Phys., 51,, 5049 (1969).
33. (a) P. J. Crutzen, Report AP-10, Institute of Meteorology, Univ. of
Stockholm, June 1972; Tell us XXVI, 47 (1974); Proceedings of the Fourth
CIAP Conference, Cambridge, Mass., Feb (1975).
(b) H. Levy, Planet. Space Sci.,20, 919 (1972); ibid., 21,, 575 (1973);
J. Geophy. Res., 78, 5325 (197377
(c) S. C. Wofsy, Annual Review of Earth & Planetary Sciences, £ (1976).
34. (a) E. Robinson & R. A. Rasmussen, J. Geophy. Res., 78, 5345 (1973).
(b) E. Robinson & R. C. Robbins, "The Changing Global Environment,"
p. Ill, D. Reidel Publishing Co. (1975).
(c) A. A. Altschuller, W. A. Lonneman, F. D. Sutterfield, & S. L.
Kopczynski, Env. Sci. & Tech., 5_, 1009 (1971).
(d) R. J. Gordon, H. Mayrsohn, & R. M. Ingels, Env. Sci. & Tech., 2^,
1117, (1968).
35. R. P. Baker, R. R. Baldwin, & R. W. Walker, Symp. Combust., 13th, Univ.
Of Utah, Salt Lake City, 1970 (1971).
76
-------
36. R. F. Hampson & D. Garvin, "Chemical Kinetic & Photochemical Data for
Modelling Atmospheric Chemistry," NBS Technical Note 866 (1975).
37. R. A. Perry, R. Atkinson & J. W. Pitts, Jr., J. Phys. Chem., SU, 296
(1977).
38. D. A. Hansen, R. Atkinson & J. N. Pitts, Jr., J. Phys. Chem., 79, 1763
(1975).
39. A. C. Lloyd, K. R. Darnall, A. M. Winer, & J. N. Pitts, Jr., J. Phys.
Chem., 80, 789 (1976).
40. E. D. Morris, Jr. & H. Niki, J. Phys. Chem., 75, 3640 (1971).
41. K. M. Wiberg, "Physical Organic Chemistry," Wiley (1964).
42. R. Atkinson & J. N. Pitts, Jr., J. Chem. Phys., 63_,3591 (1975).
43. A. V. Pastrana & R. C. Carr, Jr., J. Phys. Chem., 79_, 765 (1975).
44. R. Simonaitis & J. Heicklen, Int. J. Chem. Kinet., 5., 231 (1973).
45. I. R. Slagle, J. R. Gilbert, R. E. Grahma, & D. Gutman, Symposium on
Chemical Kinetics Data for the Lower and Upper Atmosphere, Warrenton,
Virginia, Sept (1974).
46. D. Gutman, Discussions at 12th International Symposium on Free Radicals,
Laguna Beach, CA, Jan (1976).
47. R. J. Cvetanovic, Discussions at 12th International Symposium on Free
Radicals, Laguna Beach, CA, Jan (1976).
48. J. Chang, D. J. Wuebles and D. D. Davis, J. Geophys. Res., (in press).
49. D. Davis, T. McGee, W. Heaps, D. Philen & A. Nelson, Electric Power
Research Institute Report, (in press).
50. S. Fischer, R. Schiff, E. Machado, W. Bellinger, & D. Davis, 169th
National American Chemical Society Meeting, Phil., PA, April 6-11 (1975)
51. G. J. Doyle, A. C. Lloyd, K. R. Darnall, A. M. Winer, & J. N. Pitts, Jr.,
Environ. Sci. Techno!., 9, 237 (1975).
77
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO.
FPA- 6-00/5- 77-11-1-
3. RECIPIENT'S ACCESSI ON-NO.
4, TITLE AND SUBTITLE
5. REPORT DATE
October 1977
INVESTIGATION OF IMPORTANT HYDROXYL RADICAL REACTIONS
IN THE PERTURBED TROPOSPHERE
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
D. D. Davis
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Engineering Experiment Stations
Georgia Institute of Technology
Atlanta, Georgia 30332
10. PROGRAM ELEMENT NO.
1AA6Q3 AF-n5
11. CONTRACT/GRANT NO.
R804629
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 5/74 - 7/77
14. SPONSORING AGENCY CODE
EPA/600/07
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The flash-photolysis resonance fluorescence technique has been utilized to study
the reaction kinetics of hydroxyl radicals with ten aromatic and six olefinic hydro-
carbons at 298 K and several diluent gas pressures. The aromatic compounds that were
studied include benzene, toluene,ethyl benzene, n-propylbenzene, isopropylbenzene,
hexafluorobenzene, n-propyl pentafluorobenzene, o-, m-, and p-xylenes; and the olefins
include ethylene, acetylene, propylene, 1-butene, cis-2-butene, and tetramethylethy-
lene. Based on our extensive data on OH-substituted aromatic hydrocarbon reactions,
it has been inferred that addition of hydroxyl radicals to the aromatic ring is the
dominant reaction in these systems. In the case of OH-olefin reactions, addition of 01
to the double bond seems to be a prominent path for the heavier unsaturates. From
these rate constant data the lifetimes of all these hydrocarbons in the lower
troposphere has been calculated. Utilizing the technique of laser flash photolysis,
time-of-flight mass spectrometry, attempts were made to understand the mechanisms
involved in the reactions of OH with substituted aromatic hydrocarbons.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
*Air pollution
*Photochemical reactions
*Reactions kinetics
*0xidation
*Aromatic hydrocarbons
*Alkene hydrocarbons
Hydroxyl radicals
13B
07E
07D
07B
07C
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report}
21. NO. OF PAGES
'(This page)
22. PRICE
UNCLASSIFIED
EPA Form 2220-1 (9-73)
78
------- |