EPA-600/3-77-014a
February 1977
Ecological Research Series
MECHANISMS OF PHOTOCHEMICAL REACTIONS
IN URBAN AIR
Volume I. Chemistry Studies
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are: .
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconornic Environmental Studies
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal
species, and materials. Problems are assessed for their long- and short-term
influences. Investigations include formation, transport, and pathway studies to
determine 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-600/3-77-014a
February 1977
MECHANISMS OF PHOTOCHEMICAL REACTIONS IN URBAN AIR
Volume I. Chemistry Studies
by
James N. Pitts, Jr.
Statewide Air Pollution Research Center
and
Department of Chemistry
University of California
Riverside, CA 92521
Grant No. 800649 - 13, 14, 15
Project Officer
Marcia C. Dodge
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. 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.
ii
-------
ABSTRACT
Results are presented of a research program concerned with selected
aspects of the kinetics, mechanisms and products of reactions involved in
photochemical air pollution.
Rate constants were determined, using competitive and modulation-phase
3
shift techniques, for the gas phase reaction of 0( P), atoms with a variety
of organic and inorganic species over the temperature range 296-423 K.
3
Products for the gas phase reaction of 0( P) atoms with toluene and 1-raethyl-
cyclohexene were also studied.
The products and mechanisms of the reaction of nitric oxide with
methyl peroxy radicals were investigated at 296 K using long path infrared
spectroscopic and gas chromatographic techniques.
The reactions of peroxyacetyl nitrate were investigated in the gas
phase with selected constituents of polluted atmospheres, and in the liquid
phase with a variety of organics. Chemiluminescence from the reaction of
peroxyacetyl nitrate with a series of amines was studied in the liquid phase,
The mechanism and products of gas phase reactions of ozone with a variety
of organics was investigated in low pressure flow systems using chemi-
luminescent and photoionization mass spectrometric techniques.
The N00-catalyzed geometric isomerization of 2-butenes and 2-pentenes
was studied over the temperature range 298-400 K while an investigation of
the NO -propylene photooxidation system was carried out at room temperature.
X
iii
-------
CONTENTS
Abstract ill
List of Figures vi
Lis t of Tables ix
Acknowledgments xi
I. Introduction 1
II. Conclusions 4
III. Recommendations 8
IV. Experimental Methods, Results and Discussion 9
1. Determination of absolute rate constants for the
reaction of 0(3P) atoms with organics over the
temperature range 299-392°K 10
2. Rate constants for the reaction of 0(3P) atoms with
selected olefins, monoterpenes and unsaturated
aldehydes 23
3. The reaction of 0(3P) atoms with toluene and 1-methyl-
cyclohexene 36
4. A long-path infrared spectroscopic study of the
reaction of methylperoxy free radicals with nitric
oxide 49
5. Rate constants for the gas-phase .reaction of peroxy-
acetyl nitrate with selected atmospheric constituents 62
6. Chemiluminescence detection of PAN and PBzN ...... 74
7. Solution phase reactions of PAN with molecules of
biological significance 82
8. Thermal and photochemical processes of NO and N©2 and
the application of the results to the chemistry of
polluted atmospheres 86
9. 40-meter long-path infrared studies of the propylene/
N0x/hv system 97
10. The ozone-induced chemiluminescent oxidation of
acetaldehyde , 109
11. Low-pressure gas-phase ozone-olefin reactions. Chemi-
luminescence, kinetics, and mechanisms 114
12. Photooxidation mass spectrometer studies of gas-phase
ozone-olefin reactions 150
-------
LIST OF FIGURES
Number
1.1 Plots of 2trv/tan 4> against xylene concentration for o-, m-,
and p-xylenes at room temperature .................... ....... 13
1.2 Arrhenius plots fpr the arpmatic hydrocarbons .......... ..... 15
1.3 Arrhenius plots of log k against 1000/T ... ..... .... ......... 16
2.1 Plots of ethylene yi^ld versus N2 yield for cyclppentene
alone and for fours runs with differing limonene/cyclopentene
ratios .... .................................................. 26
A rp
2.2 Arrhenius plots of log k2*Vk2 - against 1000/T .............. 27
2.3 Arrhenius plots of log k2A/k2C? against 1000/T .............. 28
3.1 Plot of selected product yields (N2 = 1) against total N20
pressure for the reaction of 0(3P) atoms with 1-methylcyclo-
hexene at 296 ± 2°K " . .'.'...I.... ...... V ..... ". ---- . ............ 44
'
4.1 Infrared spectra of C^aPNO, NO, N02; Infrared spectra of
azpmethane , NO, and 02 before and after photolysis .......... 53
4.2 Typical time dependence of CHsONO, CHsON02, N02 and NO in the
photooxidation of azomethane in the presence of NO .......... 55
5.1 Plots of d ln[PAN]/dt against reactant concentration for 03
and H20 ...... V. ........................ ..................... 65
5.2 Plpfs of ln[PAN] against time for the reaction of PAN with NO
in air diluent .......... . ................ . .................. 67
6.1 Chemiluminescent detection unit ............................. 75
6.2 Emission spectra of chemilumine scene e reactions of triethyl-
amine with PAN and ozone ........ . ........................... 78
6.3 Decay curves of chemiluminescence generated in reaction of
triethylamine with PAN and ozone ............. ....... ........ 79
8.1 Plot of the mole fraction of trans-2-butene vs time for the
cis- to trans-NO? catalyzed geometric isomerization of
cis-2-butene .......... . ........ .... ......................... 88
vi
-------
LIST OF FIGURES (cont.)
Number Page
8.2 Arrhenius parameters for the N02 catalyzed geometric isomeri-
zation of cis- or trans- 2-butene ....................... ..... 90
8.3 Arrhenius parameters for the N02 catalyzed geometric, isomeri-
zation of cis- or trans- 2-pentene ..................... ...... 91
9.1 Typical concentration-time profile for the N02 photooxidation
of C3H6 .............. ; ................................. ..... 100
9.2 Typical concentration-time profile for the N02 photooxidation
of C3H6 ..... . . ; ...... ; ..................................... -. 101
9.3 A typical GO-MS total ion chromatogram of an irradiated N02-
C3H6-air mixture ....................... ..................... 104
9.4 GC-MS computer spectrum of CHsCHO as obtained from the total
ion chromatogram of an irradiated N02-C3Hg-air mixture ...... 105
915 GC-MS computer spectrum of (013) 2CO as obtained from the
total ion chromatogram of an irradiated N02-C3H6-air mixture 106
9.6 GC-MS computer spectrum of CHsOH as obtained from the total
ion chromatogram of an irradiated N02-C3Hg-air mixture ...... 107
10.1 Emission spectra from 500 to 800 nm in the chemilumine scent
reactions of ozone with acetaldehyde and H atoms ............ 110
10.2 Emission spectra from 700 to 1100 nm from the chemilumines-
cent reactions of ozone with acetaldehyde and H atoms ....... Ill
11.1 Schematic of apparatus used in chemiluminescence and kinetic
studies [[[ 116
11.2 Chemilumine scent emission spectra in the visible region from
the reaction of 2% 03/02 with ethylene, trans- 2-butene, and
isobutene [[[ 120
11.3 Comparison of the chemiluminescent emission (700-1100 nm)
from the ozone-cis-2-butene reaction to the OH Meinel bands
from the H + 03 reaction .................................... 122
11.4 Chemiluminescent emission spectra (300-550 nm) from the
reactions of 2% 03 in 02 or N2 with ethylene, cis-2-butene,
-------
LIST OF FIGURES (cont.)
Number Page
11.5 Chemiluminescent emission spectra (300-600 ran) from the
reactions of 2% 03 in N2 or 02 with 2-butene-ds, and
isobutene-dg 125
11.6 Semilogarithmic plots of relative emission intensities and
ozone absorbance vs reaction distance for the reaction of
cJ.s-2-butene With 2% 03/02 or 2% 03/N2 128
12.1 Resonance lamp energies and ionization potentials of relevant
species 152
12.2 Kinetic behavior of SCO 4- C2H5, CH2CO, and CH3CHO observed in
the ejLs-2-butene-ozone reaction 156
12.3 Kinetic behavior of HCHO, CH3OH, HCOOH, C2H5, and H202
observed in the cis-2-butene-ozone reaction 157
12.4 Dependence of signal at m/e 104 on olefin concentration and
on time in the £is-2-butene-ozone reaction 158
12.5 Kinetic behavior of HCHO and CH3CO in the ethylene-ozone
reaction , 160
12.6 Kinetic behavior of HCHO and CH3COCH3 in the isobutene-ozone
reaction 161
viii
-------
LIST OF TABLES
Number Page
1.1 Rate constants k2 for the reaction of 0(3P) atoms with
organics 17
1.2 Arrhenius parameters for 0(3P) atom reactions 18
1.3 Comparison of the room temperature rate constants, k2, and
the activation energies, E, for n-butane and the aromatics
from the present work with selected literature values 19
A CP
2.1 Relative rate constants k2 /k2 for the reaction of 0(3P)
atoms 29
2.2 Arrhenius parameters for the reaction of 0(3P) atoms 31
2.3 Comparison of the room temperature rate constants, k2, and
activation energies, E, for ketene, acrolein and croton-
aldehyde from the present work with literature values 33
3.1 Product yields observed from the reaction of 0(3P) atoms with
toluene 40
3.2 Product yields from the reaction of 0(3P) atoms with 1-methyl-
cyclohexene (N2 = 1) 43
4.1 Columns used in gas chromatographic analysis of the products
of the photooxidation of azomethane in the presence of NO at
room temperature 51
4.2 Infrared extinction coefficients and limits of detection of
some reactants and products in the photooxidation of azo-
methane in the presence of NO at room temperature 52
4.3 Reactant concentrations, rates of product formation and
quantum yields in the photooxidation of azomethane in the
presence of NO at room temperature 56
5.1 Experimental conditions used and the rate constants k2
obtained for the reaction of PAN with S02, C02, N02, NH3,
H20, 03, isobutene and acetaldehyde 66
5.2 Initial conditions and observed first order constants, kobs>
for the reaction of PAN with NO 68
ix
-------
LIST OF TABLES (cont.)
Number Page
5.3 Chemical lifetimes of PAN in a typical Los Angeles polluted
atmosphere .................................................. 72
6.1 Chemiluminescence efficiency of amines reacting with PAN
(1000 ppm) in the gas phase and in solution ................. 76
7.1 PAN products, % of total, from reaction of isobutyraldehyde . 83
8.1 Thermodynamic and Arrhenius parameters for the N02 catalyzed
geometric isomerization of the 2-butenes and 2-pentenes ..... 92
8.2 Calculated and experimental data for the reactions of
nitrogen dioxide with selected olef ins ...................... 94
9.1 Analytical methods .......................................... 99
9.2 Reaction products from the photooxidation of C^E^ by N02 .... 103
11.1 Summary of the chemiluminescing species (200 < X < 1100 nm)
identified in ozone-olefin reactions in the presence and
absence of oxygen ........................................... 126
11.2 Reaction orders of emitting species in ozone and olef in in
the presence and absence of oxygen . . ........................ 127
11.3 Rate constants for the cis-2-butene and isobutene reactions
with ozone under varying conditions of total pressure and
oxygen concentrations ....................................... 130
11.4 Comparison of the rate constants obtained in this work with
literature room-temperature absolute values ................. 132
11.5 Absolute rates of light emission for each emitting species
in the reactions of 2% 03/02 with ethylene and cis-2-butene . 131
12.1 Observed peak mass numbers and their assignments ............ 153
12.2 Stable products observed in the reaction of ozone with
ethylene, isobutene, and cis-2-butene ....................... 162
-------
ACKNOWLEDGEMENTS
The following personnel performed the experimental work or otherwise
contributed directly to the research described in this report:
Roger Atkinson
Karen R. Darnall
Barbara J. Finlayson
Dennis R. Fitz
Haftmut Fuhr
Jeffrey S. Gaffney
D. Alan Hansen
Dennis M. Hebert
Alan C. Lloyd
Christopher T. Pate
John W. Peters
Jeremy L. Sprung
Peter H. Wendschuh
xi
-------
SECTION I
INTRODUCTION
The research work described in this report was carried out in the Depart-
ment of Chemistry, University of California, Riverside, under EPA Grant 800649
(formerly AP-109) during December 1, 1971 - November 30, 1974. The work
described deals with various aspects of the kinetics, mechanisms, and products
of reactions involved in photochemical air pollution and possible implications
for the chemical reaction mechanism employed in urban airshed models.
The principal components of an ab initio urban airshed model of the
formation of photochemical smog are an emission inventory, a meteorological
program, and a chemical reaction mechanism. The emission inventory specifies
where, when, which, and in what concentrations pollutants are emitted into
an air parcel; the meteorological program specifies the size and trajectory
of an air parcel; and the reaction mechanism specifies the chemical and, if
possible, physical (gas-to-particle conversion) transformations which the
pollutants undergo within the air parcel. Together, these three components
of the airshed model specify the rate of formation of secondary pollutants—
such as ozone, nitrogen dioxide, oxygenates, nitrogenous compounds, and
aerosol—as a function of both place and time.
Before it can be applied to an assessment of air pollution problems, an
airshed model must be validated against ambient air data characteristic of
the air basin for which it is a model. Successful validation of an ab initio
airshed model for the formation of photochemical smog would complement and
extend existing, largely empirical airshed models and would clearly be of
great utility for the cost-effective formulation of air pollution control
strategies, land-use plans, environmental impact statements, and air pollution
1 2
episode health warning systems. '
3-7
Because the chemistry of photochemical smog is exceedingly complex
even homogeneous reaction mechanisms for smog formation are unusually lengthy,
and validation of such mechanisms requires the development of a very broad
and detailed data base. Homogeneous reaction mechanisms constructed from
-------
the current data base simulate the rates of disappearance of aliphatic hydro-
carbons and the rate conversion of NO to N02 with encouraging fidelity.
However, the behavior of many photooxidation products and aromatic compounds,
such as toluene, is still modeled with considerable uncertainty, and no
existing models treat such heterogeneous processes as aerosol formation in
detail.
The types of data most urgently needed to permit further refinement and
ultimate validation of smog reaction mechanisms have been authoritatively
reviewed ' and discussed. Such data include:
• Precise concentration/time profiles for all major and significant trace
smog species formed during nitrogen-oxide-promoted photooxidations of
characteristic hydrocarbon pollutants and their major atmospheric oxi-
dation products.
• The effects of added S02 and particulate upon the course of hydrocarbon/
nitrogen oxide photooxidations, particularly the formation of sulfate
and aerosol.
• Rates of formation and properties of aerosols formed during hydrocarbon/
nitrogen oxide photooxidations.
• Photolytic behavior of photoactive secondary pollutants, especially
aldehydes and nitrous acid.
• Temperature-dependent homogeneous and heterogeneous rate and equilibrium
data for the reactions between oxides of nitrogen and water and their
oxyacid products.
• Mechanistic data defining the effects of molecular oxygen upon atmospheric
hydrocarbon oxidation pathways.
• Data on reaction rates and products, both stable and transient, formed
by the reactions of 0(3P) atoms, and particularly OH and H02 radicals,
with important atmospheric pollutants and their major oxidation products—
e.g., reactions of OH with olefins and aromatic compounds yield products
of uncertain identity and quantity.
The studies of the University of California have focused on selected
aspects of these requirements, including reactions of 0(3P) with olefins and
aromatic compounds and selected reactions of CH302 and 03.
-------
REFERENCES
1. J. N. Pitts, Jr., A. C. Lloyd and J. L. Sprung, "Chemical Reactions in
Urban Atmospheres and Their Application to Air Pollution Control
Strategies," Proceedings of the International Symposium on Environmental
Measurements, Beckman Instruments Process S.A., Geneva, October 2-4,
p. 27 (1974).
2. J. N. Pitts, Jr., A. C. Lloyd and J. L. Sprung, Chem. Brit., 11, 247
(1975).
3. P. A. Leighton, "Photochemistry of Air Pollution," Academic Press, New
York, New York (1961).
4. A. P. Altshuller and J. J. Bufalini, Environ. Sci. Technol., _5, 39 (1971).
5. K. L. Demerjian, J. A. Kerr and J. G. Calvert, Adv. Environ. Sci.
Technol., 4-, 1 (1974).
6. J. N. Pitts, Jr. and B. J. Finlayson, "The Mechanisms of Photochemical
Air Pollution and the Chemist: Past Perspectives and Future Challenges,"
Angewandte Chem. Int. Ed., 14, 1 (1975).
7. B. J. Finalyson and J. N. Pitts, Jr., Science 192, 111 (1976).
8. M. C. Dodge, "Workshop on Mathematical Modeling of Photochemical Smog:
Summary of the Proceedings," EPA-R4-73-010, U. S. Environmental
Protection Agency, Research Triangle Park, North Carolina, January, 1973.
9. J. H. Seinfeld, T. A. Hecht and P. M. Roth, "Existing Needs in the
Experimental and Observational Study of Atmospheric Chemical Reactions,"
EPA-R4-73-031, U. S. Environmental Protection Agency, Washington, D.C.,
June, 1973.
10. M. C. Dodge and T. A. Hecht, Abstracts of Papers to be presented at the
"Symposium on Chemical Kinetics Data for the Lower and Upper Atmosphere,"
Warrenton, Virginia, September, 1974, U. S. Environmental Protection
Agency, Research Triangle Park, North Carolina, August, 1974.
11. Triangle Universities Consortium on Air Pollution, "Symposium on Chemical
Aspects of Air Quality Modeling," University of North Carolina, Chapel
Hill, North Carolina, April, 1974.
-------
SECTION II
CONCLUSIONS
• Absolute rate constants have been determined for the reaction of ground
state oxygen 0(3P) atoms with a series of organics (ethylene, propylene,
benzene, toluene, o-, m-, p-xylene, 1,2,3-, 1,2,4-, and 1,3,5-trimethylbenzene)
and the inorganics NO(M = N2Q) and S02(M = N20) over the temperature range
299-392°K, using a modulation-phase shift technique. In addition, absolute
rate constants were determined for the reaction of 0(3P) atoms with n-butane,
methoxybenzene, and o-cresol at room temperature. The rate constants at room
temperature were observed to range from (1.44 ± 0.2) x 107 liter mole sec
for benzene to (2.01 ± 0.22) x 109 liter mole" sec" for propylene.
• Rate constants for the reaction of 0(3P) atoms with selected olefins
(propylene, 1-methylcyclohexene, and 1,3-cyclohexadiene) , monoterpenes
(a-pinene, 3-pinene, and d-limonene) , and unsaturated aldehydes (acrolein
and crotonaldehyde) were determined over the temperature range 296-423°K,
relative to that for the reaction of 0(3P) atoms with cyclopentene. In
addition, relative rate constants for the reaction of 0(3P) atoms with ketene
and toluene were determined at 296°K and 423QK, respectively. These relative
rate constants were placed on an absolute basis, using a rate constant for
the reaction of 0(3P) atoms with propylene derived from the literature. The
absolute room temperature rate constants obtained ranged from 2.3 x 108 liter
—1 —1 —1 —1
mole sec for acrolein to 6.5 x 10 10 liter mole" sec" for d-limonene.
• Products from the gas phase reaction of 0(3P) atoms with toluene and
1-methylcyclohexene have been studied in order to determine the effect of
aromaticity on the reaction mechanism. 0(3P) atoms were generated by the
mercury photosensitization of N20 in a static system and products were
analyzed by gas chromatography. For toluene, the major volatile products
observed were the cresols (mainly o-cresol), CO, and phenol, together with a
large amount of tar formation. In the case of 1-methylcyclohexene, the
products were the CyH^O isomers expected from the general reaction mechanism
of Cvetanovic. The differences in the products between the cyclic olefin
-------
and the aromatic hydrocarbon were explained by the differences in the rate
constants for reaction of 0(3P) atoms with the reactant and with the reaction
products. Thus, the reaction of 0(3P) atoms with the aromatic hydrocarbons
is slow and produces mainly highly reactive unsaturated products, while for
the simple olefins the initial reaction is very fast and forms largely
saturated, and hence unreactive, products. Therefore, the extent of secondary
reactions is much smaller in the latter case.
• The reaction of NO with CH302 radicals generated in the photooxidation
of azomethane (X>320 nm) was investigated at 296 ± 2°K using long-path infra-
red and gas chromatographic techniques. At short photolysis times, CHsONO
and N02 were identified as products with quantum yields of 1.7 ± 0.2 and
1.9 ± 0.3, respectively. The data show that the reaction CH302 + N02 •>
CH30 + N02 is the only reaction path for the reaction of CH302 radicals with
NO.
• Rate constants for the removal of peroxyacetyl nitrate (PAN) in the
presence of S02, isobutene, CO, CHsCHO, N02, H20, 03, NHs, and NO were deter-
mined at 296 ± 1°K by monitoring the decay of PAN in excess reactant as a
function of time by long-path infrared spectroscopy and/or gas chromatography.
The reactions with all the reactants, except 03 and NO, were sufficiently
slow that heterogeneous removal processes for PAN in their presence could not
be excluded. For the reaction of PAN with NO, the initial rates of removal
of PAN were observed to be first order in PAN and independent of NO concen-
-4 -1
tration, with a first-order rate constant of (2.8 ± 0.8) x 10 sec . A
possible reaction mechanism for the thermal reactions of PAN is proposed.
• Chemiluminescence from the reactions of PAN with a series of amines has
been investigated. In addition, spectra have been obtained for the gas-phase
reactions of PAN (Xmax ~ 650 nm) and ozone ( max ~ 520 nm) with triethylamine
at atmospheric pressure. The two chemiluminescent reactions can be distin-
guished optically, and PAN concentrations as low as 6 ppb were detected by
this technique.
• A study of the solution-phase reactions of PAN with a variety of organic
molecules (aldehydes, alcohols, amines, mercaptans, and sulfides) containing
functionality found in common biological systems has been carried out.
• The N02-catalyzed geometric isomerization of 2-butene and 2-pentene has
been studied over the temperature range 298-400°K. Both kinetic and equi-
librium data have been obtained. The combination of these results with
-------
thermodynamic calculations permits the estimation of rate constants for the
addition of N02 to the double bond of simple olefins. Comparison of these
rate constants to those for reaction of olefins with 03 suggests that atmo-
spheric consumption of olefins by processes initiated by N02 addition cannot
be significant.
• A study of the N0x-propylene photooxidation system has been carried out
using long-path infraredj gas chromatographic, and combined gas chromato-
graphic-mass spectrometric techniques.
• The vibration-rotation Meinel bands of OH(X2TT.) have been observed in
the chemiluminescent gas-phase reaction of ozone with acetaldehyde at ~3 torr
total pressure and room temperature. A weaker emission in the visible region
was tentatively identified as formaldehyde emission.
• Chemiluminescence from HCHO^A" -»• l^) t OHCxV)^^, and OH (A21 ) has
been observed in the gas-phase reactions of 2% 03 in 02, N2, or HE with a
series of simple olefins in a flow system at room temperature and at total
pressures of 2-10 torr. The vibration-rotation emission from OH(X2ir ) f<-
was virtually identical with OH Meinel band emission from the reaction,
H + 03 •*• OH(X2ir.) , + 02, confirming that H atoms are formed in 03-olefin
reactions under these experimental conditions. In the presence of 02,
glyoxal phosphorescence was identified in the 2-butene (cis or trans)
reaction, and methylglyoxal phosphorescence was tentatively identified from
the reactions of isobutene, 2-methyl-2-butene, and 2,3-dimethyl-2-butene
with ozone. The number of quanta emitted per molecule of reactant consumed
at 4.6 torr total pressure in the ethylene and cis-2-butene reactions were
estimated to be 10~7 for HCHO^A" •> XAX) and 10~5 quanta for (CHO)2(3A -»• 1A. )
_y u g
in the cis-2-butene reaction. Approximately 10 quanta were emitted in
the (9,3) transition of vibrationally excited OH per molecule of reactant
consumed, indicating that the formation of vibrationally excited OH is a
surprisingly efficient process under these conditions. In the ethylene,
cis-2-butene, and isobutene reactions, the time decay of light emission from
each excited species was exponential in 02 but nonexponential in N2 or He,
with increased emission intensities occurring at reaction times <0.1 sec in
the latter case. Rate constants, determined from the loss of 03 in both 02
and N2 as diluents, assuming 1:1 stoichiometry, were a factor of 2-5 times
greater in N2. In 10 torr of 02, the measured rate constants for the
ethylene, cis-2-butene, and isobutene reactions, respectively, were
6
-------
(1 ± 1) x lo3, (6.3 ± 1.9) x 10\ and (5.4 ± 2.3) x 103 liter mole"1 sec i.
Major products of the reaction of cis-2-butene with ozone in either 02 or
N£ as the carrier gas were, in addition to acetaldehyde, 2-butanone (possibly
from the OH-cis-2-butene reaction) and methyl vinyl ketone, which was observed
only in oxygen. The results are discussed in terms of the O'Neal-Blumstein
theory of gas-phase ozone-olefin reactions, and the possible role of these
reactions in photochemical smog formation is considered.
• The room-temperature gas-phase reactions of 03 with ethylene, propylene,
and cis-2-butene have been investigated at total pressures of ~2 torr, using
photoionization mass spectrometry. Both radical species and stable products
were identified and monitored as a function of time.
REFERENCE
1. R. J. Cvetanovic, Adv. Photochem., _!, 115 (1963).
-------
SECTION III
RECOMMENDATIONS
From the results described in the report and literature data, it is
apparent that the following data need to be obtained in order to more fully
understand the chemistry of photochemical air pollution.
a) To determine the prodiucts and mechanism of the reaction of ozone
with a series of olefins at atmospheric pressure in the presence
of 02. A full product study as a function of Os/olefin ratio,
total pressurei 02 concentration would be preferable.
b) The products of the reaction of 0(3P) atoms with aromatics should
be determined, using as wide a variety of techniques as possible
(such as photoionization mass spectroscopy, gas chromatography,
etc.). In addition, the effect of Q£ on the products and mechanism
should be investigated.
c) Rate constants, products, and mechanisms of the reaction of OH
radicals with olefins and aromatic hydrocarbons should be determined.
Products of such reactions should be determined, if possible, under
conditions applicable to polluted urban atmospheres.
d) The rate constants and products of the reactions of H02, alkoxy,
and alkylperoxy radicals with NO, N02, 02, and selected organics
should be determined.
This research work will require major efforts involving difficult
experimental systems, but such knowledge is a necessary input into detailed
chemical kinetic computer models of photochemical air pollution.
-------
SECTION IV
EXPERIMENTAL METHODS, RESULTS AND DISCUSSION
-------
1. DETERMINATION OF ABSOLUTE RATE CONSTANTS FOR THE REACTION OF 0(3P) ATOMS WITH
ORGANICS OVER THE TEMPERATURE RANGE 299-392°K
Experimental. Ground state oxygen atoms were generated by the mercury
1 2
photosensitization of N20 in a flow system. ' 253.7 nm resonance radiation
was emitted by a modulated radio-frequency-powered low pressure mercury arc
driven by a six-turn coil connected to a 14.0 MHz radio frequency generator
operating at -125 watts.
The 253.7 nm resonance line was isolated by passage through a Corning
7-54 filter plus an aqueous solution of 210 gm NiSO^/60 gm CoSOit liter .
The 253.7 radiation was collimated into the reaction cell and was monitored
by a Bausch and Lomb UV monochromator-RCA 1P28 photomultiplier combination.
The Pyrex reaction cell, 12 cm in length and 2.5 cm in diameter, had Supracil
end windows and a short sidearm fitted with a Pyrex window. Oxygen atoms
were monitored by observation of the N02 afterglow produced by the addition of
known flows of NO to the reactant stream. The afterglow emission was detected
through the Pyrex side window by a cooled EMI 9659A photomultiplier tube
fitted with a cut-off filter transmitting A£450 nm. The phase shift, 0, be-
tween the incident 253.7 radiation and the N02 air afterglow was determined
1-2
using a PAR HR8 lock-in amplifier with a Type C preamplifier. The reaction
cell was enclosed in a furnace whose temperature was measured by a thermocouple
probe. The reactant gas stream was observed to be within ±0.5°K of the
furnace temperature under all conditions used.
Absorbed light intensities in the reaction cell were determined by GC
, 3
analysis of propylene oxide from the 0(JP) + propylene reaction at ~50 torr
total pressure to be typically 1 x io16 quanta sec~ , corresponding to an
3 -7 —1
O^P) atom production rate of 5 x 10 mole liter . With these absorbed
light intensities, residence times in the reaction cell of 0.26-0.46 sec,
and the flows used (total flow rate (2.8-5.1) x 10 mole sec ), reactant
conversions ranged from <2% for propylene or 1,3,5-trimethylbenzene to 10.03%
for n-butane or benzene.
The gases used had the following purity levels, according to the manu-
facturer: N20 >98.0%; NO >99.0%; Ar >99.998%: ethylene >99.5%; propylene
>99.0%; n-butane >99.9%; S02 £99.8%; o-cresol >99.0%; methoxybenzene >99.0%.
Analysis by GC showed the following impurities in the less reactive hydro-
carbons: n-butane, -0.01% butenes, mainly trans-2-butene; benzene, 0.5%
10
-------
toluene and <0.5% xylene; toluene, 0.1% benzene and 0.05% m-xylene; o-xylene,
10.2% impurity; m-xylene, 1.0% p-xylene and 0.4% ethylbenzene; p-xylene,
0.3% m-xylene; 1,2,3-trimethylbenzene, 0.2% of the 1,3,5-isomer and 2.2% of
the 1,2,4-isomer; 1,2,4-trimethylbenzene, 1.0% of the 1,3,5-isomer; 1,3,5-
trimethylbenzene, 0.2% of the 1,2,4-isomer and 0.1% ethyltoluene. ~10% NO
in argon was passed through Linde Molecular Sieve 13X to remove any H20 and
N02 present, while N20, S02, ethylene, propylene and n-butane were dried by
passage through traps containing Drierite.
For the rate constant determinations for the aromatics, a known fraction
of the N20 flow was saturated with the aromatic vapor at 22°C (11°C in the
case of benzene). Aromatic concentrations in the gas stream were determined
before entering the reaction cell by their UV absorption using a 10.0 cm
path length cell and ;'a Gary 15 spectrophotometer. The absorption cell was
calibrated using known pressures of the aromatic as measured by a MKS Baratron
guage. NO, S02, ethylene, propylene and n-butane concentrations were calcu-
lated, with an estimated accuracy of ±3-4%, from the observed partial flow
rates and total pressure in the reaction cell. The gas stream was saturated
with mercury vapor at room temperature before entering the reaction cell.-
All flows were monitored by calibrated rotameters and controlled by needle
valves.
Results. The reactions of 0(3P) atoms with NO, S02, ethylene, propylene,
2
n-butane, and a series of aromatics were studied at room temperature. No
chemiluminescent emission was observed in the absence of NO, except in the
case of 0(3P) + ethylbenzene which precluded any rate constant determination
for this hydrocarbon.
For sinusoidally modulated 253.7 nm radiation and the following reaction
scheme: '
11
-------
Hg (6) + hv (253.7 nm) -> Hg (6^) (1)
Hg (63P!) + N20 -> Hg (6*8 ) + N2 + 0(3P)
0(3P) + reactant -> products (2)
0(3P) + NO + M -*• N02 + M (3a)
•> N02 + M (3b)
N02 •»• N02 + hv
*
N02 + M -*• N02 + M
The phase shift, 0, between the N02 afterglow emissions and the incident
12
253.7 nm radiation under the conditions used is given by '
- ' i
tan 0 = 2irv/k2[ reactant] + k3[NO][M])
where k3 = k3a + k3fc and v is the modulation frequency. In this study, the
modulation frequency was in all cases 1200 ± 1 Hz and total pressure varied
over the range 48-135 torr.
The rate constant k3 for the reaction 0 + NO + M was determined directly
using the equation
2irv/tan 0 = k3[NO][M] (II)
The fact that argon is a less efficient third body than N20 in reaction (3)
4
was corrected for using literature rate constant data for reaction (3).
For determinations of the rate constants k2, [NO][M] was held constant
and [reactant] varied. Figure 1.1 shows an example of plots of 2rrv/tan 0
against [reactant] for the xylenes at room temperature. It is seen from
equation I that such plots h^ve intercepts of k3[NO][M] and slopes of k2.
The intercepts of such plots were determined from the experimental [NO][M]
and the rate constant k3. Runs to check these intercepts agreed with the
calculated values within the expected experimental error of a single run
t
(5-10%). Table 1.1 gives the values of k2 obtained.
E/RT
The Arrhenius parameters (k = A e ' ) obtained from this data are
12
-------
m-XYLENE
p-XYLENE 9
0.4 0.8
[XYLENE] moie liter1
1.2x10
Figure 1.1 Plots of 2irv/tan 0 against xylene concentration for o-, m-,
and p-xylenes at room temperature.
13
-------
given in Table 1.2 while Figure 1.2 shows Arrhenius plots for 0(3P) +
aromatic hydrocarbons.
Discussion. As shown in Figure 1.3, for propylene there is very good
agreement with the data of Kurylo (E = 0.076 ± 0.044 kcal mole" ), Huie et
67 8
al., Stuhl and Niki and Cvetanovic and co-workers. For ethylene, the
present room temperature value is in good agreement with those of Stuhl and
79 8
Niki, ' and Cvetanovic and co-workers. While the temperature dependence
determined in this work agrees well with that of Davis et al. (E - 1.13 ±
0.032 kcal mole), their absolute values (and that of Kurylo and Huie )
lie some 15-20% higher. Earlier literature data for C2Hi» obtained by dis-
charge flow techniques range from (2.2-7.2) x 108 liter mole sec at
12—17
room temperature, with reported activation energies of 1.6 kcal
-112,13 _i 14
mole and 1.1-2.8 kcal mole
For the reaction 0 + NO + M (M = N20) the present room temperature
value of k3 agrees with those obtained by Kaufman and Kelso (ks = 4.7 x io10
liter2 mole"2 sec"1) and Atkinson and Cvetanovic (k3 - (3.9 ± 0.3) x io10
1 1
liter2 mole sec ), although the activation energy obtained in this work
(E = 0.90 ± 0.2 kcal mole" ) is lower than previously reported: -1.93 ±
IQ ?n 91
0.1 (M = N2); -1.8 ± 0.4 (M - 02); -2.0 (M = Ar) and -1.6 ± 0.3 kcal
—1 18
mole (M = N20). No immediate reason for this discrepancy is obvious.
There is no previously reported value of k2 for the reaction 0 + S02 +
M (M = N20), but the temperature dependence obtained in this work (E = 2.00 ±
0.4 kcal mole ) is in good agreement with the value of E = 2.24 kcal mole
22
for M = N2 obtained by Davis using a flash photolysis-resonance fluorescence
technique.
Table 1.3 compares the present room temperature rate constants, k2, and
activation energies, E, for n-butane and the aromatic hydrocarbons with the
literature data. It can be seen that there is generally good agreement with
23—26
the relative rate data of Cvetanovic and co-workers, placed on an abso-
lute basis using k2 (ethylene) - 3.37 x io9 e~^1270 * 300>/RT liter mole"1
sec determined in this work. The present value for n-butane falls in
27 12
between those of Herron and Huie and of Elias and Schiff.
It can be seen from Table 1.3 that the room temperature rate constants
29
obtained by Mani and Sauer using a pulsed radiolysis technique are a factor
of 3-6 higher than the present work, possibly because of the complexity of
their experimental technique. Similarly, the room temperature rate constant
14
-------
5x10
IxlO
'o
CD
CO
o
-------
IxlO
JO
i
o
CD
CO
A
0+C2H4
2.0
3,0 4.0
1000/T (°K)
5.0
5x10
4
10
xlO
10
Figure 1.3 Arrhenius plots of log k against 1000/T:
-•- This work; -0~ Davis et al; -T&— Kurylo;
, 11 8
X Huie et al; A Kurylo and Huie; O Furuyama et al;
9 Stuhl and Niki7'9
o
-------
Table 1.1
Rate constants k2 for the reaction of 0(3P) atoms with organics. The errors include
precision limits derived from least squares standard deviations and estimated accuracy
of other parameters such as pressure and composition of mixtures.
Reactant
Ethylene
Propylene
n-Butane
Benzene
Toluene
o-Xylene
m-Xylene
p-Xylene
1,2, 3-Trimethylbenzene
1,2, 4-Tr imethylbenzene
1, 3, 5-Tr imethylbenzene
Methoxybenzene
o-Cresol
Temperature °K
300.7 ±0.3
341.2 ±0.5
392.2 ±0.5
300.0 ±0.2
341.2 ±0.5
392.2 ± 0.5
301.4 ± 0.2
300.3 ±0.3
341.2 ±0.5
392.2 ± 0.5
300.4 ±0.2
341.2 ±0.5
392.2 ± 0.5
299.1 ±0.3
341.2 ± 0.5
392.2 ± 0.5
299.5 ±0.5
341.2 ±0.5
392.2 ± 0.5
300.4 ±0.5
341.2 ± 0.5
392.2 ±0.5
299.7 ± 0.4
341.2 ±0.5
392.2 ± 0.5
300.2 ± 0.3
341.2 ± 0.5
392.2 ± 0.5
299.9 ± 0.5
341.2 ± 0.5
392.2 ± 0.5
302.4 ±0.2
301.5 ±0.2
—8 —1 —1
10 x k2 liter mole sec
4.00 ± 0.40
5.15 ± 0.50
6.58 ± 0.65
20.1 ± 2.2
22.2 ± 2.3
20.1 ± 2.0
0.188 ± 0.02
0.144 ± 0.02
0.303 ± 0.035
0.69 ± 0.08
0.450 ± 0.045
0.85 ± 0.09
1.52 ± 0.15
1.05 ± 0.11
1.72 ± 0.18
2.77 ± 0.3
2.12 ± 0.21
3.09 ± 0.3
5.0 ± 0.5
1.09 ± 0.11
2.01 ±0.2
2.94 ± 0.3
6.9 ± 0.7
9.9 ± 1.0
13.0 ± 1.5
6.0 ± 0.6
8.1 ± 0.8
11.5 ± 1.2
16.8 ± 2.0
18.9 ± 2.0
22.8 ± 3.0
0.66 ± 0.7
3.5 ± 0.8
17
-------
Table 1.1 (cont.)
Reactant Temperature *
NO(M - N20) 300.5 ± Q.4
341.2 ±0.5
392.2 ± 0.5
S02(M = N20) 299.2 ± 0.5
341.2 ± 0.5
399.2 ± 6.5
K 10"10 x k, or k*
4.29
3.62
3.02
0.115
0.175
0.255
Table 1.2
—2 —1
(liter2 mole" sec )
± 0.43
± 0.33
± 0.31
± 0.015
± 0.02
± 0.03
Arrhenius parameters for 0(3P) atom reactions
Reactant A *
Ethylene
Propylene
Benzene
Toluene
o-Xylene
m-Xylene
p-Xylene
1,2, 3-Trimethylbenzene
1,2, 4-Tr ime thylbenzenc
1,3, 5-Tr imethylbenzene
A *
NO(M « N20)
S02(M - N20)
10"9 (liter mole"1 sec"1)
3.37
2.08
11.1
8.2
6.25
7.7
7.9
10.3
9.35
6.05
10~9(liter2 mole"2 sec"1)
9.6
33.2
E (kcal mole"1)
1.27 ± 0.2
0.00 ± 0.3
3.98 ± 0.4
3.10 ± 0.3
2.43 ± 0.3
2.15 ± 0.3
2.54 ± 0.3
1.60 t 0.3
1.65 ± 0.3
0.77 ± 0.3
E (kcal mole"1)
-0.90 ±0.2
2.00 ± 0.4
18
-------
Table 1,3
Comparison of the room temperature rate constants, k2, and the activation energies, E, for n-butane and the
aromatics from the present work with selected literature values.
Reactant
n-Butane
Benzene
Toluene
o-Xylene
m-Xylene
p-Xylene
10"8 x
Present work
0.188 ± 0.02
0.144 ± 0.02
0.45 ± 0.045
1.05 ± 0.11
2.12 ± 0.21
1.09 ± 0.11
k2 liter mole sec E kcal mole
Derived from ,. Derived from »
Literature relative rates* Present work Literature relative rates*
0.2512 0.18
0.1327
28 28
0.28 ± 0.07 0.16 3.98 ± 0.4 4.4 ± 0.5 3.27 i 0.7
0.36 ± 0.0729
1.4 ± 0.329 3.10 ± 0.3 2.47 ± 0.4
6.7 ± 1.629 2.43 ± 0.3
7.7 ± 2.029 2.15 ± 0.3
4.5 ± 1.429 2.54 ± 0.3
a) Ref. 23-26. Placed on an absolute basis using k2 - (ethylene) - 3.37 x io9 e~(127° * 300>''Rf iiter mole'1 sec"1.
-------
28
for benzene of Bonanno et al. obtained by discharge flow-mass spectrometry
is a factor of 2 higher than the present value, although their activation
—1 ' "
energy of 4.6 ± 0.5 kcal mole , obtained over the range 255-305°K, is in
good agreement with that determined in this work.
20
-------
REFERENCES
1. R. Atkinson and R. J. Cvetanovic, J. Chem. Phys., 55, 659 (1971).
2. R. Atkinson and J. N. Pitts, Jr., J. Phys. Chem., 78, 1780 (1974); 79,
295 (1975); _79, 541 (1975). ~"~
3. R. J. Cvetanovic, Can. J. Chem., 36. 623 (1958).
4. F. Kaufman and J. R. Kelso, Symp. Chemiluminescence, Duke University,
Durham, USA, (1965).
5. M. J. Kurylo, Chem. Phys. Letters, 14, 117 (1972).
6. R. E. Huie, J. T. Herron and D. D. Davis, J. Phys. Chem., ^6, 3311 (1972).
7. F. Stuhl and H. Niki, J. Chem. Phys., 55, 3954 (1971).
8. S. Furuyama, R. Atkinson, A. J. Colussi and R. J. Cvetanovic, Int. J.
Chem. Kinet., £, 741 (1974).
9. F. Stuhl and H. Niki, J. Chem. Phys., 5_7_, 5403 (1972);
10. D. D. Davis, R. E. Huie, J. T. Herron, M. J. Kurylo and W. Braun, J.
Chem. Phys., _56, 4868 (1972).
11. M. J. Kurylo and R. E. Huie, J. Chem. Phys., _58_, 1258 (1973).
12. L. Elias and H. I. Schiff, Can. J. Chem., 38, 1657 (1960).
13. L. Elias, J. Chem. /Phys., 38, 989 (1963).
14. A. A. Westenberg and N. deHaas, Symp. Combust., 12th, Univ. Poitiers,
France, 1968, 289 (1969).
15. J. M. Brown and B. A. Thrush, Trans. Faraday Soc., 63, 630 (1967).
16. C. Tanaka, S. Tsuchiya and T. Hikita, J. Fac. Eng. Univ. Tokyo Ser. A,
j>, 62 (1967).
17. H. Niki, E. E. Daby and B. Weinstock, Symp. Combust., 12th Univ. Poitiers,
France, 1968, 277 (1969).
18. R. Atkinson and R. J. Cvetanovic, J. Chem. Phys., 56, 432 (1972).
19. F. S. Klein and J. T. Herron, J. Chem. Phys., 4^, 1285 (1964).
20. M. A. A. Clyne and B. A. Thrush, Proc. Roy. Soc., A269, 404 (1962).
21. T. G. Slanger, B. J. Wood and G. Black, Int. J. Chem. Kinet., ,5, 615
(1973).
22. D. D. Davis, Can. J. Chem., 52, 1405 (1974).
21
-------
REFERENCES (cont.)
23. R. J. Cvetanovic, J. Chem. Phys., 33, 1063 (1960).
24. R. J. Cvetanovic, Adv. Photochem., 1., 115 (1963).
25. G. Boocock and R. J. Cvetanovic, Can. J. Chem., 39, 2436 (1961).
26. G. R. H. Jones and R. J. Cvetanovic, Can. J. Chem., 39, 2444 (1961).
27. J. T. Herron and R. E. Huie, J. Phys., Chem., .73, 3327 (1969).
28. R. A. Bonanno, P. Kim, J. H. Lee and R. B. Timmons, J. Chem. Phys., 57,
1377 (1972).
29. I. Mani and M. C. Sauer, Jr., Adv. Chem. Ser., 82, 142 (1968).
22
-------
2. RATE CONSTANTS FOR THE REACTION OF 0(3P) ATOMS WITH SELECTED
OLEFINS, MONOTERPENES AND UNSATURATED ALDEHYDES
As part of an investigation into the rates and products of ,the reactions
of 0(3P) atoms with unsaturated organic compounds involved in the formation
of photochemical air pollution, rate constants for the reactions of 0(3P)
atoms with ketene, acrolein, crotonaldehyde, a-pinene, g-pinene, d-limonene,
1-methylcyclohexene, and 1,3-cyclohexadiene have been determined relative
to that for the reaction of 0(3P) atoms with cyclopentene over the temperature
range 296-423°K. The reaction of 0(3P) atoms with cyclopentene has been
thoroughly investigated and has been extensively used to determine relative
2 3
0(3P) atoms reaction rate constants by Cvetanovic and co-workers. '
In order to test the experimental system and to place the relative,rate
constants on an absolute basis, propylene was included in the compounds
4-8
studied as its absolute rate constant is known to a good degree of accuracy
over, the temperature range used in this work. ,
Experimental. Ground state 0(3P) oxygen atoms were generated by the
mercury photosensitization of nitrous oxide at 253.7 nm:
o
) + hv —> Hg(63Px) (la)
N?0 —> Eg(61S ) + No + 0(3P) (1)
i. f o
In order to minimize short wavelength photolysis of the reactants and pro-
ducts, the 253.7 nm resonance radiation from a low pressure mercury arc was
passed through a Corning 7-54 filter to remove wavelengths <235 nm. A
conventional static high vacuum system was used, fitted with greaseless
stopcocks to minimize absorption of reagents and products. The cylindrical
quartz reaction cell, diameter 5.0 cm, length 20.0 cm, had a volume of 393 cm3
and, in order to ensure homogeneity, the gas mixture was circulated in a
total volume of 1043 cm3 by means of a magnetically driven all-glass circu-
lating fan. The reaction cell and circulating system were enclosed by a
furnace where temperature could be held constant to better than ±1°K over
the temperature range 296-423°K. Reactant pressures were measured using either
a 0-20 torr Wallace and Tiernan FA160 absolute pressure gauge or a 0-800 torr
mercury manometer.
23
-------
Reactant purities and purification procedures were as follows: ketene
was prepared by the pyrolysis of acetone and degassed and vacuum distilled
at 196°K. The resulting ketene purity level was >98% with acetone being the
major impurity. All other reagents were >98% stated purity and were further
purified by thorough degassing at 196°K or 77°K and bulb-to-bulb distillation
in vacuum. As a further check on the reactant purities, the reagents used
were gas chromatographed on a 10 ft x 1/8" 3,$'-oxydipropionitrile (9.7%
on 80/100 mesh Firebrick) column using a flame ionization detector, and mass
spectra were obtained on a Finnigan 3100D quadrapole mass spectrometer. In
all cases purity levels consistent with the above were observed.
Photolysis times were typically 5-15 minutes during which 5 cm3 samples
were periodically removed for analysis by a Carle gas sampling valve. The
sample was split into two streams and N2 from the N20 and C^^ from cyclo-
pentene were analyzed by gas chromatography. N2 was measured on a 5 ft x
%" Linde Molecular Sieve 13X column at room temperature by a thermal conduc-
tivity detector, while C^ was analyzed on a 6 ft x 1/8" Poropak Q column
at room temperature by a flame ionization detector. (18.9 ± 0.4)% C2Hij ^n N2
was used to periodically check retention times and relative responses in the
detectors.
Results. Relative rate constants were determined using the same tech-
2 3
nique developed by Cvetanovic and co-workers. ' From the reaction scheme:
0(3P) + cyclopentene — > a(C2Htt + CH2=CHCHO) + other products1 (2CP)
0(3P) + reactant A — > products (2A)
CP A
with rate constants k2 and k2 respectively, it can be shown that
k2A[A]
A
(C2H£t/N2)A k2U
where (C2Hit/N2) ~ and (C2H^/N2) are the C^/^ yield ratios in the absence
and presence of reactant A, respectively. In all cases runs were carried out
in the absence of cyclopentene to check that ethylene was not produced from
the reaction of 0(3P) atoms with reactant A. Only for ketene and acrolein
was any ethylene production observed. However, in both cases the C2Htt pro-
24
-------
duction from the reactant was calculated to cause less than 2% error in the
determination of the relative rate constants.
The 02^/^2 ratio for the reaction of 0(3P) atoms with cyclopentene was
determined before and after every rate determination and was observed to be
constant to better than ±5% at any given temperature providing that the
reaction cell was heated to >373°K under vacuum between rate constant deter-
minations for the pinenes, limonene, acrolein and crotonaldehyde. The C2Htf/N2
ratio from the reaction of 0(3P) atoms with cyclopentene alone was determined
to be 0.23 ± 0.02 over the pressure range 200-500 torr at temperatures from
296-623°K in good agreement with the work of Cvetanovic, Ring and Doyle.
For each run, at a given reactant/cyclopentene ratio, several determin-
ations of the ethylene and nitrogen yields were carried out as a function of
time. Typical results are shown in Figure 2.1, which shows the C2H^ yield
versus N2 yield for cyclopentene alone and for four runs with differing
limonene/cyclopentene ratios.
Relative rate constants k2 /k2 determined from the slopes of the plots
of C^i^ yield versus N2 yield are plotted in Arrhenius form in Figures 2.2
and 2.3, and given in Table 2.1. The relative Arrhenius parameters,
k2 A -(E—E )/RT
—— = • e , obtained by least squares analysis, are given in
k2 A
Table 2.2, along with the absolute Arrhenius parameters obtained by using
76ART 1 l
k2(propylene) = 2.4 x 1Q9 e ' liter mole sec , (k2(propylene) = 2.10 x
109 liter mole" sec" at 298°K), derived from the recent literature rate
4-8
constant data.
Discussion. In all cases, experiments were carried out under conditions
of low conversion where secondary reactions of 0(3P) atoms with the products
should have been negligible. Similarly, the reaction of 0(3P) atoms with
impurities was estimated to cause errors in the measured rate constants of <5%.
The value of k2(propylene)/k2(cyclopentene) of 0.181 ± 0.010 at 296 ±
2°K determined here is in good agreement with the ratio of 0.192 obtained by
A
Cvetanovic. Similarly, the present ratio of k2(toluene)/k2(cyclopentene) =
0.026 ± 0.003 at 423 ± 1°K agrees well with that of 0.023 interpolated from
the d^ta of Jones and Cvetanovic. When placed on an absolute basis using
the derived rate expression for 0(3P) + cyclopentene, the value of k2(toluene) =
(2.4 ± 0.3) x 108 liter mole" sec" at 423 ± 1°K agrees within experimental
error with the absolute value of 2.05 x 108 liter mole~ sec calculated from
25
-------
0
10 20 30 40
NITROGEN YIELD (ARBITRARY UNITS)
50
Figure 2.1 Plots of ethylene yield versus N2 yield for: a) cyclopentene
alone, b-e) cyclopentene/limonene = 8.64(b); 6.54(c); 3.24(d);
26
-------
10.0
5.0
2.0
o.
o _
1.3-CYCLOHEXADIENE
a-PINENE
0.5
0.2
PROPYLENE
O.I
I
2.2
2.6 3.0
IOOO/T(°K)
3.4
A,, CP
Figure 2.2 Arrhenius plots of log k2 /k2 against 1000/T
27
-------
10.0
5.0
2.0
o.
o —
1.0
0.20
0.10
0.05
0.02
0.01.
2.2
I-METHYLCYCLOHEXENE
•5
-PINENE
CROTONALDEHYDE
ACROLEIN
2.6 3.0
IOOO/T(°K)
3.4
A CP
Figure 2.3 Arrhenius plots of log k2 /k2 against 1000/T
28
-------
Table 2.1
A CP
Relative rate constants k2 /ka for the reaction of 0(3P) atoms. The error
limits are the least square standard deviations.
/
Reactant A Temperature °K k?*
Propylene 296
333
373
423
a-Pinene 296
333
373
423
3-Pinene 296
333
373
423
d-Limonene 296
333
373
423
1-Methylcyclohexene 296
333
373
423
1,3-Cyclohexadiene 296
333
373
423
± 2
± I
± 1
± 1
+ 2
± 1
± 1
± 1
± 2
± 1
± 1
± 1
± 2
± 1
± 1
± 1
± 2
± 1
± 1
± 1
± 2
± 1
± 1
± 1
0.181
0.203
0.225
0.231
1.38
1.84
2.22
2.73
1.30
1.64
2.07
2.41
5.61
6.74
7.45
8.16
4.21
3.93
3.22
2.71
4.33
3.43
3.05
2.74
v rp
V ., Lir
/k?
± 0.010
± 0.026
± 0.019
± 0.012
± 0.05
± 0.09
± 0.15
± 0.27
± 0.05
± 0.11
± 0.20
±0.12
± 0.45
± 0.50
± 0.45
± 0.57
± 0.17
±0.25
± 0.32
± 0.27
± 0.20
± 0.24
± 0.31
± 0.27
29
-------
Table 2.1 (cont.)
A CP
Reactant A Temperature °K kg /k2
Ketene 296 ± 2 0.024 ± 0.003
Acrolein 296 ± 2 0.020 ± 0.002
333 ± 1 0.035 ± 0.005
373 ± 1 0.059 ± 0.006
423 ± 1 0.081 ± 0.009
Crotonaldehyde 296 ± 2 0.044 ± 0.005
333 ± 1 0.070 ± 0.007
373 ± 1 0.100 ± 0.010
423 ± 1 0.150 ± 0.015
Toluene 423 ±1 0.026 ± 0.003
30
-------
Table 2.2
Arrhenius parameters for the reaction of 0(3P)
the least square standard deviations.
Reactant A
Propylene
Cyclopentene
ot-Pinene
B-Pinene
d-Limonene
1-Methylcyclohexene
1, 3-Cyclohexadiene
Acrolein
Crotonaldehyde
AA/ACP
A /A
0.
1.
13.
10.
19.
0.
0.
2.
2.
43
00
46
77
63
95
92
46
62
(EA-E
0
0
1
1
0
-0
-0
2
2
atoms. The indicated errors
CP — 1
)kcal mole
.50 ±0.08
.00
.34 ± 0.06
.24 ± 0.06
.72 ± 0.07
.90 ± 0.14
.•89 ± 0.11
."82 ± 0.19
.41 ± 0.05
AA liter
2.
5.
7.
6.
1.
5.
5.
1.
1.
in the activation energies are
-1 -la) A -1 3)
mole sec E kcal mole
4 x
6 x
5 x
0 x
1 X
o ..
1 x
4 x
5 x
109
109
1010
1010
10ll
109
109
1010
1010
0.076
-0.43 ±
0.91 ±
0.82 ±
0.30 ±
-1.33 ±
-1.32 ±
2.40 ±
1.98 ±
0.08
0.14
0.14
0.15
0.22
0.18
0.27
0.13
a' Placed on an absolute basis using k2(propylene) = 2.4 x io9 e liter mole sec
4-8
-------
the Arrhenius parameters obtained recently in these laboratories using
a modulation technique. Furthermore, from the activation energy for cyclo-
pentene (E = -0.43 ± 0.08 kcal mole ) an activation energy for 0(3P) +
toluene of E = 2.8 kcal mole" can be obtained from the data of Jones and
Cvetanovic in good agreement with E = 3.10 ± 0.3 kcal mole determined
recently.
Table 2.3 compares the room temperature rate constants and the acti-
vation energies determined in the present work for the reaction of 0(3P)
atoms with ketene, acrolein and crotonaldehyde with the literature values
12-17
determined using discharge flow techniques. There is seen to be general
agreement within the likely experimental errors for acrolein and croton-
12-14
aldehyde, and for ketene, the present rate constant is in good agreement
with that determined recently by Mack and Thrush, but lies between those
of Jones and Bayes and of glass et al.
It can be seen from Table 2.2 that cyclopentene, 1,3-cyclohexadiene
and 1-methylcyclohexene have similar preexponential factors and negative
activation energies. This behavior is similar to that observed for other
18
simple olefins such as cis-2-butene and tetramethylethylene. Negative
activation energies have also been reported for the other group Via atoms
19
with olefins and seem to be a general phenomena for the reaction of group
Via atoms with olefins of low ionization potential. Possible reasons for
the observation of negative Arrhenius activation energies have been discussed
18—21
by previous workers. Two recently postulated reasons for the observation
of negative Arrhenius activation energies are that either the preexponential
18 19 21
factor is temperature dependent * ' or that there is a temperature depen-
20
dent potential energy curve crossing probability.
However, a-pinene, $-pinene and d-limonene have preexponential factors
which are a factor of -10-20 higher than those for cyclopentene, 1-methyl-
cyclohexene and 1,3-cyclohexadiene and have low but positive activation
energies. It may be that for these more complex molecules abstraction
reactions are occurring together with addition of 0(3P) atoms to the double
bond. Thus the preexponential factors for 0(3P) atom abstraction from cyclo-
pentane and cyclohexane are 1.3 * 101l liter mole" sec" and 2.2 x 1Q11 liter
—1 —1 ?9 9^
mole sec respectively, ' while the preexponential factors for 0(3P)
atoms addition to simple olefins are in the region of 5 x 109 liter mole"
_! 22,24-27
sec . Similar behavior is postulated to occur in the case of the
32
-------
Table 2.3
Comparison of the room temperature rate constants, k2, and activation energies,
E, for ketene, acrolein and crotonaldehyde from the present work with literature
values.
Reactant
k2 x
io-8
liter
Present work
Ketene
Acrolein
Cr o tonald ehyd e
2.78
2.32
5.10
± 0.
± 0.
± 0.
35
23
58
mole
sec E kcal mole
Literature Present work Literature
3
1
5
1
5
.4
.7
.3
.6
14
±
+
17
12
0.315
0.416
>13 2.40 ± 0.27 2.012'13
1.98 ± 0.13 2.314
reaction of 0(3P) atoms with 1-butene where the abstraction reaction may be-
come important above ~260°K.
33
-------
REFERENCES
1. R. J. Cvetanovic, D. F. Ring and L. C. Doyle, J. Phys. Chem., 75,
3056 (1971).
2. R. J. Cvetanovic, Adv. Photochem., !_, 115 (1963).
3. R. J. Cvetanovic, J. Chem. Phys., 3J), 19 (1959); 33, 1063 (1960).
4. R. Atkinson and J. N. Pitts, Jr., J. Phys. Chem., 7£, 1780 (1974).
5. F. Stuhl and H. Niki, J. Chem. Phys., 55, 3954 (1971).
6. M. J. Kurylo, Chem. Phys. Letters, 14, 117 (1972).
7. R. Atkinson and J. N. Pitts, Jr., Chem. Phys. Letters, 2T_, 467 (1974).
8. S. Furuyama, R. Atkinson, A. J. Colussi and R. J. Cvetanovic, Int. J.
Chem. Kinetics, £, 741 (1974).
9. J. W. Williams and C. D. Hurd, J. Org. Chem., _5, 122 (1940).
10. G. R. H. Jones and R. J. Cvetanovic, Can. J. Chem., 39, 2444 (1961).
11. R. Atkinson and J. N. Pitts, Jr., J. Phys. Chem., _79, 295 (1975).
i
12. R. D. Cadle and E. R. Allen, "Chemical Reactions in Urban Atmospheres,"
C. S. Tuesday, Ed., Elsevier, p. 63 (1971).
13. R. D. Cadle, S. S. Lin and R. F. Hausman, Jr., Chemosphere, 1, 15 (1972).
14. R. D. Cadle, H. H. Wickman, C. B. Hall and K. M. Eberle, paper presented
at 167th American Chemical Society Meeting, Los Angeles, California (1974),
15. G. P. R. Mack and B. A. Thrush, J. Chem. Soc., Faraday Trans. I., 70,
187 (1974).
16. I. T. N. Jones and K. D. Bayes, Proc. Roy. Soc., A335, 567 (1973).
17. R. W. Carr, Jr., I. D. Gay, G. P. Glass and H. Niki, J. Chem. Phys.,
_49, 846 (1968).
18. D. D. Davis, R. E. Huie and J. T. Herron, J. Chem. Phys., 59, 628 (1973).
19. D. D. Davis and R. B. Klemm, Int. J. Chem. Kinet., _5, 841 (1973).
20. J. Connor, A. Van Roodselaar, R. W. Fair and 0. P. Strausz, J. Amer.
Chem. Soc., 93, 560 (1971).
21. R. Atkinson and R. J. Cvetanovic, J. Chem. Phys., 56, 432 (1972).
22. J. T. Herron and R. E. Huie, J. Phys. Chem., Ref. Data, _2» 467 (1963),
and references therein.
34
-------
REFERENCES (cont.)
23. J. T. Herron and R. E. Huie, J. Phys. Chem., 73, 3327 (1969).
24. M. J. Kurylo, Chem. Phys. Letters, 14, 117 (1972).
25. R. Atkinson and J. N. Pitts, Jr., Chem. Phys. Letters, 27, 467 (1974)
26. D. D. Davis, R. E. Huie, J. T. Herron, M. J. Kurylo and W. Braun, J.
Chem. Phys., 56, 4868 (1972).
27. R. E. Huie, J. T. Herron and D. D. Davis, J. Phys. Chem., 76, 3311
(1972).
35
-------
As part of an investigation into the rates and products of the reactions
of 0(3P) atoms with a variety of unsaturated organic compounds, the products
obtained from the reaction of 0(3P) atoms with toluene and 1-methylcyclohexene
have been studied in order to determine the effect of aromaticity on the
reaction mechanism.
Experimental. The experimental system has been described in detail in
Section 2 of this report, and only the essential details will be given here.
Ground state oxygen atoms were produced by the mercury photosensitization of
%0 in a closed circulating reaction system. The volumes of the reaction and
circulating systems were 4000 cm3 and 1043 cm3 for the studies on toluene and
1-methylcyclohexene, respectively. The reaction system was enclosed by a
furance whose temperature could be held constant to better than ±1°K over the
temperature range 295-425°K.
5 cm3 samples were periodically removed for analysis using a Carle gas
sampling valve. In all cases, the N2 yield from the N20 photosensitization
was used as an internal actinoineter to monitor the number of 0(3P) atoms
produced during the irradiations.
The experimental details for the two reaction systems studied are as
follows:
0(3P) + toluene
Reactions were carried out at 373 ± 2PK in order to avoid problems
associated with adsorption of reaction products on the glass and/or stainless
steel components of the gas sampling system. The product samples were split
into two fractions: CO and N2 were analyzed by gas chromatography on a 5 ft x
V Linde Molecular Sieve 13X column at 296 ± 2°K using a thermal conductivity
detector, while the organic products were detected by gas chromatography
using a flame ionization detector. Thus, CH^ and C2H6 were analyzed on a
6 ft x 1/8" Porapak Q column at 296 ± 2°K while the volatile phenolic products
were analyzed on a 10 ft x 1/8" 3% polyphenyl ether on DCMS Chromasorb W
column at 388 ± 2°K. In all cases gas chromatographic peaks were identified
by comparison of their retention times with those of authentic samples and
were quantified by calibration using known pressures of the reagents in the
reaction system.
A few subsidiary experiments were carried out at 296 ± 2°K in a 9.4 cm
36
-------
pathlength cylindrical quartz cell fitted with NaCl end windows. The reactions
were monitored by in-situ infrared absoprtion spectroscopy using a Perkin-
Elmer 221 spectrophotometer.
All chemicals used were of >98% purity.
0(3P) + 1-methylcyclohexene
CO and N2 were analyzed as described above while the volatile organic
products were analyzed by gas chromatography on a 10 ft x 1/8" 9.7% g,8-oxydi-
propionitrile on 80/100 mesh Firebrick column at 363 ± 2°K using a flame
ionization detector.
Identification of the gas chromatographic peaks was made by comparison
of their retention times with those of authentic samples, wherever possible,
and for the organic products, by elucidation of their structures from nmr,
infrared and mass spectrometric data.
In order to have sufficient reaction product to obtain pure samples for
spectral identification, a flow system similar to that described by Grovenstein
and Mosher was used and samples of the reaction products separated and
, -1
collected by preparative gas chromatography. A flow of 3.5 cur sec of ^0
was bubbled through mercury, then through 1-methylcyclohexene and the
resulting flow stream irradiated in a quartz photolysis tube by a spiral low
pressure mercury resonance lamp. Condensible products and unreacted 1-
methylcyclohexene were trapped out at 196 ± 2°K, with a typical reactant
conversion of -11%. This flow system was operated at 296 ± 2°K and 735 ±
10 torr total pressure (mainly N20). Identical gas chromatographic peaks
were observed with this system as with experiments carried out at lower con-
versions and at lower total pressures in the closed circulating system.
With the 6,3-oxydipropionitrile column operated at 296 ± 2°K, six of the
seven product peaks were collected in quantities sufficient to obtain nuclear
magnetic resonance (NMR) (Varian A60 and A60D instruments), infrared (Perfcin-
Elmer 221 spectrophotometer) and mass spectra (Finnigan 3100D and 3200E
quadrupole mass spectrometers).
l-Methyl-l,2-^epoxycylohexane was synthesized by an adaption of Hibbert
fy
and Burt's procedure. 3.0 cm3 of 1-methylcyclohexene (0.025 moles) was
added to a solution of 5.2 gm of m-chloroperbenzoic acid (85% purity, 0.025
moles) in 100 cm3 of methylene chloride. The solution was stirred for 3-4
hours at 273°K. The precipated m-chlorobenzoic acid was removed by filtration
and the resulting solution was washed with a) excess 10% sodium sulfite
37
-------
solution to remove any remaining peracid, b) 10% sodium carbonate to remove
any remaining benzoic acid and c) water to remove carbonate and any traces
of benzoic acid. The solution was then dried over anhydrous sodium sulfate,
filtered and the methylene chloride removed under vacuum. Gas chromato-
graphic analysis of the product showed the 1-methyl-l , 2-epoxycylohexane to
be >98% purity, with the remainder being mainly unreacted 1-methylcyclohexene
Commercially available 1-methylcyclohexene and 2-methylcyclohexanone had
purity levels >98% as confirmed by combined gas chromatography-mass spectro-
metric analyses*
Calibration of the gas chromatograph for CO and N2 was carried out by
gas sampling known pressures of these compounds.
Calibration using liquid samples of 1-methyl-l , 2-epoxycylohexane and 2-
methylcyclohexanone showed them to produce equal responses, in terms of peak
areas, and it was assumed that this response was valid for liquid injection
3
of all the other CyH^O isomers observed in this work. Gas phase sampling
via the Carle valve from the reaction system of identical pressures of 1-
methyl-l,2-epoxyhexane and 2-methylcyclohexanone yielded a relative response
for these two compounds of 2.43:1 at 296 ± 2°K and 1.17:1 at 423 ± 1°K.
This deviation from unity of the relative responses for gas sampling probably
results from absorption of the more polar compounds on the stainless steel
components of the gas sampling system. When the relative amounts of 1-methyl-
1,2-epoxycyclohexane and 2-methylcyclohexanone produced at 400 torr total
pressure and 296 ± 2°K were corrected for the relative response factor of
the gas sampling system, they were within 2% of that obtained by direct liquid
injection of the product mixture collected from the flow system at 735 ± 10
torr total pressure and 296 ± 2°K. Responses for gas sampling of the other
products at 296°K were then derived by compariosn of the direct liquid in-
jection of the flow study products with the gas phase sampling of the reaction
carried out at 400 torr total pressure, assuming the product distribution to
be identical.
Gas chromatographic calibration factors for samples from the gas phase
system at 423 ± 1°K were based on the fact that only a 17% difference in re-
sponse between 1-methyl-l , 2-epoxycyclohexane and 2-methylcyclohexanone was
observed. Responses for the other products were derived by interpolation
based on their relative retention times. A maximum error of ±10% in the
product yields was expected in this case.
38
-------
Results and Discussion
0(3P) + Toluene. This reaction was studied extensively at 373 ± 2°K
over the pressure range of 80-400 torr of N20, with toluene pressures ranging
from 2-10 torr. The reaction products observed were CO, o-cresol, p-cresol,
phenol, an unknown and a reddish-yellow tar. The product yields, relative
to the N2 yield, are given in Table 3.1 along with the yields of the volatile
addition products expressed as a percentage of the total volatile addition
products. No benzaldehyde, benzyl alcohol or biphenyl were detected. In
addition, small amounts of CH^ and C2H6 were observed in experiments carried
out at 296 ± 2°K and 423 ± 1°K, with the CH^/% ratio being determined as 0.01
at 296 ± 2°K.
Room temperature experiments, monitoring the reaction by infrared absorp-
tion spectroscopy, showed the production of CO and CHij and small amounts of
a very broad carbonyl band at -1725 cm . The latter band remained after
evacuation of the cell suggesting that it was associated with the tar. An
NMR spectrum of the reddish-yellow tar material in acetone-de/CDCls showed
no evidence of aromatic protons. Instead a very broad signal from methine
protons and from protons in methyl groups were observed, suggesting a mixture
of materials.
Table 3.1 also gives the product data obtained at room temperature by
Jones and Cvetanovic and by Grovenstein and Mosher. It can be seen that
there is general, agreement on the products obtained from this reaction, and
that the volatile addition products account for only a small fraction (15-
25%) of the 0(3P) atoms consumed. However, there are differences in the
relative amounts of the individual phenolic compounds observed, with Grovenstein
1 4
and Mosher finding more m-cresol than either Jones and Cvetanovic or the
present work. Also, the present observation of phenol is in agreement with
1 4
Grovenstein and Mosher, while Jones and Cvetanovic did not detect this
product. In all cases extensive tar formation was observed, and the present
finding of the lack of aromaticity in the tar is in agreement with the work
of Boocock and Cvetanovic5 on the reaction of 0(3P) atoms with benzene. The
detection of CH^ and C2H6 in the present work suggests the presence of CH3
radicals which can abstract and H atom to form CHi^ or recombine to form C2H6.
The available evidence suggests that the initial reaction step is that
of addition of an oxygen atom to the aromatic ring, predominately in the
ortho position, followed by isomerization to phenolic compounds or ring
39
-------
Table 3.1
Product yields observed from the reaction of 0(3P) atoms with toluene
Product A
Phenol
o-Cresol
m-Cresol
p-Cresol
CO
H20
CH4
C2H5
This work
A/N2 % Yieldd)
0.02 8.7 ± 1.9
0.20 80.5 ± 1.7
i 0.01e) I 4.7 ± 0.7e)
1 1
0.140 ± 0.018
0.018)
0.0058)
Ref 4b) Ref lc)
A/N2 % Yieldd) % Yieldd)
8.0
0.12 78 58.8
If) if) 15'3
0.034 ' \ 22*'
« ' 17.9
0.093(100 torr)
-0.059(680 torr)
0.07
Ptotal = 80~*°° torr N2°J T - 373 ± 2°K. Includes 6.1% of an unknown.
ftotal m 385 ± 5 torr N2°; T = 298°K' No Pressure dependence of the cresol yields
noted on pressure from 100-680 torr N20.
Ptotal = 74° ± 4 t0rr Nz0' T = 3°3 ± 10K" Volatile addition products analyzed only.
d)
e)
f)
Yield of the observed volatile addition products = 100%.
m-, p-cresol not separated; mass spectra shown m-cresol <10% of the p-cresol.
m-, p-cresol not separated; infrared spectra show m-cresol £4-5% of total cresol
yield.
8) At 296 t 2eK.
40
-------
cleavage contraction to form a variety of highly reactive olefinic species
which will consume 0(3P) atoms in secondary reactions, ultimately producing
a tar or polymer containing few or no aromatic protons. Thus at 373°K, the
rate constant for the reaction of 0(3P) atoms with toluene is 1.25 x io8 liter
-1 -I6
mole sec while it is expected that the rate constants for the reaction
of 0(3P) atoms with the olefinic product species will be ~1 x 1Q10 liter
-1 -1 7-10
mole sec . Also, it is likely that the reaction will undergo fewer
secondary reactions at higher temperatures where the rate constant for 0(3P)
atoms with toluene will approach those for 0(3P) atoms with the olefinic;
products, as may be indicated by the data in Table 3.1. A possible reaction
sequence is shown below for attack at the ortho position:
0(3P)
CH3
A/"0
LJ
CH3 CHO
X
+ CH2 = CH-CH = CH-CH = CH-CHO
+ a variety of other non-aromatic unsaturated hydrocarbons
The formation of phenol presumably proceeds by addition of an 0(3P) atom at
the 1-position followed by an intermolecular methyl migration and subsequent
H atom abstraction by the phenoxy radical to form phenol and by the methyl
radical to form CHi,..
Q(3P) + Methylcyclohexene. The products from this reaction were studied
at 296 ± 2°K over the range 55-400 torr total pressure of N20 and at 423 ± 1°K
41
-------
at 170 torr total pressure of N20 with typically 1-3 torr of 1-methylcyclo-
hexene. As noted above, the product peaks observed on the $,3-oxydipropioni-
trile column under these conditions were identical with those obtained from
the flow system at 296 ± 2°K and 735 ± 10 torr total pressure of N20. The
six products collected by preparatory gas chromatography from the flow system
study were all shown by combined gas chromatography-mass spectrometry to be
isomers of molecular formula C7Hi20« NMR and infrared spectra were obtained
for all of these products and £wo of the major ones were found to have NMR,
infrared and mass spectra and gas chromatographic retention times identical
to those of 2-methylcyclohexanone and l-methyl-l,2-epoxycyclohexane. The NMR,
infrared and mass spectra of the other collected unknowns allowed them to be
identified as 1-methylcyclopentenecarbaldehyde, methylcyclopentylketone, 2-
methyl-5-hexenal and a product peak which was identified as a mixture of 1-
hepten-6-one and 2-methyl-2-hexenal, the latter probably a mixture of cis
and trans-, all of which are in accord with the mechanism shown below. The
mass spectrum of the seventh product peak was obtained using combined gas
chromatographic-mass spectrometric analysis, and its possible identity is
discussed below. CO was also detected as a reaction product.
Table 3.2 shows the product/N2 ratios obtained at 296 ± 2°K and 423 ± 1°K,
while Figure 3.1 shows some of the product/N2 ratios at 296 ± 2°K plotted
against the N20 pressure. It can be seen that all of the addition products
show a slight increase in yield with increasing total pressure as expected
if they arise from a highly energetic intermediate which can be quenched to
yield stable addition products. On the other hand, the CO yield decreases
with increasing pressure, as expected for a fragmentation product formed
from an energetic intermediate. The data at 423 ± 1°K are very similar to
those obtained at 296 ± 2°K showing that the effect of temperature on the
system to be very slight.
By analogy with the very extensive work of Cvetanovic and co-workers '
on the reaction mechanism for the reactions of 0(3P) atoms with a variety
of olefinic systems, the following simplified reaction scheme can account
for all the addition products observed:
42
-------
Table 3.2
Product yields from the reaction of 0(3P) atoms with 1-methylcyclohexene (N2 - 1)
p
T°K total MCXO MCXN MCPA MCPN MHA Xj X2 CO SO(3P)
torr
296 ±2 55 0.315 0.070 0.050 0.030 0.036 0.040 0.004 0.020 0.565
±±±±±±±±±
0.063 0.021 0.010 0.009 0.006 0.010 0.0005 0.005 0.125
296 ± 2 115 0.306 0.066 0.036 0.029 0.037 0.042 0.003 0.014 0.533
±±±±±±±±±
0.017 0.002 0.002 0.004 0.002 0.004 0.0002 0.004 0.035
296 ± 2 210 0.355 0.090 0.044 0.036 0.051 0.056 0.006 0.010 0.648
±±±±±±±±±
0.028 0.005 0.005 0.001 0.007 0.003 0.0007 0.002 0.052
296 ± 2 400 0.369 0.122 0.062 0.042 0.053 0.070 0.007 0.007 0.732
±±±±±±±± +
0.015 0.009 0.003 0.004 0.004 0.004 0.0003 0.002 0.014
423 ± 1 170 0.340 0.070 0.038 0.032 0.042 0.053 0.005 0.580
±±±±±±± ±
0.015 0.004 0.003 0.001 ( 0.004 0.003 0.0003 0.030
MCXO = l-methyl-l,2-epoxycyclohexane
MCXN = 2-methylcyclohexanone
MCPA = 1-methylcyclopentenecarbaldehyde
MCPN = methylcyclopentylketone
MHA = 2-methyl-5-hexenal
Xi = -60% l-hepten-6-one + -40% 2-methyl-2-hexenal
X2 - unknown, tentatively identified as 3-methyl-2-oxepene (see text)
43
-------
0.5r-
^0.4
en
Q
_i
UJ
en
QL
0.3
1-0.2
o
Z)
Q
n
u-
0
-o—
o
o
A/CHO
0
200 300
TOTAL PRESSURE TORR
400
500
Figure 3.1 Plot of selected product yields (N2 = 1) against total NgO pressure for
the reaction of 0(3P) atoms with 1-methylcyclohexene at 296 ± 2°K.
(Data for 2-methyl-5-hexenal and products Xj and K£ are omitted for
clarity).
-------
CH3 CHO
CH3C-CH2CH2CH2CH2
(C)
CH3COCHCH2CH2CH2CH2
(D)
CHO
I
CH3CHCH2CH2CH = CH2
CH3COCH2CH2CH2CH = CH2
CH3
cis-, trans-" C
/
CHO
CHCH2CH2CH3
l-Methyl-l,2-epoxycyclohexane may be formed from either of the initially
formed biradicals (A) or (B) by ring closure, while 2-methylcyclohexanone
may be formed from (A) by a 1,2-H atom shift or from (B) by 1,2-methyl
migration. 1,2-ring contraction leads to 1-methylcyclopentenecarbaldehyde
and methyl cyclopentyl ketone from (A) and (B) respectively, probably via
either a concerted reaction, previously shown to occur for the reaction of
45
-------
0(3P) atoms with cyclopentene at 77°K, ' or via ring opening a to the
oxygen atom position to give the biradicals (C) and (D) which can then
either recyclize to 1-methylcyclopentenecarbaldehyde and to methyl cyclo-
pentyl ketone respectively, or can isomerize to form the open chain unsatu-
rated aldehydes and ketones observed. Fragmentation, such as to CO, of the
vibrationally or electronically excited products formed will be more impor-
tant at low pressures in the absence of stabilizing collisions and hence the
addition product yields will increase with pressure while the fragmentation
product yields will decrease, as observed.
In this system an estimate of the relative amounts of the biradicals (A)
and (B) formed can be obtained from the yields of 1-methylcyclopentenecarb-
aldehyde and methyl cyclopentyl ketone and also from the yields of the open
chain unsaturated ketones and aldehydes. It can be seen from Table 3.2 that
the biradical (A) is formed -60-65% of the time, in agreement with the work
Cvetanovic and co-workers ' ' which shows that the 0(3P) atom adds pre-
dominately to the less substituted of the two doubly bonded carbon atoms.
Analogous to the observation of dihydropyran from the reaction of 0( P) atoms
with cyclopentene, it is possible that the unknown product, X2, observed
in the present work is the unsaturated ether, 3-methyl-2-oxepene, formed
from the cyclization of an isomeric form of the biradical (C):
CHO CHO'
\!-CH2CH2CH2CH2- < > ^>C-CH2CH2CH2CH2-
CH3 CH3
(C)
This is not inconsistent with the mass spectrum observed for this product,
which shows predominant mass peaks at m/e = 112, 97, 71 and 69. In addition,
the short retention time observed on the $,$'-oxydipropionitrile column is
consistent with the low polarity expected for this compound.
The present reaction system shows great similarity with the reactions
of 0(3P) atoms with cyclohexene (major products 1,2-epoxycyclohexane, cyclo-
46
-------
hexanone and cyclopentenecarbaldehyde) and cyclopentene, the latter being
extensively studied by Cvetanovic, Ring and Doyle. There the major pro-
ducts observed were the analogous 1,2-epoxycyclopentane, cyclopentanone and
cyclobutenecarbaldehyde, together with equal amounts of ethylene and acrolein.
Smaller amounts of ring-opened products identified as 4-pentenal, dyhydropyan
and possibly 2-pentenal were also observed.
Conclusion. From this and previous work ' ' the reactions of 0(3P)
atoms with simple cyclic olefins and with the aromatic hydrocarbons can be
satisfactorily explained in terms of a general mechanism such as shown above.
The difference in products between the cyclic olefins and aromatic hydrocarbons
arises because of the difference in the rate constants for reaction of 0(3P)
atoms with the reactant and with the reaction products. Thus the reaction
of 0(3P) atoms with the aromatic hydrocarbons is slow and produces largely
highly reactive unsaturated products (the rate constants for reaction of
0(3P) atoms with o-cresol have, however, been shown to be only -8 times
1 8
faster than with toluene at temperature). However, for the simple olefins
the initial reaction is very fast and forms largely saturated, and hence
unreactive, products. Thus the extent of secondary reactions are much smaller
in these cases, as observed.
47
-------
REFERENCES
1. E. Grovenstein, Jr. and A. J. Mosher, J. Amer. Chem. Soc., 92, 3810 (1970).
2. H. Hibbert and P. Burt, Org. Sym. Coll., 1, 494 (1944).
3. J. H. Purnell, "Gas Chromatography," Wiley and Sons, New York, (1962).
4. G. R. H. Jones and R, J. Cvetanovic, Can. J. Chem., 39, 2444 (1961).
5. G. Boocock and R. J. Cvetanovic, Can. J. Chem., 39, 2436 (1961).
6. R. Atkinson and J. N. Pitts, Jr., J. Phys. Chem., 7j), 295 (1975).
7. R. J. Cvetanovic, Adv. Photochem., J^ 115 (1963).
8. J. S. Gaffney, R. Atkinson and J. N. Pitts, Jr., J. Amer. Chem. Soc.,
97_, 5049 (1975).
9. J. S. Gaffney, R. Atkinson and J. N. Pitts, Jr., J. Amer. Chem. Soc.,
97_, 6481 (1975).
10. J. T. Herron and R. E, Huie, J. Phys. Chem. Ref. Data, 2., 467 (1973).
11. R. J. Cvetanovic, J. Chem. Phys., 23, 1375 (1955); 25, 376 (1956).
12. R. J. Cvetanovic, Can. J. Chem., 36, 623 (1958).
13. S. Sato and R. J. Cvetanovic, Can. J. Chem., 36, 279, 970, 1668 (1958);
37_, 953 (1959).
14. R. J. Cvetanovic and L. C. Doyle, Can. J. Chem., 38, 2187 (1960).
15. R. J. Cvetanovic, J. Phys. Chem., 74, 2730 (1970).
16. R. J. Cvetanovic, D. F. Ring and L. C. Doyle, J. Phys. Chem., 75^ 3056
(1971).
17. M. D. Scheer and R. Klein, J. Phys. Chem., 7±, 2732 (1970).
18. R. Atkinson and J. N. Pitts, Jr., J. Phys. Chem., 79_» 541 (1975).
48
-------
4. A LONG-PATH INFRARED SPECTROSCOPIC STUDY OF THE REACTION
OF METHYLPEROXY FREE RADICALS WITH NITRIC OXIDE
The reaction of alkylperoxy radicals with nitric oxide is generally
1-3
assumed to proceed by the oxidation of nitric oxide to nitrogen dioxide
with formation of an alkoxy radical:
R02 + NO —> RO + N02 (1)
Reaction (1) is believed to be an important route for oxidizing NO to N02 in
photochemical smog. In addition, the alkoxy radical, RO, may react further
4
in the atmosphere to produce H02 which also converts NO to N02. However, a
mass spectrometric study indicated that reaction (1) with R = CH3 does not
occur as written but rather proceeds as follows:
50-65%
CH302 + NO
50-35%
CH302NO (2a)
-> HCHO + HONO (2b)
Secondary reactions of the HCHO and of the CH302NO adduct with 02 were postu-
lated to form methyl nitrate and formic acid, the observed products. More
recent studies suggest that the reaction of NO with CHs02 occurs via
reaction (1) approximately 79 ± 8% of the time.
In this work a reinvestigation of this important reaction has been
carried out by long-path infrared (LPIR) spectroscopy and gas chromatography
(GC), using millitorr reactant concentrations. The results of this study
confirm that methylperoxy radicals do react with nitric oxide as shown in
reaction (1); no evidence for the existence of alternate reaction paths was
found.
Experimental. Azomethane (Merck, Sharp and Dohme) was degassed at
liquid nitrogen temperature and then passed through Ascarite (Arthur H. Thomas,
Co.) to remove traces of carbon dioxide. Nitric oxide (Matheson, >99.0%)
was passed through Linde molecular sieve 13X to remove any N02 and water
present. N02 (Matheson, >99.0%) was distilled over oxygen. The 02 (Liquid
Carbonic, >99.95%), N2 (Matheson, >99.995%), and He (Liquid Carbonic, >99.995%)
were used as received.
49
-------
7 8
Authentic samples of methyl nitrite and methyl nitrate were prepared
by standard procedures. The methyl nitrite was purified by passage through
sodium bicarbonate and Ascarite. The methyl nitrate was purified by GC.
Irradiations were carried out in an FEP Teflon-coated cylindrical vessel
(25 cm diameter, 95 cm length) which housed the White cell optics of the 40 m
path length Perkin-Elmer Model 621 infrared spectrometer. The cell was
evacuated to 6.0 x 10~ torr between runs. Unfiltered light from a medium
pressure 1200 watt mercury arc (Hanovia Model 3A-44V) entered the sample tank
through six 7.5 x 7.5 Cm ports of 6 mm thick window plate which transmitted
light of X>320 nm. Hence the 366 nm mercury line was the primary photolytic
9
wavelength. For several runs the light intensity was reduced with wire
screens in order to determine its effects on the reaction.
Reaction mixtures were prepared by expansion from a calibrated volume
of known pressures (MRS Baratron Type 90, 0-10 torr pressure gauge) of the
reactants into the reaction chamber of the LPIR followed by pressurizing to
760 torr with He or N2- Initial concentrations were: CHa^CHs, 33-369 mtorr,
NO, 25-76 mtorr; 02, 2200-5320 mtorr. These reactant concentrations were
chosen to minimize the thermal oxidation of NO to N0£ ' and the reaction
12 13
of CH3 with NO. ' Dark runs confirmed that the thermal conversion of NO
to N02 was negligible during the short duration of these experiments (-15 min),
The extinction coefficients for CH3N2CH3, CHsONO, CH30N02, and N02 were
determined from Beer's Law studies of measured millitorr concentrations of
each of these compounds in 760 torr of He or N2. During an experiment,
_i
methyl nitrite (CH3ONO) was determined from its absorbance at 800 cm
Nitrogen dioxide (N02) was determined from the total absorbance at 1600 cm
by subtracting out that due to CH3ONO. Methyl nitrate (CH3ON02) was measured
from its absorbance at 1300 cm
In some runs, methyl nitrite, methyl nitrate and other products were
also monitored by GC or combined gas chromatography-mass spectrometry (GC-MS).
Table 4.1 lists the columns, detector and operating conditions used.
A search for HCHO was also carried out by flushing the contents of the
cell through two traps in series containing distilled water and subsequently
14
identifying HCHO by the chromotropic acid test.
Actinometry was done by photolyzing measured concentrations of azomethane
±n 760 torr He and following the rate of nitrogen formation using Column E
of Table 4.1. Since the quantum yield for N2 formation from azomethane is
50
-------
Table 4.1
Columns used in gas chromatographic analysis of the products of the photooxidation
of azomethane in the presence of NO at room temperature
Column
designation
A
B
C
D
E
Column description
3 m x 3 mm 10% Carbo-
wax 600 on 100/120
acid washed firebrick
3 m x 3 mm 10% Carfco-
wax 600 on 100/120
acid washed firebrick
3 m x 3 mm 10% Carbo-
wax 600 on 100/120
acid washed firebrick
3 m x 3 mm 20% XF
1150 cyanosilicone on
80/100 HMDS treated
chromosorb
6 m x 6 mm Linde
molecular seive, 13X
Carrier gas
flow rate
(ml/min)
25
50
50
23
30
Column
temperature
(°C)
23
50
23
23
0
Compounds
identified
CH3ONO,CH3ON02
CH3ONO,CH3ON02
CH3N02
CH3ONO,CH3ON02
CH3ONO
N2, 02
Model &
detector
Varian Aero-
graph Hy-Fy
600, electron
capture de-
tector
Varian Aero-
graph 1400
with flame
ionization
detector
Finnigan
Model 3100
combined gas
chromatograph
mass spec-
trometer
Varian Aero-
graph 1400
with flame
ionization
detector
Per kin-Elmer
900, thermal
conductivity
detector
51
-------
1.0, the rate of its formation is also the rate of light absorption by
azomethane, I . As a check, N2 was also monitored during one experiment with
cl
NO and 02 present and its rate of formation was shown to agree with that
determined in actinometry runs.
Results. Figure 4.1 shows typical infrared spectra of the individual
reactants and products. These spectra agree well with those reported in the
1 f\ 1 ft
literature. The limits of detection of the products and the relevant
extinction coefficients are given in Table 4.2.
Table 4.2
Infrared extinction coefficients3 and limits of detection of some reactants
and products in the photooxidation of azomethane in the presence of NO at
room temperature.
Compound
CH3N2CH3
CH3ONO
N02
CH3ON02
Wave number,
-1
cm
1000
800
1600
1600
1300
Extinction coefficient,
-1 -1
mtorr m
1.46 x
5.63 x
3.03 x
9.15 x
9.58 x
io-5
io-4
io-4
io-4
io-4
Detection limit,
mtorr
-
0.5
-
0.6
0.5
To base 10
Figure 4.1c gives the infrared spectrum of a typical reactant mixture
before and after photolysis to 1% conversion of azomethane. It is seen that
the only detectable products are N02, CH3ONO, and small amounts of CH3ON02.
Extensive analysis by GC (Table 4.1) confirmed the presence of CH3ONO and
CH3ON02 and revealed traces of CH3N02 and one other unidentified product.
Small amounts of HCHO were also detected by the chromotropic acid test.
The observation of these products is in agreement with an earlier LPIR study
of this system, as well as with the more recent GC studies of this reaction.
19
52
-------
WAVELENGTH (microns)
4 _5 6 _7 8 9 10 12
BACKGROUND
CH3ONO (11.4 mtorr)
CH3ON02( 4.2 mtorr)
N02( 8.0 mtorr)
NO (100 mtorr)
AFTER 1%
PHOTOLYSIS
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800
1
WAVENUMBER (cm)
Figure 4.1 (a) Infrared spectrum of the empty cell and
of 11.4 mtorr CHsONO; (b) Infrared spectrum
of: 100 mtorr NO; 8.0 mtorr N02; (c) Infra-
red spectrum before and after 1% photolysis
of 340 mtorr azomethane, 39 mtorr NO, and
3130 mtorr 02.
53
-------
Figure 4.2 shows a typical plot of product formation and NO loss with
time. Both CH3ONO and N02 grow linearly during the initial stages of the
reaction when CH30N02 formation is negligible. At longer reaction times
the N(>2 concentration levels off and the CH30N02 begins to increase non-
linear ly- Finally, at low NO concentrations, the rate of CH3ONO formation
falls, while that of CH30N02 increases substantially.
The low extinction coefficient for NO at 1900 cm" (4 x 10 mtorr m )
precludes its accurate measurement by infrared at the low concentrations
used in these experiments. For example, the point at 65 minutes in Figure 4.2
corresponds to the limit of detection of NO in this system. Hence the NO
data even at short reaction times are only accurate to about ±15%.
The quantum yields of CH3ONO and N02 were calculated as the ratio of
their initial rates of formation to the rate of azomethane loss. Because
q on 71
both N02 and CH3ONO absorb light of A>320 nm, ' ' their quantum yields
were calculated under conditions such that their loss by photolysis was
small. Since the extinction coefficients for N02 and CH3ONO at 366 nm are
1.5 x 102 and 48 1 mol cm respectively, compared to 3 1 mol cm for
9
CH3N2CH3, all quantum yields were determined under conditions such that
[CH3N2CH3] [CH3N2CH3]
[CH3ONO]
Table 4.3 gives, for a series of runs of varying initial reactant con-
centrations and light intensity, (1) the initial rates of product formation
as a ratio to the azomethane concentration and (2) the corresponding quantum
yields. The N02 quantum yield is 1.9 ± 0.3 while that for CH3ONO is 1.7 ±
0.2 where the specified errors represent one standard deviation. An upper
limit for the quantum yield of CH3ON02 is 0.01 from its initial rate of
formation. Typically, CH3ON02 was not detected by GC (detection limit 4 x
-3
10 mtorr) in the first 3 minutes of photolysis.
Discussion. The following reactions describe the photooxidation of
azomethane in the presence of the nitric oxide under these conditions:
CH3N2CH3 A>32° ^ > 2CH3 + N2 *N2 = l.O15 (3)
M 1 I13
CH3 + 02 -£L-> CH302 k = (3.1 ± 0.3) x 108 1 mol sec (4)
54
-------
o
o
h-
<
oc
UJ
o
z
o
o
60 80
TIME (min.)
100
120 140
Figure 4.2 Typical time dependence of CH3ONO, CH3ON02, N02 and NO
in the photooxidation of azomethane in the presence of
NO. Initial conditions, azomethane 132 tntorr; 02 2830
tntorr; NO 38 mtorr.
55
-------
Table 4.3
Reactant concentrations, rates of product formation and quantum yields in the photooxidation
of azomethane in the presence of NO at room temperature.
Initial react ant concentrations
[CH3N2CH3J
mtorr
369
340
322
289
253
217
204
149
132
97
33
241
268
239
270
336
250
295
275
163
153
172
349
314
239
3 1 mtorr =1.3
b RCH3ONO and *i
[N0]o
mtorr
37
39
40
39
36
30
37
36
38
36
25
67
52
44
30
37
37
39
38
57
76
56
41
41
40
parts
._ are
2
•
,rt , CH3ONO x 1Q3D
1 Vjo 1
°
mtorr
Ia = (7.0
3040
3130
3340
3040
3040
3040
3040
3040
2830
3040
3040
3040
3040
3040
3040
5320
3800
2280
2200
3040
3040
3040
\ -
-------
CH3 + NO -£_> CH3NO k = (2.4 ± 0.2) x 1Q<* i mo!'1 sec~l (5)
CH3 + N02 —> CH3N02
CH302 + NO —> CH30 + N02
CH30 + NO -
—> CH3ONO
(8)
—> ECHO + HNO
GH30 + N02 -
—> CH3ON02
(9)
—> HCHO + HONO
CH30 4- 02 —> HCHO + H02 (10)
Under these experimental conditions, removal of CH3 by reaction with
12 13
NO ' will occur <10% of the time, particularly since the NO concentration
is decreasing during the reaction. At low azomethane conversions (-1%) the
contribution of reaction (6) to the removal of CH3 radicals is also small,
although it does account for the trace quantities of CH3N02 observed by GC.
The reaction (8) of methoxy radicals with NO can occur either by com-
22-25
bination (step a) or by abstraction (step b) with k ,/k = 0.143.
Qu 8
Similarly, reaction (9), which is minimal at short reaction times, can also
23
occur by combination or abstraction with k h/k = 0.08. Estimates of the
rate constant ratio r^ range from 1.223 to 2.9.26'27 Hence reaction (9)
ks
will become competitive with reaction (8) at longer reaction times when the
23 25 28
NO concentration approaches that of N02. Reaction (10) is too slow ' '
(k10 = 1.6 x IQ3 1 mol"1 sec"1) to be significant.
The experimental results are consistent with the above scheme in that
(1) the major products at short photolysis times are CHaONO and N02, (2) N02
accumulates linearly with time until its removal by both reaction (9) and
photolysis becomes competitive with its rate of formation, and (3) small
amounts of HCHO and CH3N02 are observed. At longer reaction times, reaction
(9) becomes competitive with reaction (8) and hence the CH3ON02 concentration
increases and N02 decreases. Simultaneously, the rate of CH3ONO formation
57
-------
falls. HCHO was not detected by IR since its detection limit in this system
is ~3 mtorr and the maximum yield of HCHO from reaction (8b) in any of these
runs was 2.3 mtorr. Under these conditions, loss of HCHO by photolysis is
9
negligible.
23 29
Estimates of the rates of reaction of HNO with itself and with 02
indicate that these rates are sufficiently slow that HNO should accumulate
in this system. While the IR extinction coefficient for HNO is unknown, it
is not surprising that the low yield is below the present detection limit.
The same is true of the small yields of CH3NO anticipated from reaction (5).
At these low conversions, HONO is also expected to be undetected.
Both our experimental results and the kinetic data available in the
literature suggest that for the purposes of kinetic analysis, at short
photolysis times where the loss of N02 is negligible, reactions (5), (6),
(9) and (10) may be neglected. Using a simplified scheme consisting of the
remaining reactions (3), (4), (7) and (8), the following relations can be
derived:
d[N02]
N02 = dt = 2.0 Ia[CH3N2CH3]o (I)
d[CH3ONO]
KCH3ONO = - ^ - = 1.71 Ia[CH3N2CH3]o (II)
where I is the rate of light absorption by azomethane in min and [CH3N2CH3]
a o
is the initial azomethane concentration. Hence R^ /[CH3N2CH3] and
R „ -^Mn/[CH3N2CH3] should be independent of initial reactant concentrations,
' O
as found (Table 4.3). From the average values of these ratios, I = (6.5 ±
-L _i _4 a _i
1.0) * 10 min from N02 formation and I = (7-0 ± 0.7) x 10 min from
a -4 -1
CH3ONO formation, within 7% of the value (7.0 ± 0.7) x 10 min determined
in separate actinometry experiments. This scheme also predicts quantum yields
of 2.0 and 1.7 for N02 and CH3ONO respectively, in excellent agreement with
the results given in Table 4.3.
The carbon and nitrogen mass balances in the initial stages of the
reaction are reasonably good. For example, in Figure 4.2, 1.9 ± 0.20 mtorr
N02 and 1.8 ± 0.09 mtorr CH3ONO have been produced at 10 minutes while 4.7 ±
0.7 mtorr of NO and 0.92 ±0.1 mtorr of CH3N2CH3 have reacted. According to
the above mechanism, 3.7 ± 0.6 mtorr of NO is predicted to react during this
58
-------
time interval. Since k /k = 0.86, the carbon and nitrogen balances are
ba o
110 ± 20% and 85 ± 30%, respectively. The particularly large uncertainty
in the nitrogen balance is due to the inaccuracies in measuring NO by infra-
red spectroscopy as discussed above, which precludes accurate mass balance
computations in the present system. At long reaction times, however, it
appears that the nitrogen balance may be somewhat poorer. For example, at
140 minutes the percentages of carbon and nitrogen accounted for in the
observed products are 103 ± 20% and 71 ± 30%, respectively. This is likely
due to secondary reactions producing nitrogen compounds which are not easily
detected by LPIR and GC or which may be adsorbed on the walls. Since this
study was concerned only with the mechanism at short reaction times where
all the carbon and nitrogen could be accounted for, no further investigation
of the poor nitrogen balance at longer photolysis times were done.
Our observations of the time dependence of methyl nitrate formation are
in agreement with those of earlier investigators in that it is formed with
an "induction period" which depends on the rate of light absorption. This
is understandable since according to the above mechanism,
= -4 I [CH3N2CH3] (III)
Higher rates of light absorption correspond to higher rates of NO loss, hence
reaction (9) forming CH3ON02, will become competitive with reaction (8) at
shorter reaction times. In addition, it appears that with the rate of loss
of NO given in equation (III), little NO would remain after the induction
periods found by these investigators and hence secondary reactions might
well then produce HCOOH. It is puzzling, however, that CH3ONO was not detected
in their studies in the early stages of the reaction, either by mass spec-
trometry or gas chromatography.
In conclusion, our results confirm that under these experimental conditions,
methylperoxy radicals oxidize NO to N02. We find no evidence of alternate
modes of reaction occurring between these two species.
59
-------
REFERENCES
1. H. Niki, E. E. Daby and B. Weinstqck, Adv. Chem. Ser. No. 113, 16 (1971).
2. A. P. Altshuller and J. J. Bufalinl, Photochem. Photobiol., i, 97 (1965).
3. T. A. Hecht and J. A. Seinfeld, Environ. Sci. Technol., j>, 47 (1972).
4. J. Heicklen, K. Westberg and N. Cohen, Center for Air Environmental
Studies Report No. 115-69, The Pennsylvania State University (1969).
5. C. W. Spicer, A. Villa, H. A, Wiebe and J. Heicklen, J. Amer. Chem. Soc.,
15, 13 (1973).
6. R.'simonaitis and J. Heicklen, J. Phys. Chem., 78, 2417 (1974).
7. W. H. Hartung and F. Crossley, Org. Syn. Coll., ,2, 363 (1943).
8. A. P. Black and F. H. Babers, Org. Syn. Coll., 2_, 412 (1943).
9. J. G. Calvert and J. N. Pitts, Jr., "Photochemisty," John Wiley and
Sons, Inc., New York, N. Y., 1966, and references therein.
10. J. Heicklen and N. Cohen, Advan. Photochem., _5_, 157 (1968).
11. D. H. Stedman and H. Niki, Environ. Sci. Technol., T_, 735 (1973).
12. N. fiasco, D. G. L. James and R. D, Suart, Int. J. Chem. Kinet., 2^, 215
(1970).
13. N. Basco, D. G. L. James and F. C. James, Int. J. Chem. Kinet., 4_, 129
(1972).
14. P. W. West an4 B. Sen, Z. Analyt, Chem., 153, 12 (1956) and references
therein.
15. M. H, Jones and E. W. R, Steacie, J. Chem. Phys., 21, 1018 (1953).
16. R. H. Pierson, A. N. Fletcher and E. Gantz, Analyt. Chem., 28, 1218
(1956). ~~~
17. P. Klaboe, D. Jones an4 E. R. Lippincott, Speptrochimica Acta, 23A,
2957 (1967).
18. E. R. Stephens and M. A. Price, IR spectrum of CH30N02, personal
communication, 1974.
19. P. L. Hanst and J. G. Calvert, J. Phys, Chem., 63, 2071 (1959).
20, G. R. McMillan, J. Kumari and D. L. Snyder, "Chemical Reactions in
Urban Atmospheres," Ed. C. S. Tuesday, Elsevier Publishing Co., pp. 35-
44 (1969,).
60
-------
REFERENCES (cont.)
21. I. T. N. Jones and K. D. Bayes, J. Chem. Phys., ,59, 4836 (1973) and
references therein.
22. H. A. Wiebe and J. Heicklen, J. Amer. Chem. Soc., ^5, 1 (1973).
23. H. A. Wiebe, A. Villa, T. M. Hellman and J. Heicklen, J. Amer. Chem.
Soc., 95, 7 (1973).
24. G. E. McGraw and H. S. Johnston, Int. J. Chem. Kinet., _!, 89 (1969).
25. W. Glasson, Abstracts, 167th National Meeting of the American Chemical
Society, Los Angeles, CA, March 31 - April 5, 1974, Phys. No. 41.
26. G. Baker and R. Shaw, J. Chem. Soc., 6965 (1965).
27. P. Gray, R. Shaw and J. C. J. Thynne, Progress in Reaction Kinetics,
±, 63 (1967).
28. J. Heicklen, Adv. Chem. Ser.^ No. 76, 23 (1968).
29. K. L. Demerjian, J. A. Kerr and J. G. Calvert, Advan. Environ. Sci. &
Technol., _4, 1 (1973).
30. L. H. Jones, R. M. Badger and G. E. Moore, J. Chem. Phys., 19, 1599
(1951).
61
-------
5. RATE CONSTANTS FOR THE GAS PHASE REACTION OF PEROXYACETYL NITRATE
WITH SELECTED ATMOSPHERIC CONSTITUENTS
The gas phase reactions of PAN with several compounds typically present
in photochemically polluted urban atmospheres have been studied with a view
to determining the rate constants for these reactions in order to assess the
importance of these processes for the removal of PAN in polluted urban atmo-
spheres. In addition, a product analysis was carried out for the reaction
of PAN with NO.
Experimental. Most reactions were carried out in a 65 liter FEP Teflon-
lined aluminum tank which enclosed the 40 meter pathlength multiple reflection
white cell optics of a Perkin-Elmer 621 spectrophotometer. The reaction tank
-4
could be routinely evacuated by an oil diffusion pump to £7 x 10 torr.
Reactants and products were monitored in situ by long-path infrared (LPIR)
spectroscopy using experimentally determined or literature extinction coef-
ficients. Thus PAN was monitored by its infrared absorption bands at 5.76
and 12.61 u using the extinction coefficients determined by Stephens. The
reaction of PAN with S02 was studied in a 23 liter cylindrical Pyrex cell
connected via a Carle gas sampling valve to a gas chromatograph fitted with
2-4
an electron capture detector. In this case PAN was quantitatively analyzed
on a 22" x 1/8" Teflon column packed with Carbowax 600 on 60/80 firebrick
with an N2 flow rate of 30 cm3 min"1 at 296 ± 1°K.
Reactant mixtures were introduced into either reaction vessel from a
conventional mercury-free high vacuum gas handling system. Reactant pressures
in calibrated volumes of this gas handling rack were measured by either an
MKS Baratron gauge (0-10 torr) or by a 0-800 torr Wallace and Tiernan FA 160
absolute pressure gauge.
A product analysis was undertaken for the reaction of PAN with NO using
a Finhigan 3100D combined gas chromatograph-mass spectrometer with a 10 ft x
1/8" stainless steel column packed with Carbowax 600 on firebrick, and also
by conventional gas chromatography using an electron capture detector equipped
with a 20 ft x 1/8" stainless steel column of Carbowax 400 on 60/80 firebrick.
Peroxyacetyl nitrate (1000 ppm in tanks pressurized by N2) was used as
received. CO (>99.5% purity) was passed over glass beads at 77°K in order to
remove any traces of iron carbonyl present, while NO (>99.0% purity) was passed
through activated Linde Molecular Sieve 13X to remove any N0£ and H20 present.
62
-------
Isobutene (>99.8%), S02 (>99.98%), NH3 (>99.99%), CH3CHO (>99.3%) and N02
(>99.5%) were subjected to several freeze-thaw vacuum distillations prior
to their use. Approximately 2% ozone in oxygen was produced by passing
oxygen (>99.99% purity) through a Welsbach Model T-408 ozonizer and its
concentration was monitored in situ by LPIR spectroscopy using the 9.5 y
adsorption band and the published absorptivity. In the reactions involving
water vapor, portions of room air at ~50% relative humidity were admitted
to the reaction vessel which were subsequently diluted with a synthetic air
mixture to atmospheric pressure and the resulting H20 concentration was
measured by LPIR spectroscopy at 5.02 y. LPIR spectroscopic analysis showed
that the only other significant contaminant in the H20 air mixture was C02.
In all cases the pressurizing gas was a synthetic mixture of N2 (>99.995%)
plus 02 ( 99.99%) or N2 (>99.995%). All rate constant determinations were
carried out at 296 ± 1°K with the added reactant in large excess of the
initial PAN concentration, so that the reactant concentration remained
essentially constant throughout the run. PAN concentrations were monitored
by LPIR spectroscopy and/or electron capture gas chromatography as a function
of time after the reagents were admitted to the reaction vessel.
Results. PAN decays were routinely determined in the absence of any
added reactant. The decays were observed to be first order in PAN and PAN
half-lives were typically of the order of 35-40 hours. Thus, for example,
0.045 ppm of PAN in the 23 liter Pyrex cell had a decay rate of kx - (3.0 ±
0.3) x 10~* min"1 while 8.0 ppm of PAN in the 65 liter Teflon-coated aluminum
-4 -1
tank had a decay rate of kj = (2.45 ± 0.3) x 10 min .
S02> CH3CHO, H20, iso-C^Hg, CO, N02> NH3, 03
With these reactants in large excess (except for NH3 — see below) pseudo-
first order plots of In [PAN] against time were observed to be linear with
slopes which increased linearly with the reactant concentration. Hence
[PAN]
In
-------
PAN + reactant —> products (2)
and kj is the first order rate constant for the loss of PAN in the absence
of other reactants. Plots of the PAN decay rate, defined as (t-tQ) In [PAN]Q/
[PAN] , against the reactant concentration are shown in Figure 5.1 for the
reactants H20 and 03. Linear plots were obtained for the reactants H20, N02
and CHsCHO, although a large degree of scatter was observed for N02 as the
reactant and the rate constants k2 derived from the slopes of these plots
are given iii Table 5.1 along with the initial concentrations of PAN and
reactant. For CO, S02 and isobutene only upper limits for k2 could be
obtained from the experimental data. For 03 a plot of PAN decay rate against
03 concentration was observed to be curved (Figure 5.1) and the value of k2
given in Table 5.1 is that obtained from the initial slope as the 03 concen-
tration approaches zero. In the case of NH3, the ammonia concentration was
observed to decay rapidly either in the presence or absence of PAN presumably
due to wall absorption. Hence the PAN and NHg concentrations were simulta-
neously monitored and k2 was obtained directly from the expression
-d[PAN]/dt = k2 [PAN][NH3] (II)
NO
With NO in excess of PAN in air diluent, pseudo-first order plots of
In [PAN] against time showed curvature for reaction times typically exceeding
one half-life of PAN, as shown in Figure 5.2. However, PAN decay rates,
defined as (t-tQ)~ In [PAN]J[PAN], derived from the initial linear portions
of such plots were observed to be independent of the initial NO and PAN
concentrations as shown in Table 5.2 and Figure 5.2 and were approximately a
factor of 60 higher than the PAN decay rates in the absence of NO. Hence
the reaction initially obeys the rate expression
-d[PAN]/dt - k , [PAN] (III)
obs
with the PAN decay rate being identified with k , .
obs
Changing the diluent from air to N2 or the addition of -12000 ppm H20
(-50% relative humidity) caused no change in ko|jS within the likely experi-
mental errors, although a much larger degree of scatter in the data was
64
-------
200
400
600
1
0
1 1
4
L~OJ rr-:-
8
i 1
12
i
X 103
[H20] ppm
Figure 5.1 Plots of d ln[PAN]/dt against reactant concentration
for 03 (initial PAN concentrations 5.1-6.4 ppm (0),
3.0 ppm (0), 9.7 ppm (§)) and H20 (A, initial PAN
concentration 7.6-8.3 ppm).
65
-------
Table 5.1
Experimental conditions used and the rate constants 1^2 obtained for the
reaction of PAN with S02, CO, N02, NHa, H20, 03, isobutene and acetaldehyde.
Reactant
S02
iso-C.Hs
CO
CH3CHO
N02
H20
NH3
Reactant concentration
ppm
1260 - 3540
51
232
83 - 312
26 - 52
5900 - 12,000
13 - 660
28 - 50
Initial PAN ^ _1 a)
Concentration ppm k2 ppm min
0.042 <2 x 10"8
5.8 <1 x 10"6
11.0 <2 x 10"7
3.9 - 5.1 (1.1 ± 0.2) x
3.0 - 9.3 (2.6 ± 1.2) x
7.6 - 8.3 (3.3 ± 0.5) x
s a)
2.9 - 9.6 -8 x 10
4.3 - 8.4 (2.6 ± 1.0) x
io-6
io-6
10"8
IO-5
a)
*
b)
Indicated error limits are the estimated overall errors which include
least square standard deviations of plots of equation I and the estimated
accuracy limits of other parameters such as PAN and reactant concentrations.
Derived from the initial slope ([03] -> 0) of Figure 1.
66
-------
N0 = 26ppm
N0=93ppm
40 80
TIME (minutes)
Figure 5.2 Plots of ln[PAN] against time for the reaction of PAN
with NO in air diluent.
67
-------
Table 5.2
Initial conditions and observed first
reaction of PAN with NO.
Diluent [PAN] ppm
Air 4.1
4.2
4.3
5.1
6.6
7-6
7.9
17.6
Air + 12000 ppm fl20 3.2-8.2
N2 2.8
3.6
3.8
4.0
4.1
4.2
4.3
order rate
[NO] ppm
52
26
109
50
98
50
93
211
46-202
50
50
50
25
75
101
101
constants, k . , for the
1 a)
102 x k , min
obs
1.9 ± 0.2
1.6 ± 0.1
1.8 ± 0.1
1.8 ± 0.1
1.4 ± 0.1
1.6 ± 0.1
1.5 ± 0.1
1.3 ± 0.1
Mean 1.6 ± 0.2
Mean 1.5 ± 0.6 (5 runs)
1.9 ± 0.2
1.9 ± 0.1
1.8 ± 0.1
200 ± 0.1
2.0 ± 0.2
2.3 ± 0.1
2.3 ± 0.1
Mean 2.0 ± 0.2
Error limits are least square standard deviations.
68
-------
observed for air with 12000 ppm H20 diluent. The overall value of k
0 obs
determined from the data in Table 5.2 is k ^ = (1.7 ± 0.5) x 10 min"1
obs '
where the indicated error is the estimated overall error limit. This value
is in good agreement with kQbs = 2.06 x io~2 min"1 determined by Schuck et
al. in a 20-liter stirred flow reactor.
A product analysis of the reaction of PAN with NO was carried out with
N2 as diluent, using a combined gas chromatograph-mass spectrometer and infra-
red techniques. The observed products were N02, C02, CH3ONO, CH3ON02, CH3NO
and CH3N02. The yields of N02, C02, CH3ONO and CH3ON02 relative to the loss
in the PAN concentration were determined to be 1.7 ± 0.3, 0.54 ± 0.05, 0.02 ±
0.01 and 0.08 ± 0.02, respectively. CH3NO and CH3N02 were observed to be
present in small amounts (<0.05 of the PAN consumed).
Because of the N0-N02•thermal conversion in air diluent, no quantitative
product analysis study was carried out in this case. However, the same major
products (N02, C02, CHONO and CH3ON02) were observed with the C02 yield being
similar (AC02/APAN = 0.5 ± 0.1) to that observed for PAN + NO in N2 diluent.
Discussion. The decay of PAN in the absence of added reactant, although
first order in PAN, could be either a homogeneous unimolecular loss process
or heterogeneous involving wall reactions. Although the decay rates k^ were
observed to be similar in either the 23-liter Pyrex cell or the 66-liter
Teflon-coated vessel, no firm conclusions as to the mode of PAN decomposition
can be drawn from this data alone.
For the reaction of PAN with NO, the rate law of -d[PAN]/dt = k0bs[pAN],
determined in N2 or air diluents, and the value of kdbs obtained are in agree-
ment with the work of Schuck et al., as is the observation of curvature in
the ln[PAN] versus time plots. A mechanism which can account for the kinetics
of this reaction as well as the products observed is as follows:
0 1 °
CH3C-OON02 . > CH3C-00- + N02
CH3C03 + NO —> CH3C02 + N02 (3)
CH3C02 —> CH3 + C02 (4)
69
-------
CH3 + 02 — > CH302 (5)
CH3 + NO — > CH3NO (6)
CH3 4- N02 — > CH3N02 (7)
8 9
CH302 + NO — > CH30 + N02 (8) '
CH30 + NO — > CH3ONO (9)
CH30 + N02 — > CHON0 (10)
Reactions (5-10) account for the observed products in both air and N2 diluent,
with reactions (6), (7) occurring in N2 but reaction (5) predominately in air.
Using the steady-state approximation for the CH3C03 radical,
-d[PANl k1k3[PAN][NO]
dt ~ kCNOj] + k3[NO] U '
When k3[NO] » k-[N02], then -d[PAN]/dt = ^ [PAN] in agreement with the
initial stages of the reaction, with kj = k0bs' However, when this inequality
no longer holds, then the PAN decay rate decreases as the ratio [N02]/[NO]
increases, as occurs in the late stages of the reaction, and finally, when
k , [N02] » k3[NO], the reaction should obey the rate law
-d[PANl klk3 [PAN] [NO]
dt " k_x [N02]
It is likely that for the other reactants a similar scheme applies, with
reaction (3) being replaced by
CH3C03 + reactant — > products (11)
N02 + reactant — > products (12)
If k_1[N02] » kn [reactant] or kx 2[ reactant ], the observed first order depen-
dence on PAN and reactant concentrations will be observed. However, because
of the slowness of the removal of PAN in the presence of N02, H20, CH3CHO and
NH3, the involvement of heterogeneous processes cannot be excluded. In the
70
-------
case of 03 as reactant, the observed dependence of the PAN decay rate on the
03 concentration (Figure 1) may fit an equation similar to equation (IV) :
-d[PAN]
dt = k.j.CNO;,] + ka[03]
where ka = k^ or ki2- This would account for the first order dependence on
03 concentration at low 63 concentrations, tending to a zero order dependence
at high 03 concentrations.
For atmospheric purposes the half-life, TT,, of PAN is of interest. Table
%
5.3 gives the calculated PAN half-lives using the rate constants obtained in
this work, assuming them to be those for the homogeneous gas phase reactions,
and typical ambient atmosphere concentrations of the reactants studied. For
NO, the minimum PAN half-life of ~0.7 hrs. occurs with NO in excess of N02.
In typically polluted atmospheres the presence of N02 will increase the half-
life of PAN substantially. It can thus be seen that the important atmospheric
loss processes of PAN are by reaction with NO and possibly H20, although the
latter may be heterogeneous in character.
71
-------
Table 5.3
Chemical lifetimes of PAN in a typical Los Angeles polluted atmosphere
Reactant
cob)
iso-CttH8b)
CH3CHOC)
N02d)
so2e)
NH36)
03d>
H20d>
N0f>
Reactant concentration
ppm
10
0.004
0.05
0.1
0.03
0.09
0.2
12,000
PAN lifetime3
TI, hours
>5
>3
2
4.4
>2
5
~7
29
>0.7
x IO3
x 10^
x 10^
x IO1*
x IO7
x IO3
x IO2
a T, = 0.693/k[reactant]
^
W. A. Lonneman, S. L. Kopczynski, P. E. Darley and F. 0. Sutterfield,
Environ. Sci. and Technol., *J, 229 (1974).
c\
' Estimate
Los Angeles Country APCD, Annual Report, 1973.
e)
Characterization of Aerosols in California by Science Center, Rockwell
International (ARE Contract No. 358) Dec« 15, 1973.
See text
72
-------
REFERENCES
1. E. R. Stephens, Anal. Chem., 36, 928 (1964).
2. E. F. Darley, K. A. Kettner and E. R. Stephens, Anal. Chem., 35, 589
(1961).
3. E. R. Stephens and M. A. Price, J. Chem. Ed., 50, 351 (1973).
4. R. G. Smith, R. J. Bryan, M. Feldstein, B. Levachi, F. A. Miller and
E. R. Stephens, Health Laboratory Science, J3, 48 (1971).
5. E. R. Stephens, F. R. Burleson and E. A. Cardiff, J. Air Poll. Contr.
Assoc., 15, 87 (1965).
6. F. L. Hanst, E. R. Stephens, W. E. Scott and R. C. Doerr, Anal. Chem.,
_33, 1113 (1961).
7. E. A. Schuck, E. R. Stephens, M. A. Price and K. R. Darnall, 163rd
American Chemical Society Meeting, Boston, MA, U.S.A. (1972).
8. C. T. Pate, B. J. Finlayson and J. N. Pitts, Jr., J. Amer. Chem. Soc.,
j)6, 6554 (1974).
9. R. Simonaitis and J. Heicklen, J. Phys. Chem., 7£, 2417 (1974).
73
-------
6. CHEMILUMINESPENCE DETECTION OF PAN AND PBzN
Our investigations of the chemical reactivity of peroxyacetyl nitrate
(PAN) have led to the discovery of chemiluminescent reactions of PAN with
tertiary amines which potentially possess the capability of providing a more
versatile and sensitive means of monitoring PAN in the atmosphere than the
present gas chromatographic technique. The usefulness of such reactions as
the basis for monitoring instrumentation has already been realized in the
now standard ethylene-ozone chemiluminescent reaction for measuring ambient
ozone levels. *
Experimental. The reaction vessel consisted of a silvered 1-liter chemi-
luminescence cuvette provided with two gas inlets centered above the optical
window which formed the bottom of the cuvette. Light generated in the reaction
cuvette was chopped mechanically at 330 Hz and the output of the photomultiplier
(E.M.I. 9659 QB) was fed into a lock-in amplifier (Princeton Applied Research,
Model HR8). Cutoff filters were mounted in a wheel arrangement between the
optical window and mechanical chopper (Figure 6.1). Low-resolution spectra
were obtained with a Bausch and Lomb monochromator (Model 33-86-03) located
between the chopper and multiplier phototube. The amplified signals were
displayed on a 100 mV potentlometric Sargent recorder (Model SRG). The mono-
chromator-raultiplier phototube unit was calibrated by use of a standard lamp
(G.E. quartz iodine lamp EPI-1452, 1000 W) and the spectra were corrected for
spectral sensitivity.
Identical spectra were obtained whether the amine was added to the
reaction cuvette in a stream of nitrogen or whether an excess of amine (liquid-
vapor equilibrium) was present in the cuvette. Variable concentrations of
both PAN and ozone were prepared and passed into the cuvette at a constant
flow rate (-10 cc/sec). All experiments were performed at room temperature.
Peroxyacetyl nitrate (1000 ppm, pressurized tanks) was kindly provided
by the University of California Statewide Air Pollution Research Center at
Riverside. Ozone (~2% V/V) was produced by passing oxygen (Matheson, ultra-
high-purity grade) through a Welsbach Model T-408 ozonizer. Triethylamine
(Matheson, Coleman & Bell or Mallinckrodt gave identical results) was checked
for purity by glc (>99.95%) and used without further purification.
Results and Discussion. Various amines (in the gas and solution phase)
were tested for their chemiluminescence efficiency when reacted with PAN (see
74
-------
I
CUVETTE
•TURNABLE FILTER HOLDER
/ /
Ui
HIGH
VOLTAGE
IN
WTT////////////.
FILTER INPUT
\
T.IH-'.I I '
• •' ;•• .'• '.*•'••'•'•'.'••••'.'•.
• • . •., «• •. • ••••...
J L
CHOPPER SYSTEM
COOLED PHOTO-MULTIPLIER
•^SIGNAL TO LOCK-IN AMPLIFIER
Figure 6.1 Chemiluminescent detection unit
-------
Table 6.1). For both the gas and solution phase, triethylamlne was found to
yield the highest light output when reacted with PAN. The emission from the
liquid-phase reaction of 1% triethylamine in acetone with PAN gives constant
Table 6.1
Chemiluminescence efficiency of amines reacting with PAN (1000 ppm) in the
gas phase and in solution.
Compound
Ammonia
Methylamine
Ethylamine
Propylamine
Butylamine
D ime thy lamine
Diethylamine
Dipropylamine
Diisopropylamine
Dibuty lamine
Diisobutylamine
Tr ime thy lamine
Triethylamine
Aniline
N, N-Dimethylaniline
Diphenylamine
Tr ibenzy lamine
Tr iethy lamine
Vapor pressure (torr) 25°C
gas
gas
(in water) saturated gas stream
300.8
53.7
112 (in water)
57.5
saturated
saturated
saturated
18.00
128 (in water)
saturated
liquid phase, 1% in acetone
11 it
ii ii
ii ii
ii it
W
(Et3N in acetone = 100)
0
0
3
1.0
0.7
0
11.5
4.7
2.4
10
4
16
40
0.8
52.5
1.3
18.6
100
reproducible Chemiluminescence for a given PAN concentration with the intensity
immediately dropping to the dark-current level as the flow of PAN is terminated.
76
-------
This is in contrast to the gas phase reaction which produces a long lasting
afterglow. Concentrations of PAN as low as 6 ppb were detected with a signal-
to-noise ratio of about 3. This is nearly as sensitive as present gas chro-
matographic techniques. Improvement of the light detection system may permit
measurement of PAN concentrations as low as 1 ppb or less.
We have recently found that the gas phase reaction of peroxybenzoyl
nitrate (PBzN) with triethylamine is also chemiluminescent. The emission
spectrum using cutoff filters shows a sharp maximum from 590 to 610 nm.
Concentrations of PBzN at least as low as 5 ppb could be detected by this
method.
Corrected low-resolution spectra of the light emissions produced in the
reactions of PAN (1000 ppm) and ozone (20,000 ppm) with triethylamine vapor
were obtained independently and are shown in Figure 6.2 for comparison. The
relative intensity maximum was arbitrarily set at unity. The maxima were
separated by approximately 130 nm.
In control experiments, methylnitrate, ethylnitrate, ethylnitrite, and
N02, all of which are possible contaminants of PAN, were mixed with triethyl-
amine vapor in the reaction cuvette and showed no evidence of luminescence.
In a similar control experiment, the reaction of nitric oxide (1000 ppm) with
triethylamine resulted in a weak luminescence which was enhanced in the
presence of air. However, the emission observed from 10 ppm nitric oxide
was insignificant compared to that produced from 0.4 ppm PAN.
The reaction of PAN with triethylamine produced a slowly decaying after-
glow which lasted for several minutes even at the lowest PAN concentrations
employed. This behavior is contrasted to that of the corresponding ozone
reaction in which the emission rapidly decayed after the ozone flow ceased.
The respective decay curves are illustrated in Figure 6.3.
To distinguish the chemiluminescence induced by the reaction of tri-
ethylamine with concentrations of PAN and ozone found in polluted urban
atmospheres, various combinations of cutoff filters were tested. A 500-nm
cutoff filter (transparent: X <550 nm) in conjunction with a 665-nm cutoff
filter (transparent: X >665 nm) provided relative intensity values character-
istic for both the PAN and ozone induced emissions. In the case of ozone-
triethylamine reaction, a value of I665/I550 <3- was predicted from the
77
-------
oo
1.0
0.9
0.7
UJ
2 0.6
j£j 0.5
< 0.4
Lj' _,
ca 0.3
0.2
O.I
a
400
PAN
500
X (nm)
600
700
Figure 6.2 Emission spectra of chemiluminescence reactions of triethylamine with
PAN and ozone. Relative intensity maximum arbitrarily set at unity.
-------
VO
CO
21
UJ
LJ
_J
UJ
cc
03 flow stop
Et.N afterglow
-------
monochromator spectra shown In Figure 6.2; for the PAN-triethylamine reaction,
a value of Ie65/I500 >3- was predicted. The measured values of I665/I500 were
0.35 and 3.44 for the ozone-amine and the PAN-amine reactions, respectively.
This combination of cutoff filters affords the most appropriate means of
distinguishing the two emissions at low concentrations of PAN and 03 where
the use of a monochromator is not feasible.
To test the cutoff filter technique, the chemiluminescence produced by
a mixture containing a 20-fold excess of ozone to PAN, a ratio not uncommon
in polluted ambient air, was investigated. In this particular case, a value
of Ie65/I500 of 1-35 was found, indicating a much greater chemiluminescence
efficiency for the PAN-amine reaction in contrast to the ozone-amine reaction.
In an experiment designed to measure the chemiluminescence efficiency
of the PAN-amine reaction relative to the corresponding Os-amine reaction,
1.5 ppra of 03 produced the same uncorrected overall integrated light inten-
sity as 0.4 ppm PAN in reaction with the amine. As an approximate correction
for spectral sensitivity of the multiplier phototube over the integrated
emission spectrum, the relative quantum efficiencies at the X values of
the emissions produced by the PAN and 63 reactions, 0.35 and 0.79, respectively,
were applied. From these data, the chemiluminescence efficiency of the PAN
reaction was calculated to be approximately 10 times greater than the corre-
sponding ozone reaction.
The development of an improved chemiluminescence method of detecting
PAN has been attempted. Based on a suggestion by Dr. Hanst of the EPA, a
photowheel-chopper system was built to provide elimination of the ozone-amine
chemiluminescence (X 520 nm) which tails into the PAN-amine emission
max
(X 665 nm). A modulation technique (using a rotating filter wheel with
nicix
quandrants alternately consisting of Corning 2-73 and 3-74/neutral density
filters) was used with a lock-in amplification to provide a d.c. signal for
the 03-amine emission and an a.c. signal for the PAN-amine emission. The
resulting lock-in amplifier response (sensitive to a.c. only) thus gives a
measure of the PAN concentration. With this system the ozone-amine emission
is successfully eliminated but at the cost of a 10-fold reduction in detection
of PAN sensitivity.
80
-------
REFERENCES
1. C. W. Nedergragt, A. Van der Horst and T. Van Duljn, Nature, 206, 87
(1965).
2. J. A. Hodgeson, J. P. Bell, K. A. Rehme, K. J. Krost and R. K. Stevens,
Joint Conference on Sensing of Environmental Pollutants, Palo Alto,
California, November 8-10, 1971, AIAA Paper No. 71-1067.
3. R. G. Smith, R. J. Bryan, M. Feldsteln, B. Levadie, F, A. Miller and
E. R. Stephens, Health Lab. Sci., 8., 48 (1971).
81
-------
7. SOLUTION PHASE REACTIONS OF PAN WITH MOLECULES OF BIOLOGICAL SIGNIFICANCE
The destruction or alteration of a functional group in a molecule of
biological significance is likely to alter its biochemical modes of reaction,
ultimately harming the cell. PAN has already been implicated in the phyto-
pathology of several species and has been shown to be lethal to mice in
2
sufficiently high concentrations. The chemical reactions actually responsible
for these effects are not known.
In order to understand these chemical effects, a study of the reactions
of PAN with a variety of organic molecules (aldehydes, alcohols, amines,
mercaptans and sulfides) containing functionality found in common biological
compounds such as amino acids, proteins, fatty acids, and carbohydrates, was
undertaken. The reactions of PAN in solution have been followed primarily
by NMR spectroscopyj although infrared spectra of the product solution were
frequently used and gas chromatogfaphy was used to follow one reaction.
Aldehydes. PAN oxidizes aldehydes to the corresponding acids according
3
to the following general reaction:
0
RCHO + CH3C-0-0-N02 —> RC02H + CH3C02H + CH3N02 + CH3ONO + C02
In either benzene or carbon tetrachloride, the relative reaction rates for
several aldehydes are:
(CH3)3CCHO > (CH3)2CHCHO > 0CHO = CH3CH2CHO = CH3CHO
Although the final reaction products were not solvent-dependent, the
reaction pathway varied with the solvent. In CCl^, the acid chloride corres-
ponding to the aldehyde was observed as an intermediate which was converted
to the acid. A comparison of products derived from PAN in benzene and carbon-
tetrachloride is given in Table 7.1 for the reaction of PAN with isobutyraldehyde.
There is some evidence that the reaction in CDClg is first order in PAN.
82
-------
Table 7.1
PAN products, % of total, from reaction of isobutyraldehyde
Solvent
CCli^
Benzene
CH3C02H
73
39
CH3ONO
7
12
CH3N02
12
35
CH3C1
8
0
Others
0
14
Alcohols. Secondary alcohols (2-propanol, a-methylbenzyl alcohol, cyclo-
hexanol) react with PAN as follows:
0 OH C
11 t it
CH3C-OON02 + R-CH-R1 —> R-C-R' + CH3C02H + HN03(?)
The reactions were run in three solvents: in acetone-dg, the reaction is rela-
tively clean; in CCl^, one unidentified product occurs in low yield; in benzene,
several unidentified trace products are detected.
Amines. PAN has been found to react with ammonia, primary amines (methyl-
amine, ethylamine, n-propylamine, n-butylamine) and tertiary amines (triethyl-
amine and N,N-dimethylaniline). In CDC13 solution, the reactions with ammonia
and primary amines suggest the following stoichiometry and products (no by-
products were detected):
OHO
it i »
CH3C-00-N02 + N-R « CH3-C-NHR + 02 + HN02
H
The amides were identified by comparison of their spectra with those of inde-
pendently synthesized compounds. HN02 was identified by its characteristic,
highly structured UV absorption spectrum. In the solution phase reaction of
n-propyl amine, oxygen production was observed as a vigorous evolution of gas.
In solution, PAN reacts very rapidly with tertiary amines, the reaction
being complete in one minute when reactant concentrations are both 0.1 M.
The reaction with triethylamine yields one major amine product (80 to 90%),
83
-------
as yet unidentified, along with nitrogen di§&id§ (4§i§§6§4 ly its
lumineseent emission) and acetic aeid. The expected. amifle §Ht4e may al§Q
present in yields o£ 10^ or less,
Her cap tans. The reaction o€ FAM wilh a 4- t© i=»f©34 eseess ef
mereaptan has fcg«n run in deuterochloroferro (CDClg) and §qiieQu§
The number of molss of mereaptan eonsuwed per mole of FAM w§§ 3 to 4 in
and 2 to 3 in acetone-dg. The major products are shown ift the |§ll§wilii
reaction:
CH3COOH02 + CH3iH —J> CHgCOI
Small amounts of methyl methanethiogulfOR»t» C^gSOgiCHf) sad «?§iee§ ©I §§vefai
other compounds were also detected.
If only one equivalent or less @f methyl mgreaptan is reggted wi€h FAN
in C3)C13, the initial products are Ci3§§CH3 m»d CM|iMO» the l&tter tafUiy
disappears, accompanied by a vigorous evolution ef ga§ and the mfpesf^Bet ©f
an NM1 emission peak corresponding to CH|SfCH3. Ultifflaeely, the
converted to CH3S02SCl3 and at l«a§t
-------
REFERENCES
1. 0. C. Taylor, J. Air Pollut. Control Assoc., JL9, 347 (1969).
2. K. I. Campbell, G. L. Clarke, L. 0. Emik and R. L. Plata, Arch. Environ.
Health, 15, 739 (1967).
3. P. H. Wendschuh, C. T. Pate and J. N. Pitts, Jr., Tetrahendron Letters,
3JL, 2931 (1973).
4. P. H. Wendschuh and J. N. Pitts, Jr., unpublished results.
5. P. H. Wendschuh, H. Fuhr, J. S. Gaffney and J. N. Pitts, Jr., J. Chem.
Soc., Comm., 74, (1973).
6. K. R. Darnall, D. M. Hebert, J. M. Hughes, J. W. Wilson and J. N. Pitts, Jr.,
unpublished results.
85
-------
8. THERMAL AND PHOTOCHEMICAL PROCESSES OF NO AND N07 AND THE APPLICATION OF
THE RESULTS TO THE CHEMISTRY OF POLLUTED ATMOSPHERES
Introduction. The N02 catalyzed geometric isomerization of 2-butene
and 2-pentene has been examined over the temperature range 298° to 400°K.
Both kinetic and equilibrium data have been obtained. The combination of
these results with results from thermodynamic calculations permits the esti-
mation of rate constants for the addition of N02 to the double bond of simple
olefins. Comparison of these rate constants to those for the consumption of
olefins by 03 suggests that atmospheric consumption of olefins by processes
initiated by N02 addition to an olefin double bond cannot be significant in
the atmospheric consumption of olefins.
Experimental. cis-2-Butene (Matheson research grade, 0.7% trans-2-butene
impurity), trans-2-butene (Matheson research grade, 0.3% cis-2-butene impurity),
cis-2-pentene (Chemical Samples, 0.8% trans-2-pentene impurity), and trans-2-
pentene (Chemical Samples, 50.1% cis-2-pentene impurity) were thoroughly de-
gassed by bulb-to-bulb distillation, analyzed by glc for hydrocarbon impurities,
and used without further purification. N02 (Baker CP grade) was mixed with
an equal pressure of 02 (Matheson research grade used as received) and allowed
to stand overnight to remove any NO and N20s impurity (N203 J NO + N02; 2NO +
02 ^ 2N02). The N02 was then thoroughly degassed by bulb-to-bulb distillation
and used without further purification.
The all-glass reaction system (total volume 2.7 fc) consisted of a 2.3-fc
bulb with two exit arms connected by a gas circulating loop. Mounted in the
circulating loop was a circulating pump driven by magnetic coupling to an
external motor. The reaction system was housed in a simple oven constructed
of wood and insulating board. Two fans with motors mounted outside the oven
provided efficient circulation of the air within the oven. Oven temperatures
were measured using ice bath referenced thermocouples. The oven had a tempera-
ture range of 25-125°C and a temperature stability of $0.5°C. The reaction
system was serviced by a conventional high vacuum line. Greaseless 0-ring
stopcocks were used throughout the reaction system and the high vacuum line.
Reaction system pressures were measured using an MKS Baratron gauge (0.1-300
torr).
The gaseous reaction mixture was directly sampled using a Carle micro-
i
volume gas sampling valve (sample loop volume -200 y£) connected through a
86
-------
greaseless 0-ring stopcock to the circulating loop of the reaction system.
Glc analyses for cis and trans olefin were performed on a 10% Nad on alumina
column (3 ft, 1/8 in.; 100-120 mesh, acid washed, activated alumina) operated
at ~75°C. The Varian Aerograph gas chromatograph was equipped with a flame
ionization detector. Glc peaks were displayed on a 1-mV potentiometric
recorder equipped with a Disc integrator. The response and linearity of the
entire analytical train (gas sample valve, gc, recorder) were calibrated,
using standard gas mixtures prepared using the Baratron gauge. The combined
precision of the gc analyses and calibrations was better than ±1%.
Results. The equations used to develop the experimental data were
identical for all four olefins. They will be described only as they apply
to cis-olefin. Exchanging the symbols C and T will yield the corresponding
equations for trans-olefin.
For the reaction
N02
cis-olefin > trans-olefin (1)
the initial rate of isomerization at low conversion may be written
d[T]/dt = kcis[N02]0[C]0 , (2)
where k is the experimentally determined rate constant, T and C denote
cis
trans- and cis-olefin and the zero subscripts denote initial concentrations.
This expression may also be written
[T]t/[C]0 = kcis[N02]0t (3)
Since [T]t/[C]0 = 0- + tC]t/[T]t)"1 by measuring [C]t/[T]t as a function of
time (t), the observed rate constant, kcis> cai* be determined easily and
precisely.
All experiments were run to conversions of less than 5%. As expected
for such low conversions, the observed dependence of [C]t/lT]t on t was
always linear. Figure 8.1 illustrates the precision of the linearity
routinely obtained. Nevertheless, this apparent linearity must mask the
perturbing effects of some reverse isomerization.
87
-------
0
40 60
TIME (mln)
80
100
Figure 8.1 Plot of the mole fraction of trans-2-butene ys time
for the cis- to trans-NOg catalyzed geometric isomer-
ization of cis-2-butene. The least-squares slope of
the plot is 2.64 ± 0.03 * 10"^ min"1. Reaction
conditions were 9.00 torr of cis-2-butene, 0.100 torr
of N02, and 54.4 ± 0.2°C. The nonzero intercept is
due to the presence of a trace of trans-2-butene
impurity in the starting cis-2-butene.
88
-------
Normal kinetic treatment of the following reaction scheme
cis-olefin + N02 > N02 + trans-olefin
yields eq. 5, where KT(, = ktrans/kcis- Iterative use of
[C]O+[T]O
eq. 5 alowed the raw rate constants (k . or k ) calculated using eq. 3
cis trans
to be corrected for the perturbing effects of reverse isomerization. To do
this the raw values for k . and k were used to calculate an initial
cis trans
set of Arrhenius parameters for these rate constants. An initial value of
(or K) was then calculated from these Arrhenius parameters. Substitution
of this value into eq. 5 now allowed a new set of values for k . (or k )
H cis trans
to be calculated. This process was repeated until a consistent set of values
was obtained for k . and k
cis trans
Figures 8.2 and 8.3 present plots of log k vs. 6 where 6 = 2.3RT.
The lines are the least-squares lines calculated from the data. The slope
and intercept of each least-squares line give the Arrhenius parameters of
that reaction. Table 8.1 summarizes the experimental results:
89
-------
380
TEMPERATURE (°K)
360 340 320
300
6.4
10,000/0
Figure 8.2 Arrhenius parameters for the N(>2 catalyzed geometric
isomerization of cis- or trans~2-butene. Plot of log
k vs 9"1 where 0 = 2.3RT: (0) £is-2-butene; (•) trans-2
butene.
90
-------
380
TEMPERATURE (°K)
360 340 320
300
1.0 h
6.0
6.4 6.8
10,000/0
7.2
Fisure 8 3 Arrhenius parameters for the N02 catalyzed geometric
Fxgure8.3 £rhe P^ ^ ^ ^^.2 tene. Plot of log
k vs 6-1 where 9 ^T.SRT: (0) cls-2-pentene; (•) trans-2-
pent ene
P
91
-------
Table 8.1
to
Thermodynamic and Arrhenius parameters for the N02 catalyzed geometric isomerization of the 2-butenes
and 2-pentenes
Olefin
-1 -1
k(l mole sec at 25°)
-1 -1
Log A(l mole sec )
E(kcal mole"1)
AH°, 298(kcal mole"1)
r
AH°, 298(eu)
cis-2-Butene trans- 2-Butene
0.159 ± 0.002 0.053 ± 0.001
7.86 ± 0.06 7.65 ± 0.08
11.8 ± .1 12.2 ± 0.1
-1.00 ± 0.02
-1.15 ± 0.03
cis-2-Pentene trans-2-Pentene
0.183 ± 0.005 0.038
7.49 ± 0.06 7.72
11.2 ± 0.1 12.5
-1.16 ± 0.07
-0.81 ± 0.10
± 0.001
± 0.06
± 0.1
-------
For the reaction
ka N02
N02 + RCH + CHR' - > RCH-CHR1
"V
I II
several thermodynamic quantities may be defined as follows:
where k& - ^ ea and
-T
AH°'C u
ab
ekt AS*/R , . . AS°'C /R
— e b andA = Ae ab'
As described elsewhere AH°'Cab and AS°'Cab can be estimated from thermo-
* 3
chemical data and ASb can be estimated by incremental methods.
The activation energy for C-C bond rotation in the radical II should be
-1 4
small (~5 kcal mole ) compared to the activation energies (Ea and Ejj) for
the formation and unimolecular decomposition of the radical. Therefore,
values of Ea for the 2-butene and the 2-pentenes may be calculated directly
from the experimental data presented above. For simple olefins, however,
Efc is unlikely to be very sensitive to whether R or R1 is an alkyl group of
an H atom. Therefore, equation (4) permits Et to be estimated from the
calculated 2-butene and 2-pentene values for AHab and Ea. Using a 3a error
limit gives Efc = 12.8 ± 2 kcal mole"-'- for the unimolecular decomposition of
the radical (II) derived from any simple olefin.
Once a value for Efc is available, Ea can be estimated for any olefin
for which AH° b can be calculated. Then, because Aa can be calculated from
ASb* and S° b, k can also be estimated. Table 8.2 summarizes the results
of our thermodynamic calculations. It should be noted that unsymmetrical
olefins give two values of ASt,.
In polluted urban atmospheres, at least during the day, the concentrations
of 03 and N02 are roughly comparable. Therefore, since the rate constants
93
-------
Table 8.2
vo
Calculated and experimental data for
Olefln
Acetylene
Ethylene
Propylene
1-Butene
Iso-Butene
cis-2-Butene
trans- 2-Butene
1-Pentene
cls-2-Pentene
trans-2-Pentene
AH°'C .
ab
kcal mole
2.60
2.19
1.75
1.70
1.51
-0.14
0.61
4.46
-0.37
-0.17
1.29
-0.16
-0.65
0.15
-0,56
0.24
AS°'C .
ab
eu
-25.0
-24.1
-25.4
-27.7
-25.7
-28.4
-25.2
-30.0
-27.0
-25.9
-26.1
-28.6
-28.5
-28.1
-27.1
-26.7
b
eu
4.5
4.6
0.8
4.8
0.8
4.8
3.9
5.1
1.4
1.4
1.2
4.8
1.4
1.6
1.4
1.6
the reactions of nitrogen dioxide with selected olefins.
1 mole sec
58.1
93.6
7.40
17.4
6.24
12.0
37.5
6.15
4.41
7.44
6.21
10.8
2.05
2.76
4.13
5.59
Ea
kcal mole
14.4
14.0
13.6
13.5
13.3
11.7
12.4
16.3
11.4
11.6
13.1
11.6
11.2
12.0
11.3
12.1
calc
k x 1Q3 a^ k x 10^
a TTO2
. . -1 -1
1 mole sec
0.017
0.053 1.8
0.009
0.022
0.011
0.348
0.031
6.3
12.3
0.359
0.301 /
[ 0.301 51.0
o.oooil
0.189 23.0
0.023 32.0
0.016
0.321
0.140
0.049
0.243
0.085
0.337 6.6
0.189
0.328
For unsymmetrlcal olefins, the overall rate constant Is the sum of the rate constants for the formation of each of
the two possible Intermediates.
b)
See Ref. 5.
-------
for atmospheric consumption of simple olefins by 03 are all 103 - 101* 1 mole
sec , it is obvious that consumption of olefins by reaction pathways initiated
by N02 addition to the double bond of an olefin cannot compete kinetically with
olefin comsumption by 63.
95
-------
REFERENCES
1. J. L. Sprung, H. Akimoto and J. N. Pitts, Jr., J. After. Chem. Soc.,
^3, 4358 (1971).
2. J. L. Sprung, H. Akimoto and J. N. Pitts, Jr., J. Amer. Chem. Soc.,
96, 6549 (1974).
3. S. W. Benson and H. E. O'Neal, Nat. Stand. Ref. Data Set., Nat. Bur.
Stand., No. 21 (1970).
4. S. W. Benson, K. W. Egger and D. M. Golden, J. Amer. Chem. Soc., 87,
468 (1965).
5. S. Jaffe in "Chemical Reactions in Urban Atmospheres," C; S. Tuesday, ed.,
Elsevier, New York, N.Y. (1971).
96
-------
9- 40 METER LONG-PATH INFRARED STUDIES OF THE PROPYLENE/NQv/hv SYSTEM
Experimental. The experimental system consisted of a long-path infra-
red (LPIR) spectrometer (Perkin-Elmer Model 621; dual beam, 40-m pathlength),
whose sample and reference tanks (Perkin-Elmer Model 198; diam 27 cm, length
115 cm, volume -66 liters) were connected to a conventional high vacuum
system. To minimize wall reactions, the optical benches housed in the sample
and reference tanks and the interior surface of each tank were coated with
FEP Teflon (T^ = 11 hr for the thermal decay of 03). Both tanks were evacuable
to better than 7 * 10~4 torr (1 ppm). A 20-liter, all-glass reaction vessel
in which preliminary studies can be run was connected to the high vacuum
line. Stainless steel valves and 0-ring stopcocks were used throughout the
system.
Irradiations of the sample tank were carried out using a medium pressure
mercury arc (Hanovia, 1200 watt). Light from the arc entered the sample
tank through six quartz windows (7.5 x 7.5 x 0.6 cm), mounted axially along
the top of the tank. The windows transmitted light of wavelength X>210 nm.
Sets of cut-off filters were used to block light of wavelengths shorter than
290, 300, or 320 nm, respectively. When the 300-nm cut-off filters were
used, the photolysis rate of N02 was 0.042 min"1.
Pressures in the 20-liter glass reaction vessel and in the greaseless,
mercury-free, high-vacuum line were measured using a Wallace and Tiernan
gauge (0 to 800 torr) or an MKS Baratron capacitance manometer (10~3 to 10 torr),
Pressures in the LPIR sample or reference tank were measured using the Wallace
and Tiernan gauge or an Alphatron gauge (1Q-4 to 10~3 torr). Expansion of
known pressures of reactants, measured with the MKS Baratron capacitance
manometer, from calibrated volumes on the vacuum line into the sample tank
of the LPIR allowed low concentrations (1 to 500 ppm) of reactants to be
accurately metered into the tank.
All reagents (hydrocarbons, aldehydes, NO, and N02) and diluents (N2,
02, and He) were obtained in the highest purity available. Organic gases
and N02 were further purified by repeated freeze-thaw cycles using appropriate
cryogens. Traces of N02 and H20 were removed from NO by passage through a
molecular sieve (Linde 13X) at liquid nitrogen temperature. Ultra high purity
N2, 02, or He were used as received. Ozone in 02 was prepared using a
Welsbach ozonizer and was used without further purification.
97
-------
Adsorption of ozone on silica gel, followed by elution with % or He,
allowed mixtures of 03 in N2 or He to be prepared which contained low
concentrations of 02([03]:[02] -100:1). Ozone concentrations were determined
by infrared absorption at A = 9.48 p, where e = 3.74 x 10~4 ppnf1 nT1.
Table 9.1 summarizes the analytical techniques which were used to
analyze reaction mixtures produced in either the glass tank, used for pre-
liminary studies, or the LPIR sample tank. All of the sampling methods
were routine except for the use of chromosorb adsorbents to effect enhance-
ments of product concentrations by factors of 100 to 1000. In this case,
volumes up to the entire contents of the LPIR sample tank were evacuated
through tubes packed with chromosorb. The tubes were transferred to an
oven and the adsorbed compounds were thermally desorbed at 120°C and collected
in a liquid nitrogen-cooled trap. This condensate was then analyzed by flash
vaporization onto the GC column of the Finnigan 3100D GC/MS.
Results. The following characterization experiments were run in the
66 Si FEP Taflon-coated sample tank of the LPIR reaction system:
(1) Dark decay of 03: x^3 , , = 11.0 hrs.
(2) Dark decay of N02: T>°2d - 73.4 hrs.
(3) Photolysis of N02 in N2: k, - 0.042 rnin" for X>300 nm using a
medium pressure Hg arc.
The response and linearity of the entire analytical train (LPIR sample
tank, gas sample valve, gas chromatograph, recorder) was calibrated by
removing and analyzing known pressures (MRS Baratron capacitance manometer)
of authentic samples of all reactants (€3%) and most major reaction products
(C2H5CHO, (CH3)2CO, CH3CHCH20, C2H5ONO, CH3ON02, CH3CHO) which were analyzed
by gas chromatography.
A complete study has been made of the reaction products formed by the
N02 photooxidation of C3H6 (typical conditions: [N02]0 = 25 ppm, [C3H6]0 -
125 ppm, synthetic air [N2/02 = 4.0], rel. hum. » 0, medium pressure Hg arc,
300 nm cut-off filters, k*J°2 - 0.042 min"1). Figures 9.1 and 9.2 present
typical concentration/time profiles for these experimental conditions.
98
-------
Table 9.1
Analytical methods
Instrument
Sampling method
Compounds detected
Sensitivity
40-m LPIR spectrometer
In situ
N02, CO, C02, carbonyl
compounds, olefins, PAN,
nitrites, and nitrates
5 to 100 ppm
Dual-column flame ionization
Carle valve (1.0-ml
sample loops)
Hydrocarbons, aldehydes
(except formaldehyde),
ketones, epoxides, and
other oxygenates
>0.1 ppm
v£>
Single-column electron capture
Carle valve (0.5-ml
sample loops)
PAN, nitrites, nitrates, £0.1 ppm
nitro compounds, 02
Chemiluminescence detector
Calibrated leak
(flow ~3 ml s'1)
03, NO
£0.1 ppm
Gas chromatograph/mass
spectrometer (GC/MS)
Syringe samples (100 ml)
adsorption technique
Same as above for GC
methods
1 ppm
£3 ppm
-------
I40r-
s
a.
a.
120
100
80
60
40
20
0 CH3CH»CHe
• CH3CHO
A N02
20
30 40
TIME (MIN)
50
60
70
Figure 9.1 Typical concentration-time profile for the N(>2 photooxidation
of C3H6.
100
-------
I-1
o
OL
Q.
3.0
2.0
1.0
0 CH3CHCH2
ACH3CH2CHO
• PAN
A CH3ON02
0 CH3CH2ON02
10
20
30
40
50
G3
60
70
80
TIME (WIN)
Figure 9.2 Typical concentration-time profile for the N02 photooxidation of
-------
Table 9.2 presents the products identified, their methods of identification
and their approximate yields. Figure 9.3 presents a typical total ion
chromatogram of a product mixture analyzed by GC/MS. Product sampling for
this analysis was done by Chromosorb adsorption. The presence of acetone
impurity in the carrier gas of the GC/MS and its condensation in the liquid
N2~cooled sample loop of the GC accounts for the unreasonably high yield
of acetone in this total ion chromatogram. Figures 9.4, 9.5 and 9.6 present
representative computer reduced mass spectra of three products [CI^CHO,
(CH3>2CO, CHsOH] present in this total ion chromatogram.
Samples of a CsHg/NOx photooxidation product mixture were adsorbed on
Chromosorb adsorbents and analyzed by GC/MS (Carbowax 600 column). This
analysis detected CHsCHCl^O and C2H5CHO, the first having been previously
undetected by the procedures then used to analyze for oxygenates from hydro-
carbon/NOx chamber photooxidation runs. Adoption of the Carbowax 600 column
used in the GC/MS analysis for GC analysis of chamber oxygenates directly
confirmed the presence of these two 0(3P) atom/propylene addition products
in chamber product mixtures. CH3ON02 was also identified by GC/MS analysis
of chamber product mixtures sampled by adsorption on Chromosorbs. Fourier
interferometry had previously detected an alkyl nitrate, tentatively identi-
fied as C2H50N02, in the C3Hg/NOx photooxidation product mixtures run in the
evacuable chamber.
102
-------
Table 9.2
Reaction products from the photooxidation of C3H6 by
NO-
Product
Analytical technique
LPIR FID-GCa) EC-GCb) GC-MSC'd)
Approximate yield
at 115 min (ppm)
C2H5CHO / / /
(CH3)2CO / / /
CH3CHCH20 / /
CH3CHO / , / /
HCHO /
CH3ONO / /
CH3ON02 / / /
C2H5ONO / /
C2H5ON02 / / /
PAN /
HC02CH3e) /
CH3C02CH3e) /
CH3OH /
C2H5OH /
CO /
C02 /
03 UV absorption (Dasibi)
2.9
0.3
2.7
29.8
major
minor
2.5
trace
0.1
2.6
minor
trace
trace
trace
minor
minor
0.2
10 ft, 1/8 in, 10% Carbowax 600 on 80/100 C-22 firebrick at 50°C
20 ft, 1/8 in, 10% Carbowax 600 on 80/100 C-22 firebrick at 50°C
20 ft, 1/8 in, 10% Carbowax 600 on 80/100 C-22 firebrick at 50°C
Samples removed from LPIR by adsorption on chromosorbs.
Possibly formed by esterification of the related acid during adsorption
on chromosorb or collection in Iq. N2 trap.
103
-------
ION OfttWflTOGftfln
100
O
*•
SOQ SSO
60S
SPECTRUM
Figure 9.3 A typical GC-MS total ion chromatogram of an irradiated
mixture.
-------
ACETALDEHYDE
SPECTRUM NUMBER 60
-59
LPIR2
100
Ld
CD
fe50
UJ
a
UJ
o
a:
LU
a.
)
«•
)-
-
-
-
l(
3
tnrmtrl
-
50
-30
" i
-150
LU
2
LU
LU
Q_
M/E
Figure 9.4 GC-MS computer spectrum of CE^CRO as
obtained from the total ion chromato-
gram of an irradiated
mixture.
105
-------
ACETONE
SPECTRUM NUMBER 115
-III
LPIR2
iuvj-
.
^l>
^£
UJ
Q.
UJ
CO
CD
feso-
UJ
LU
^
UJ
Q.
n
w
K
nf
3
-Junjll
>
50
•
•
-30
•*
-J
1
H
fe
•is y
UJ
o
o:
LJ
CL
n
w
M/E
Figure 9.5 GC-MS computer spectrum of (CH3)2CO
as obtained from the total ion chro-
matogram of an irradiated
air mixture.
106
-------
METHANOL
SPECTRUM NUMBER 193
-186
LPIRS
100-
•40
UJ
CL
LU
CO
QQ
fe.so-
LU
|
LU
O
tr
LU
a.
0-
-20!
LU
O
ir
LU
a.
•0
M/E
Figure 9.6 GC-MS computer spectrum of CHsOH as
obtained from the total ion chromato-
gram of an irradiated N02-C3H6-air
mixture.
107
-------
REFERENCES
1. P. L. Hanst, E. R. Stephens, W. E. Scott and R. C. Doerr, Anal. Chem.,
33, 1113 (1961).
108
-------
10- THE OZONE-INDUCED CHEMILUMINESCENT OXIDATION OF ACETALDEHYDE
Experimental. A conventional gas handling and flow system with aim
flow tube was used. Emission spectra were recorded with a 0.3 m McPherson
scanning monochromator (600 grooves/mm blazed at 500 nm) and a dry ice
cooled EMI 9558A (500 to 800 nm) or 9684A (700 to 1100 nm) photomultiplier.
A lock-in amplifier (Princeton Applied Research, Model 120) and a mechanical
chopper (360 Hz) were used to enhance the emission signal to noise ratio.
Approximately 3 mole percent ozone in oxygen (determined by UV absorb-
ance of 03 at 250 nm) was produced by passing Matheson ultra-high purity
(>99.95%) grade oxygen through a Welsbach ozonizer (Model T-408). Flow rates
of the 03/02 mixture were typically 300 ymoles/sec.
Formaldehyde was produced by the method of Spence and Wild from para-
formaldehyde (Matheson, Coleman and Bell) and was used without further
purification. The source and percentage of total organic impurities as
measured by gc analysis (Carbowax 400/chromosorb W column, 60°C, flame
ionization detector) of the remaining aldehydes were as follows: propion-
aldehyde (Matheson, Coleman and Bell, <0.3%), benzaldehyde (Matheson, Coleman
and Bell, 0.2%) and acetaldehyde (Matheson, Coleman and Bell, 0.2%;
Mallinckrodt, 0.1%; Chemical Samples Company, 0.02%; Aldrich, 0.08%, Eastman
0.8%). In searching for a detectable emission, reactant flow rates, total
pressure and linear flow rate were varied. The acetaldehyde flow rate during
spectral runs was typically 5 ymoles/sec and the total pressure was about
3 torr.
Results and Discussion. In addition to the observed OH emission, an
emission in the visible about a factor of forty less intense than the OH
emission with a maximum of 430 nm was also recorded. Although the signal
to noise ratio was not sufficient to obtain a well-resolved spectrum, a
2 3
comparison to the visible emission from the ozone-olefin reactions ' shows
that this emission is probably formaldehyde fluorescence.
Figures 10.1 and 10.2 give the chemiluminescent emission spectra of the
acetaldehyde oxidation from 500 to 1000' nm. For comparison, the vibration-
rotation bands of OH(2n±) produced in the same apparatus by the reaction
H + 03 -> OH+ < 9 + 02 (1)
109
-------
(a) 03/02 + CH3CHO
I I I
LJ
£ (b) 03+ H— OH;S9+ 02
UJ
cc.
(7.1) (8.2) (5.0)(9.3) (6.1) (7.2) (8.3) (4.0) (9.4)(5.1)
_L
J_
500
550
600
650
X —<
700
750
800 nm
(UNCORRECTED FOR SPECTRAL SENSITIVITY )
Figure 10.1 Emission spectra from 500 to 800 rm in the chemiluminescent
reactions of ozone with (a) acetaldehyde. Total pressure
4.0 torr; spectral bandwidth 9 nm. (b) H atoms. Total
pressure 4.0 torr; spectral bandwidth 6.9 nm.
110
-------
CO
LU
h-
z
UJ
<
_l
(T
(a) CH,CHO
(b) 0
OH
v^9
(8.3) (4D) (5.1)
, 04) i
(62)
(7.3)
(8.4) (3.0) (9.5) (4 )
i I
(5.2)
700
800
900
(UNCORRECTED FOR SPECTRAL SENSITIVITY )
1000
HOOnm
Figure 10.2 Emission spectra from 700 to 1100 nm from the chemiluminescent reactions of
ozone with (a) acetaldehyde. Total pressure 3.3 torr; spectral bandwidth
5.3 nm. (b) H atoms. Total pressure 2 torr; spectral bandwidth 2 nm.
-------
are also presented in these figures. Since the spectra are identical, the
chemiluminescence produced by the acetaldehyde-03/02 reaction must also be
A
due to the Meinel bands of the OH radical. The similarity of the vibrational
distribution of the two spectra suggests that reaction (1) is the likely
source of the vibrationally excited OH produced during the oxidation of
acetaldehyde by 03.
Because only the acetaldehyde oxidation was chemiluminescent, a rather
surprising result, the possibility of light emission due to impurities in the
acetaldehyde, particularly unsaturated hydrocarbons, was examined. The fact
that the same emission was observed when five different sources of acetaldehyde
were used argues strongly against tshis possibility. Furthermore, even if all
the organic impurities in the acetaldehyde samples were olefins, one would
have expected to observe an emission at most a factor of one hundred less
2
intense than is observed. Since the observed emission was only a factor of
two less intense than the ozone — tetramethylethylene OH emission, we there-
fore conclude that the observed emission is not due to impurities.
These preliminary results do not permit the formulation of a detailed
reaction mechanism; the mechanisms of ozonolysis of aldehydes are known to
5—8
be very complex. While there is no obvious explanation of why only the
acetaldehyde oxidation is chemiluminescent, studies of the thermal oxidation
of aliphatic aldehydes have revealed differences in the oxidative rates and
o
mechanisms of formaldehyde as compared to higher aldehydes. In the case of
propionaldehyde, the presence of 3 hydrogens may alter the decomposition mode
of intermediates which, in the acetaldehyde oxidation, lead to excited OH.
Further investigations of the detailed mechanisms of these reactions
are currently in progress.
112
-------
REFERENCES
1. R. Spence and W. Wild, J. Chem. Soc., 338 (1935).
2. B. J. Finlayson, J. N. Pitts, Jr. and H. Akimoto, Chem. Phys. Letters,
3.2, 495 (1972).
3. W. A. Kummer, J. N. Pitts, Jr. and R. P. Steer, Environ. Sci. Technol.,
_5, 1045 (1971).
4. A. B. Meinel, Astrophys. J., Ill, 555 (1950).
5. P. S. Bailey, Chem. Rev., _58, 925 (1958).
6. R. E. Erickson, D. Bakalik, C. Richards, M. Scanlon and G. Huddleston,
J. Org. Chem., 31, 461 (1966).
7. H. M. White and P. S. Bailey, J. Org. Chem., _30, 3037 (1965).
8. A. A. Syrov and V. K. Tsykovskii, J. Org. Chem., USSR, j6, 1406 (1970).
9. J. F. Griffiths and G. Skirrow, Oxid. Combust. Rev., 3_, 47 (1968).
113
-------
11. LOW-PRESSURE GAS-PHASE OZONE-OLEFIN REACTIONS.
CHEMILUMINESCENCE, KINETICS. AND MECHANISMS
Extensive mechanistic investigations of the liquid-phase reactions of
ozone with olefins have identified many of the reaction intermediates and
have established the Criegee zwitterion mechanism as a major reaction path-
way. ~ Until recently, the Criegee mechanism has also been widely assumed
to apply to the initial steps of the gas-phase reaction. However, in the
gas phase at room temperature, unimolecular decomposition or rearrangement
of the "zwitterion" (more likely a biradical in the gas phase) is expected
to predominate over recombination with the carbonyl fragment. ' Although
several reports of gas-phase products, including secondary ozonides, '
are consistent with this mechanism, it fails to explain the formation of
11 12
free radical intermediates ' in low-pressure (~2 torr) ozone-olefin
12
reactions and of "unusual" ozonolysis products, both at low and high total
7,8,13,14
pressures.
Recently, evidence on the nature of certain excited intermediates in
gas-phase ozone-olefin reactions has appeared. Thus, chemiluminescence from
the room-temperature gas-phase ozone-ethylene reaction at atmospheric pressure
was first observed by Nederbragt, et al., and further investigated by Warren
and Babcock and Hodgeson and coworkers. Subsequently, studies '
in our laboratories showed that at low pressures (~1 torr) all simple olefins
chemiluminesce on reaction with 2% 03/02- Meinel band emission from vibra-
tionally excited OH radicals was observed for the first time in these oxi-
dations, and formaldehyde fluorescence was tentatively identified in the
reactions of ethylene, propylene, and 1-butene.
Concurrently, on the basis of thermochemical-kinetic calculations, O'Neal
23
and Blumstein proposed alternatives to the gas-phase Criegee mechanism.
These involve internal hydrogen abstractions of the initial molozonide in
addition to its decomposition to the Criegee fragments. Their mechanism
rationalizes most of the "unusual" products observed in previous studies, as
well as the production of chemiluminescence.
Reported here are the results of further detailed studies of low-pressure
ozone-olefin reactions, using more sensitive and more quantitative spectro-
scopic techniques to provide the requisite resolution for conclusive identi-
fication of the light-emitting species and to permit estimates to be made
114
-------
of their absolute rates of light emission. Additionally, the kinetics of
the light intensities and of ozone decay, measurement of the self-heating
of the gas mixture in these exothermic reactions, analysis of the major
stable products, and the effects on these parameters of replacing the oxygen
carrier gas by nitrogen are presented. This evidence, along with our recent
results on intermediate species obtained using a photoionization mass spec-
12
trometer (hereafter referred to as PMS), is discussed in terms of the
O'Neal-Blumstein mechanism.
Experimental. The flow system used in these studies is shown schematically
in Figure 11.1. Along the length of the Pyrex flow tube (5-cm i.d., 1-m length)
were five substrate inlet jets (Ji-Js) and five Pyrex bead-type thermistors
(TI-TS) (Fenwal Electronics Inc.) which had been calibrated against a quartz
thermometer. Pyrex or quartz windows were attached to the ends of the flow
tube using Teflon-lined end couplings. Flow tube pressure was measured
with a Statham 0-5 psia transducer (A) calibrated against an MKS Baratron
gauge. Flow tube pressures were not corrected for pressure drop along the
flow tube, as the Poiseuille equation predicts a pressure drop of ~1 v, £0.1%
of the total pressure. The flow tube was connected to a conventional high-
vacuum line via Teflon stopcocks and Teflon-lined ball and socket joints.
Where appropriate, a fluorinated grease (Krytox 240AC) was used to minimize
03 decomposition. The flow tube was periodically cleaned by baking overnight
at 565°C.
For studying the kinetics of the chemiluminescent emission of individual
excited species, appropriate optical filters (C) were used to isolate the
emission which was detected by a cooled EMI 9558A photomultiplier (B). The
following optical filters were used: (1) for HCHO^A" -»• 1A-i)t cutoff filters
isolating the 373 < A < 473 nm region; (2) for OHtX2^)^^ Meinel band
emission, a cutoff filter transmitting X >555 nm; (3) for OH(A2E -> X2^),
a 317 ± 6 nm interference filter; and (4) for the emissions at 517-520 nm,
cutoff filters isolating the 510 < X <555 nm region.
The kinetics of 03 consumption were followed by monitoring the 253.7 nm
absorption of 03 using a pen-ray lamp (D) and RCA 1P28A photomultiplier (G).
The filter combination (E,F) isolated the 253.7 nm line and reduced scattered
light at other wavelengths.
Chemiluminescent emission spectra were scanned using a McPherson 0.3 m
scanning monochromator (H) and a cooled photomultiplier (I) (EMI 9684B,
115
-------
03-CARRIER GAS
DILUENT
GAS
H
OLEFIN
DG
FT
A
H-
T5
B
K
Figure 11.1
Schematic of apparatus used in chemiluminescence and kinetic studies:
A, pressure transducer; B, EMI 9558A photomultiplier; C, optical filters;
D, pen-ray mercury lamp; E, Corning 7-54 filter; F, 254 ± 26 nm filter;
G, RCA IP28A photomultiplier; H, 0.3-m McPherson scanning monochromator;
I, photomultiplier; K, Carle gas sampling valve; L, gas chromatograph;
Jj-Js, inlet jets; TX-TS, thermistors.
-------
9558A or 9659QB) whose output was amplified by a lock-in amplifier (PAR Model
120, operated at 360 Hz) or a photon counter (SSR Instruments Model 1120
amplifier discriminator, and Model 1105 data converter console) and displayed
on a potentiometric recorder. To enhance the observed emission intensities,
a plane mirror was placed opposite the monochromator at the downstream end
of the flow tube.
Gas chromatographic analysis of the cis-2-butene-ozone reaction was
carried out using a Carle gas sampling valve (K) with 5 cm3 sample loops
attached to a Varian Aerograph Model 1200 gas chromatograph (L) with a flame
ionization detector. The columns used were a 10 ft by 1/8 in. 20% SE-30
column on 6/8 firebrick, operated at room temperature, and a 10 ft by 1/8 in.
10% 3,3-oxydipropionitrile column, on 80/100 HMDS Chromosorb P, operated at
90°C. Products were identified by comparison of their retention times to
those of authentic compounds introduced into the Carle valve as a mixture
with air.
03/diluent and substrate flow rates were measured using calibrated flow-
meters. Flow rates were typically 20-30 pmol sec of olefin and -130 ymol
_-|
sec" of the 03/diluent gas mixture at total flow tube pressures of 2-10 Torr.
Linear flow velocities were -80 cm sec corresponding to a mean flow tube
residence time of -1.3 sec.
The absolute rate of light emission from individual emitting species
o /
was determined by comparison of the light intensity in the appropriate
wavelength region from the 0-NO-reaction, IO_NO» to that from the ozone-olefin
reaction, In n,, where
Ug—UJL
VNO = ks[oHNO]
and
VOL = kxCOsl [olefin] (II)
k and k^ are the respective absolute rates of light emission in that wave-
s ^
length region.
0(3P) atoms were generated by the microwave discharge of N2 and subsequent
25
titration of the N atoms by nitric oxide
117
-------
N + NO —> N2 + 0(3P) (1)
with the 0(3P) atom concentration being obtained from the end point of the
titration. The 10.3% NO/N2 was passed through a trap at Dry Ice temperature
containing Linde Molecular Sieve 13X to remove any N02 and H20 present.
The concentrations of 03 and olefin were calculated, assuming 1:1
stoichiometry, from their initial concentrations using the measured rate
constant for 03 decay under these experimental conditions and the reaction
time to the observation port.
Then, using the experimentally determined values of IQ_NO and ^-Os-OL
and the known value of k , kjj was calculated. The total rate of light
s
emission from that emitting species was then determined from the fraction
of the total light emission from that excited species transmitted by the
filter. For HCHO^A" -> ^j) this fraction, isolated using a 430 ± 11 nm
interference filter, was measured from its photoexcitation emission spectrum
by planimetry. Similarly, in the cis-2-butene reaction, the rate of light
emission from glyoxal from 510 to 555 nm, isolated using cutoff filters, was
measured and the fraction this comprised of the total emission was calculated
from the photoexcitation emission spectrum. It was assumed that at these
total pressures no glyoxal fluorescence would occur due to rapid intersystem
26—28
crossing from the singlet to the triplet state.
99
The total OH Meinel band emission which extends to 4.5 y could not be
recorded in this study due to instrumental limitations. Therefore, the inten-
sity of a fraction of the (9,3) band was measured, ahd the total rate of light
emission in the (9,3) band was calculated from the known transmission character-
istics of the filter.
Ethylene, cis-2-butene. and isobutene (Matheson Research Grade >99.8%)
were used without further purification. Gas chromatographic analysis of
these olefins showed no detectable impurities. 2-Butene-d8 (mixture of cis
and trans) and isobutene-dg (Merck Sharp and Dohme, with stated D atom purities
of >98% and £99%, respectively) were used as received.
Ozone (~2% in 02) was prepared from 02 (Matheson ultra-high purity)
using a Welsbach ozonator Model T-408. 03/N2 or 03/He (both containing <0.2%
02) were prepared by selective absorption of 03 on silica gel30'31 at -78°C
followed by elution with the diluent gas (N2 or He).
H2 (>99.999%), He (>99.95%), and N2 (>99.95%) were used as received. H
118
-------
atoms were generated by methods described previously.
Formaldehyde was prepared from paraformaldehyde (Matheson Coleman and
Bell) by the method of Spence and Wild. Acetaldehyde (Matheson Coleman
and Bell), acetone (Mallinckrodt Spectra Grade), and biacetyl (Matheson
Coleman and Bell) were thoroughly degassed at liquid nitrogen,temperature
and used without further purification. Glyoxal was prepared from glyoxal
trimer (Matheson Coleman and Bell) by heating a mixture of trimer and P205
and collecting the glyoxal distillate in a liquid nitrogen trap. Volatile
impurities in the glyoxal were removed by degassing at Dry Ice temperature.
Results
Identification of Chemiluminescing Species. Electronically excited
formaldehyde, ECHO^A"), has now been confirmed as a Chemiluminescing species
common to all the ozone-olefin reactions that were studied at sufficient
resolution to resolve the vibrational structure. Typical chemiluminescent
emission spectra (350 < A < 600 nm) are shown in Figure 11.2; also shown for
33
comparison is the formaldehyde fluorescence excited by tesla coil discharge
and recorded with the same detection system.
The reactions of cis- and trans-2-butene with ozone have identical chemi-
luminescent emission spectra, an example of the latter being shown in Figure
11.2ii. The 520-nm peak from these reactions is assigned to glyoxal phosphor-
escence (3A -> *A ) by comparison with the (0,0) band obtained by photoexcitation
of glyoxal vapor at 430 ± 11 nm. The remaining vibrational bands of the trip-
34 35 ' '
let system ' are obscured at these pressures by the much stronger Meinel
band emission of the OH-(X2ir ) I
-------
(i)
i I I I i i i I I I I 1 1 L
360
380
400
420 440
460
480 500 nm
360 380 400 420 440 460 480 500 520
(Hi)
HCHO (W-'A,)
360
360
400
420
440
460
480
500
SZO
S40
560
580 nm
Figure 11.2
Chemilumine scent emission spectra in the visible region
from the reaction of 2% 03/02 with: (i) ethylene,
total pressure 4.5 torr, spectral slit width 3.2 nm;
(ii) trans- 2-butene . total pressure 3.5 torr, spectral
slit width 2.5 nm; (iii) isobutene, total pressure 4.4
torr, spectral slit width 3.2 nm; (iv) isobutene, total
pressure 4.4 torr, spectral slit width 1.9 nm; (i-iii)
HCHOOA" •*• ^i), total pressure 4.9 torr, spectral slit
width 2.5 nm; (ii) (CHO)2(3AU -> ^g) , total pressure
1.33 torr, spectral slit width 0.4§ nm.
120
-------
37
Yardley and coworkers was in good agreement and hence it is believed that
this is the emitting species. The chemiluminescence at A>440 nm (Figure 11.2iii)
may be the sum of formaldehyde fluorescence and methylgloyxal fluorescence.
The chemiluminescent emission spectrum from the cis-2-butene-ozone reaction
in the region 700 < X < 1100 nm is compared in Figure 11.3 to that from the H
+ 03 reaction under similar conditions of temperature and pressure. Reaction
2 is known ' to produce the Meinel bands of vibrationally excited OH^ .
v'<9,
with v' = 9 corresponding to the exothermicity of the reaction
H + 03 —> OH(XTT1)v^9 + 02 (2)
AH0° = -77 kcal/mol
Both reactions have virtually identical vibrational distributions especially
in that emission from levels v' > 9 is not observed. In addition, the
rotational distributions are similar, further confirming H atoms as the
precursors of the OH chemiluminescence in ozone-olefin reactions.
Secondary reactions of the major products cannot be responsible for the
40
Meinel band chemiluminescence since only acetaldehyde gives this emission
on reaction with 63 and the intensity in this case is at least a factor of
10 less than that observed in the cis-2-butene-ozone reaction.
To determine whether oxygen was involved in the chemiluminescent processes,
spectra from the reactions of ozone with ethylene, cis-2-butene, and isobutene
were recorded first in N2 and then in 02 as diluent. Use of He rather than
N2 caused no detectable change in the spectra. Figure 11.4 shows that the
HCHO^A" -> 1A-i) fluorescence is unaffected by the removal of 02. However,
the Meinel bands appeared to increase in intensity and the (0,0) band of
electronically excited OH(A2E+ -»• X2^) (hereafter referred to as OH ) was
observed at 306.4 nm (the (1,0) band at 281.1 nm was not detected). While the
intensity of OH* in the presence of 02 was not sufficient to be recorded using
a monochromator, a signal was observed using a 317 ± 6 nm interference filter
in the cis-2-butene and isobutene reactions.
Because of the increased intensity of the OH Meinel bands in N2, it
could not be ascertained whether the 517 and 520 nm peaks observed in the
isobutene and ^is-2-butene reactions in 02 were still present. Therefore,
in order to eliminate the OHf emission, the spectra from the reactions of
121
-------
(a) 03 + W
CO
-z.
LU
(b)
KJ
to
UJ
LJ
o:
(8,3) (4,0) (9H)(5,I) (6,2)
700
800
900
1000
HOOnm
Figure 11.3
Comparison of the chemiluminescent emission (700-1100 nm) from
the ozone-cis-2-butene reaction to the OH Meinel bands from
the H + 63 reaction: (i) total pressure 3.2 torr, spectral
slit width 1.2 nm; (ii) total pressure 1.02 torr, spectral
slit width 0.72 nm.
-------
(i)
.........
300
400
500
600nm
Figure 11.4 Chemiluminescent emission spectra (300-550 nm)
from the reactions of 2% 03 in (a) 02 and (b)
N2 with: (i) ethylene, total pressure 1.79
torr, spectral slit width 9.0 nm; (ii) cis-
2-butene, total pressure 1.58 torr, spectral
slit width 7.4 nm; (iii) isobutene, total
pressure 1.60 torr, spectral slit width 7.4 nm.
123
-------
the corresponding perdeuterated olefins with 03/02 and 03/N2 were recorded,
as shown in Figure 11.5. As expected, no OD emission was observed up to
700 nm. This was confirmed by the lack of OD emission from the D + 03
reaction in the same system. It appears that 02 is necessary for production
of the 517 nm peak tentatively identified as methylglyoxal phosphorescence
in the isobutene-ozone reaction. However, the experiments with the 2-butene-
ds were inconclusive since the glyoxal-d2 phosphorescence was reduced by
41
more than a factor of 2.5 even in 02, suggesting a kinetic isotope effect.
Table 11.1 summarizes the emitting species identified from the chemi-
luminescent emission spectra of the ozone-olefin reactions in the wavelength
region 200 to 1100 nm.
Kinetics of Light Emission
a. Intensity of Light Emission as a Function of Reactant Concen-
tration. The dependence of each of the emission intensities on olefin and
ozone concentration was investigated under pseudo-first-order conditions.
Reaction orders were determined from least-squares analyses of the slopes of
log-log plots of emission intensity vs. reactant concentration. Table 11.2
shows that the emission intensities were first order in each reactant, with
the exception of the isobutene reaction where the order of the 517 nm peak
in ozone was 0.39 ± 0.02 in 02 as carrier gas. With N2 as carrier gas, the
light emission generally exhibited a maximum at very low olefin concentrations
and then decreased with increasing olefin concentration.
b. Decay of Light Intensity and Ozone. For ethylene, cis-2-butene,
and isobutene, the intensities of each of the chemiluminescent emissions and
the 03 concentration were determined as a function of time with 02 or N2 as
the diluent gas over total pressures of 2-10 torr. In order to maintain
pseudo-first-order conditions, the initial olefin concentrations were a mini-
mum of six, and typically ten, times the initial ozone concentration. Decay
rates were determined from the least-squares slopes of plots of the logarithms
of either emission intensity or ozone absorbance against reaction time.
Typical data for the cis-2-butene reaction with ozone are shown in Figure 11.6.
Similar behavior was observed for the ethylene and isobutene reactions.
Figure 11.6i shows that in 02 as carrier gas the emission intensities
and the ozone concentration both decay exponentially at times >0.1 sec. The
ratio (decay rate of emission intensity)(rate of ozone decay) was determined
for each run from plots such as those shown. The average value of this ratio
124
-------
600nm
300
550
COOnm
Figure 11.5 Chemiluminescent emission spectra (300-600 nm)
from the reactions of 2% 03 in (a) N2 and (b)
02 with: (i) 2-butene-dg. total pressure 1.72
torr, spectral slit width 9.0 nm; (ii) iso-
butene~ds» total pressure 1.93 torr, spectral
slit width 9.0 nm.
125
-------
Table 11.1
Summary of the chemiluminescing species (200 < X < 1100 nm) identified in
ozone-olefin reactions in the presence and absence of oxygen.a
Olefin
Ethylene
cis-2-Butene
trans-2-Butene
Propylene
1-Butene
Isobutene
2-Methyl-2-
butene
2,3-Dimethyl-2-
butene
Chemiluminescing species
CH2CO-
Carrier t * * * *b
gas OH OH HCHO (CHO)2 CHO
02 + +
N2 + + +
02 + + + +
N2 + + + ?
02 + + +
02 + + c
02 + + c
02 + + + +
N2 + + +
02 + + +
02 + + +
+ denotes positive identification of emitting species; a blank indicates
that chemiluminescence from this species was not observed.
Tentative identification; see text.
Not studied at high resolution. '
126
-------
Table 11.2
Reaction orders of emitting species in ozone and olefin in
*j
absence of oxygen, respectively.
the presence and
Emitting species
Olefin
Ethylene
Isobutene
cis-2-Butene
Isobutene
cis-2-Butene
Carrier
gas
02
N2
02
N2
02
N2
02
02
HCHO OHf
Ozone
1.04 ± 0.04 1.04 ± 0.03
1.09 ± 0.07 1.03 ± 0.05
1.06 ± 0.04 0.93 ± 0.05
0.99 ± 0.04 1.00 ± 0.01
0.95 ± 0.03 1.14 ± 0.01
0.89 ± 0.03 1.01 ± 0.03
Olefin
1.06 ± 0.05 0.97 ± 0.04
0.96 ± 0.06 1.12 ± 0.11
517- or
520-nm peak
0.39 ± 0.02b
0.97 ± 0.02C
1.00 ± 0.05b
1.10 ± 0.09°
a Errors given are least-squares standard deviations.
Tentatively identified as methylglyoxal phosphorescence.
° Assigned as glyoxal phosphorescence, (CHO)2(3Au •*• l&g> •
127
-------
20 40 60 80 100
REACTION DISTANCE (cm)
o
u-i
•> CO
CU "O
c
tt)
o
4-1 O,
CU 3 W
a fO cu
(3 I w
« CN M
4J I -
ca co
g
0 M-l
o o o
•H O
4J CO i-t
O *J
co
>
eu
o
a
l Cvl
-H ro
4-1 •
-H IT)
C
CD
M
CO
,a
n)
4J
•H -rl
a)
m 3
°f
CO CM
S ii
rH -H
O. CJ
•H O
J
C
JJ O
•rl -H
H *J
nt u
oo tt)
o a)
rH K
•H
CU
O
rH
M-l
> -rl
i-l 4J
O O
B n)
a a)
w
CM
CM cO
I
VO O
H 4J
bO
•H
128
-------
In the pressure range of 2-10 torr was approximately 1.4 for all emissions.
However, this ratio was observed to decrease with increasing oxygen or olefin
concentration. For example, at 1.8 torr total pressure, this ratio was 1.6
for formaldehyde fluorescence; on addition of 02 to a total pressure of 9.6
torr, this ratio fell to 1.0.
In N2 as carrier gas, high initial light intensities occurred, followed
by nonexponential decays as seen in Figure 11.6ii. The apparent increase in
the intensity of the Meinel bands (Figure 11.4) on the replacement of 02 by
N2 is thus due to the increased initial light intensity rather than to an
overall increase at all reaction times. While the ozone decayed even more
rapidly than in 02 at times <0.1 sec., a slower exponential decay followed.
* t
c. Relationship of OH and OH . In the cis-2-butene reaction,
* t
the relationship of the OH emission intensity to that of OH in both 02 and
N2 as diluent gas was investigated by simultaneously following both emission
intensities as a function of reaction time. The least-squares slopes of the
* t
log-log plots of OH emission intensity vs. OH intensity for each run in a
series of experiments at varying total pressure and reactant concentrations
averaged 0.97 ± 0.04 in the presence of 02 and 0.91 ± 0.06 in N2.
Similar investigations were carried out for the reactions of ethylene
and isobutene with 03 in N2 as carrier gas. While the data were more scattered
than that for the cis-2-butene reaction, the average slopes of the log-log
plots were 1.09 ± 0.08 and 1.35 ± 0.11, respectively.
Kinetics of Ozone Decay. With olefin at least a factor of 6 in excess
over 03, the rate of decay of ozone was measured under similar conditions as
those used for recording the chemiluminescent emission spectra. Varying
flows of nitrogen or oxygen were then added to a maximum total pressure of
10 torr, and the rate of ozone decay was measured at each pressure. The
results of these experiments for the cis-2-butene and isobutene reactions
are given in Table 11.3.
Although the overall uncertainties in the rate constants measured are
approximately ±30%, Table 11.3 shows that the measured rates are larger by
a factor of 2-5 in the presence of N2 as compared to those measured in 02
and decrease slightly with increasing 02 concentration, as has been observed
47-50
since.
Under these experimental conditions, the rate constant for the ethylene-
ozone reaction was estimated to be 1 ± 1 * 103 i mol~ sec" in 02 and
129
-------
Table 11.3
Rate constants for the cis-2-butene and isobutene reactions with ozone under varying conditions of total pressure and
oxygen concentrations.*
Experimental conditions
03/02 from ozonizer used; no
additional diluent added
O3/O2 from ozonizer used; O2
as diluent
O3/H2 from silica gel trap used;
02 added as diluent
03/H2 from silica gel trap used;
no additional diluent added
O3/>2 from silica gel trap used;
•2 added as diluent
cis-2-Butene
Total pressure, torr t
1.8-2.1
3.9-4.0
4.5-5.9
7.3-7.4
8.9-9.6
5.5-5.6
8.9-9.1
1.9-2.0
5.4-6.1
9.1-10.0
io'4x
mo!'1
9.2 ±
7.3 ±
6.7 ±
7.1 ±
6.3 ±
6.5 ±
6.0 t
17 ± 5
13 t 4
16 ± 5
k,
sec
2.3
2.0
1.8
2.0
1.9
1.9
1.9
Total pressure,
1.7-1.9
3.9-4.2
6.4-6.6
9.3-9.7
4.4-4.5
1.9-2.0
4.2-4.5
6.8-7.1
9.6-9.9
Isobutene
10~ x k,
torr I mol sec
1.4 ± 0.4
0.82 ± 0.28
0.63 ± 0.23
0.54 ± 0.23
0.84 t 0.29
4.2 ± 1.2
3.8 ± 1.3
3.0 ± 1.0
3.7 ± 1.1
* Error limits shown are the estimated overall errors which include both the precision of the data as well as
experimental errors
-------
5 ± 2 x io3 4 mol""1 sec"1 in N2.
The rate constants from this work are compared in Table 11.4 with litera-
ture room-temperature absolute values.6'10'42-50 The values in 02 are seen
to be in general agreement with the literature, indicating no significant
change in the kinetics with total pressure.
Stable Product Analysis. The major stable products identified by gc
from the cis-2-butene-ozone reaction at 2 to 10 torr were acetaldehyde and
2-butanone in both 02 and N2 as diluents and methyl vinyl ketone in 02 only.
Two unidentified products were also observed in 02. The acetaldehyde yield
increased relative to that of the 2-butanone with increasing pressures of 02.
These product identifications are supported by the observations of mass peaks
corresponding to all of these products in earlier photoionization mass spec-
trometer studies of this reaction.
Absolute Rates of Light Emission. The total rate of light emission from
each emitting species in the ethylene and cis-2-butene reactions with 2% 03/02
is given in Table 11.5; that for OH refers to light emission from the (9,3)
transition only.
Table 11.5
a -1 -1
Absolute rates of light emission (X,;mol sec ) for each emitting species
in the reactions of 2% 03/02 with ethylene and cis-2-butene at a total
pressure of 4.6 torr.
Olefin
Ethylene
cis-2-Butene
HCHO
1 x 1Q~4
7 x 10~2
Emitting species
OH (CHO) 2
1 x 10~4
5 x 10~3 0.7
ft dc 3t
For HCHO AND (CHO)2 , the total rate of light emission, kx(TOT)» calculated
as described in the text, is given. For OHt, the rate of light emission is
that in the (9,3) transition only, which is expected to be a function of the
reactant concentrations (see Discussion).
Temperature Rise Studies. In order to investigate the self-heating of
the gases7'8'42 during these exothermic reactions, the temperature rise was
131
-------
Table 11.4
Comparison of the rate constants obtained in this work with literature
room-temperature absolute values.
Olefin
Rate constant * 10
H mol sec
-3
Ref
Ethylene
cis-2-Butene
Isobutene
0.93 ± 0.09
1.8a
0.8
2.1 ± 0.8
1.06
1.14
1.68
1 ± 1(02)
5 ± 2(N2)
85
90(N2)
200
29
97
75.6
120
60-92(02)
130-170(N2)
8.8 ± 2.2
3.7
14 + 2
8.2
7.01
10.8
5-14(02)
30-42(N2)
46
45
6
43
47, 49
48
50
This work
10
45
6
48
49
50
This work
44
6
45
48
49
50
This work
Dynamic system, k = 1.6 ± 0.2 x 1Q3 a mol""1 sec"1.
--? -1 -1
10
i3 £ mol'1
Static system, k = 2.0
sec
132
-------
recorded at each of the five thermistors for the reaction of a series of
olefins with 03/02 or 03/N2. In the reaction with 03/02, the magnitude of
the maximum temperature rise increased in the order ethylene < propylene ~
1-butene ~ iosbutene < cis-2-butene < 2-methyl-2-butene ~ 2,3-dimethyl-2-
butene which is also the order in which the rate constants increase.6~10'42~50
Furthermore, the temperature rise increased with flow rate of either ozone
or olefin and decreased with increasing reaction time, showing that the heat
release is related to the amount of reaction occurring.
The reactions of ethylene, cis-2-butene. and isobutene with 03/N2 showed
a greater temperature rise than the corresponding reactions in oxygen, al-
though the increase in temperature was confined to shorter reaction times in
agreement with the above observations. The maximum temperature rise observed
A § A
was <7°C, which would not have a significant effect on the rate constants ' '
45,47,49,50 , ,
measured here.
Discussion
cis-2-Butene Reaction. The mechanism of ozone-olefin reactions
23
proposed by O'Neal and Blumstein is applied to the ozone-cis-2-butene
reaction in Scheme I. Included are recent modifications to the original
scheme relevant to our studies.
At a total pressure of ~2 torr, it is anticipated that (1) a-hydrogen
abstraction should occur at a rate approximately 2.5 faster than the Criegee
fragmentation assuming some stabilization of the Criegee biradical by polar-
ization of charge, (2) collisional stabilization of the excited a-keto hydro-
peroxide (reaction a) and decomposition (reactions b to e) should be competi-
tive such that both the a-keto hydroperoxide and the intermediates and stable
products predicted by reactions b to e should be observed, and (3) B-hydrogen
abstraction of the peroxy biradical will occur at a rate slower by a factor
of ~8 than the a-hydrogen abstraction path, leading to the formation of an
excited 1,2-dioxetane.
One can now compare these predictions to the experimental data for the
^is-2-butene-ozone reaction obtained in this and earlier studies.
a. pH1" Meinel Band Emission. Application of simple unimolecular
(RRK) theory to the thermochemistry predicted by O'Neal indicates that due
to the -140 kcal/mol released in the rearrangement of the biradical to acetic
acid (step f) its decomposition (step f) should occur at a rate several
orders of magnitude greater than collisional quenching at 2 torr. It is
133
-------
CO
CH3
H
O OOH
CH;
II I
,—C—CH— CH3
CH,
H
0 OOH
II I
;—CH—CH3
CH3CHO + (CH3COOHf
I b'
CH3- + -COOH
. . . . -COOH
t / / ° ° V ' C0= + H'
IM h/ ( II II )
1* /C^VCH,—c—c—H/ +
J-\rtYT I / ^ * I
L-£-* CHj
/ 9 9 V
1 H H I -I- Wf
\CH3—C—C—CH3/ "*" -
CH3OH
CH,CC- + HCO-
0 O .
CH3C—CHCH3 + -OH
6
HX
OH 00-
CH:,—CH—CH
Criegee
> CH3CHO + CHjCHOO-
split
CH3C<
0
CH3CH —* (CHjCOOHf
• CH3- + -COOH
CO, + H-
CH:1CH(OH)CHO + HCHO
CH
II
[H—C—CH,O-]*
III
Scheme I
[H-C-C--H]*
II II
o o
-------
also possible that the acetic acid formed in step b may contain enough
of the reaction exothermicity to decompose in an analogous manner. Since
12
the observed yield of CH3CHO at 2 torr is ~1, but CH3COOH was not observed,
the yield of H atoms may approach unity under our experimental conditions.
The dominant loss process for H atoms formed in step f' and possibly b1
12
under these conditions must be the well-known chemiluminescent reaction
229'38'39 with 03.
A simplified scheme for the formation and loss of OH is then given by
°3 + cis-2-butene
k3 = 7
-> PH + other products
(3)
H + 03 > OH ,,Q + 02
3 in v <9 *•
k2 = 1.6 x io10 v ~y
(2)
52
H + cis-2-butene
= 5 x IO8
Ci+Hg
(4)
53-55
03
k5 = 1.1 x IO9 (v1 = 2)
-4.4 x IO9 (v? = 9)
H02 + 02
-> 0 + 02 + OH
-> H + 202
(5)
56
OH + cis-2-butene ^
(CH3CH(OH)CHCH3)'
CH3CH(OH)CHCH3
(6)
OH
k7 == 3.4(v' = 1) sec
-16(v' = 9) sec
-1
-1
-> OH + hv
(7)
56
where the rate constants are in liters per mole second units unless other-
wise specified and 0 is the fraction of reaction 3 giving H atoms. Hence
135
-------
t e3k2[03]2 [olefin]
[OH ]
{k5[03] + k6[olefin] + k7}{k2[03] + k,, [olefin]}
where k5 = kfkM[M]/(kr + %[!!]), kf, kr, and kji are the rate constants for
the steps outlined in reaction 6 above, and in all cases k7 « k5[03].
The overall rate constant for reaction 6 for ground state OH(v' = 0)
with cis-2-butene at a total pressure of ~1 torr is ~4 x 1010 H mol
sec . ' ' The observed enhancement of the OH-propylene adduct mass
peaks as the total pressure was increased from 1 to 4 torr suggests that at
least for propylene the adduct can also decompose, one path possibly being
that back to the reactants. Because of the high room-temperature rate con-
stant, the activation energy for the reaction of OH(v' = 0) with cis-2-butene
59
is negative and hence an increase in the vibrational energy of the reacting
OH radical is anticipated to substantially increase the rate of decomposition
of the adduct, kj-, while not significantly affecting the forward rate, kf .
Hence the overall rate constant, kg, is expected to decrease as the vibrational
energy content of the OH radical increases.
An estimate of kg for vibrationally excited OH can be obtained as follows:
assuming an activation energy for the forward reaction of ~1 kcal/mol and
AHf = -28 kcal/mol for the adduct, CH3CH(OH)CHCH3, then from unimolecular
theory,61 kr - 2 x io8 sec"1^' = 7-8). Assuming kM[M] = 2 x io7 sec"1, then
1 — 1
kg ~ 4 x IO9 & mol sec , which, though very approximate, can now be applied
to the kinetics of OH formation.
In the experiments to determine the reaction order of OH with respect
to olefin, [03] > 2 x [olefin], thus k2[03] » k^ [olefin] and hence equation
III simplifies to
6k3[03] [olefin]
[OH 1
k5[03l
Using kg * 4 x io9 X, mol"1 sec , k5[03] > k6 [olefin], and hence [OHf] «
[olefin], consistent with Table II.
Similarly, when studying the dependence of OH on [03], [olefin] > 10[03]
Then k2[03] > kijolefin] and k6[olefin] > k5[03]. Hence [OHf] « [Q3], as
obseryed .
136
-------
In order to calculate the total rate of light emission from all vibra-
tional levels of OH , and hence the yield of OH1", the vibrational distribution
must be known. It was not possible in these experiments to determine this
distribution and hence only a crude estimate of the total rate of light
/:n
emission can be obtained. Garvin, et al., have shown that at 4.2 torr
total pressure in 02 and H2 the vibrational distribution of OH1" from reaction
2 is Boltzmann and corresponds to a vibrational temperature of 9250°K. How-
ever, in the presence of olefin with which OH reacts, the lower vibrational
levels may be preferentially removed leading to a higher effective vibrational
temperature. Calculating ke as a function of OH vibrational energy content
as described above, and modifying the Boltzmann distribution found by Garvin,
/TO j.
et al., for the increased removal of OH in lower vibrational energy levels,
the H atom yield per molecule of reactant consumed is estimated to be approxi-
mately 1 within order of magnitude error limits, under these experimental
conditions.
b. HCHO^A" -> *Ai). The experimentally observed first-order
kinetics of formaldehyde fluorescence are consistent with the production of
electronically excited HCHO form the decomposition of an excited 1,2-dioxetane,
step g of Scheme I.
Using k3 = 7 x 104 Jl mol~ sec" for the cis-2-butene-03 reaction at
-3 -1 -1
4.6 torr (Table 11.3) and kx(TOT) = 7 x 10 £ mol sec for formaldehyde
fluorescence (Table 11.5), the number of light quanta emitted from electronically
excited formaldehyde per molecule of reactant consumed is ~1 x 10 . These
low values are consistent with the concerted process (step g) since both the
64
decomposition of 1,2-dioxetanes to produce the first excited singlet state
and HCHO fluorescent emission are inefficient processes.
Other ^sources of HCHO(lA") such as
CH30 + OH —> HCHO + H20 (8)
cannot be definitely excluded, particularly at the low pressures and relatively
high OH concentrations present under these conditions. Reaction 8 is thought
to be responsible for the weak formaldehyde fluorescence observed as a general
phenomenon in the low-temperature gas-phase oxidation of organic compounds.
Although OH radicals are known to be present in the cis-2-butene-ozone reaction,
no CH30 radicals were detected in the PMS studies.12 This does not preclude
137
-------
the production of CE^O since if its loss is by reaction with olefin with a
rate constant of 107-108 H mol sec , then its steady state concentration
would have been below the detection limits.
c. Glyoxal Phosphorescence. Although glyoxal phosphorescence
has been observed from the recombination of HCO radicals, this seems highly
68
unlikely in the presence of excess 02 which reacts rapidly with HCO. In
addition, this recombination, which is second order in HCO, would not be
kinetically consistent with both the first-order dependence of light Inten-
sity on reactants and with the direct production-of HCO from the ozone-olefin
adduct via steps c and c'.
The observation of glyoxal phosphorescence in the presence of 2 Torr
Q f.
of 02, and indeed at higher pressures, is rather surprising. One ener-
getically possible route involves an alternate mode of decomposition of the
1,2-dioxetane formed by 3-H abstraction
OH
I
[CH3-CH-CH-CH2] —> C2H50- + [H-C-CH20«] (9)
I I M
0—0 0
followed by reaction of the excited alkoxy radical with 02.
[H-C-CH20«] + 02 —> H02 * [H-C-C-fl] (10)
0 00
While (9) and (10) are speculative, it appears that the partitioning of energy
in these reactions may result in sufficient energy input to the glyoxal to
cause electronic excitation. Furthermore, this mechanism is consistent with
the observed kinetic isotope effect (Figure 11.5), the first-order kinetics
of the light emission, and the observation of glyoxal phosphorescence at
36
pressures approaching atmospheric.
d. Electronically Excited OH. OH has been observed in the
H + 03 reaction as a product of energy pooling.
OHt(X2Tri) + OHf(xV) —> OH(A2Z+) + OH(X2Tri) (11)
However, the width of the (0,0) band envelope of OH(A2£+ ->• X2ir ) from
138
-------
reaction 11 was greater than that from the ozone-olefin reactions. In addition,
the observed first-order dependence of electronically excited OH* on the
vibrationally excited OH1" radical rules out reaction 11 as the source of OH*.
One possible source of electronically excited OH* radicals in this
system involves 0 atom reactions. Thus, Becker, Kley, and Norstrom70 in
their studies of OH chemiluminescence in the 0-C2Ek flame observed rota-
tionally cool emission from OH . The reactions of an 0 atom with either an
H atom or an excited HCO radical to produce OH* were suggested.
0 + H — > OH (12)
0 + HCOf — > OH* + CO (13)
A mode of production of 0 atoms in the ozone-olefin system is the
decomposition of 03 by OH radicals with v1 > 3.
03 — > 0 + 02 + OH (5b)
*
The OH emission intensity is strongest for the ethylene reaction,
* t
relative to the HCHO and OH emissions, supporting an 0 atom precursor as
the relative rates of reaction of 0(3P) with ethylene, isobutene, and cis-2-
butene are 1:25:24.
e. Effects of Oxygen. In nitrogen as diluent, the initial steps
of the ozone-cis-2-butene reaction as outlined above must remain the same.
However, the secondary reactions of radicals which react rapidly with 02
will be strongly dependent on its presence. For example, CH3 and HCO which
react with 02 under these conditions, will, in N2 as diluent, react with 03.
CH3 + 02 + M - 12 - =r? CH3°2
kut - 1011 H2 mol sec
o2
= 3.4 x 109 £ mol sec
co <15>68
The products of their reactions with 03 are likely more reactive than the
CH302 and H02 produced in reactions 14 and 15, respectively. For example,
j j 66,72
if excited CH30 and HCO are formed, H atoms could be produced.
139
-------
CH3 + 03 —> 02 + [CH30]* (16)
—> H + HCHO
HCO + 03 —> 02 + [HC02] (17)
—> H + C02
Therefore in the absence of 02 one anticipates increased rates of 03
decay due to secondary radical attack and possibly increased Meinel band
emission intensity if H atoms are indeed produced in reactions 16 and 17.
This is in qualitative agreement with our experimental observations
that: (1) the measured rate constant for 03 decay in N2 as diluent is larger
than in 02, (2) a rapid loss of 03 occurs in the first <0.1 sec which is
accompanied by intense light emission, particularly of OH and OH , and (3)
a measurable temperature rise occurs, particularly in the vicinity of the
mixing jet.
The nonexponential decay of each of the light emissions in N2 as diluent
may then be due to a change in the relative rates of competing reactions
such as (5) and (6) over the time period studied, since reaction 6 is a
termination step in the chain formed by reactions 2 and 5c. Similar processes
must also occur to some extent in 2 torr of 02 since the emitting species
generally show first-order behavior on 03 (Table 11.2) and yet the rate of
decay of light intensity is generally greater than the rate of 03 decay.
Increasing the pressures of 02 up to 10 torr appears to further quench one
or more radical precursors to the chemiluminescence since the ratio (rate of
decay of light intensity)/(rate of 03 decay) approaches unity at ~10 torr of
02.
f. Intermediates and Products. Many of the intermediates and
stable products predicted by Scheme I were observed either by photoionization
12
mass spectrometry, gc, or both. Thus, HCO and H02, possibly from reactions
£Q I O
c, c1, and 15, were observed. In addition, HCHO, CH3OH, CH3CHO, CH2CO,
and an ozone-olefin adduct believed to be the a-keto hydroperoxide,
CH3CH(OOH)COCH3, were detected.12
On the other hand, 2-hydropropanal which should have been formed in
equal quantities to HCHO (step g) was not seen. In addition, while CH302
12
and C2H50 radicals were observed, their nonsteady state behavior suggested
140
-------
that they were not primary reaction products but were formed in the mass
spectrometer. Hence, their production is uncertain, although reactions 14
and 9 may be possible contributors.
In the PMS studies, ketene was also observed as a major product.12 Its
formation can be rationalized in terms of a "classical" rearrangement of the
Criegee zwitterion as suggested by Scott, Hanst, and co-workers.5'6'75
CH3CHOO —> CH2=C=0 + H20 (18)
Significant quantities of methyl vinyl ketone and 2-butanone were
observed by gc and peaks corresponding to these products were observed by
12
photoionization mass spectrometry. The identification of 2-butanone by gc
provides the best explanation, of the m/e 72 peak although it might contain
some contribution due to methylglyoxal (reaction c).
The 2-butanone may be due to the reaction of OH with cis-2-butene. The
mechanism of this reaction is not clear, although an OH-olefin adduct has
been observed mass spectrometrically. Preliminary studies of the OH-cis-
2-butene reaction in the presence and absence of oxygen indicate that both
2-butanone and acetaldehyde are formed in significant yields, with the ratio
of acetaldehyde to 2-butanone increasing on the addition of oxygen, consis-
tent with our observations.
Radical abstraction from cis-2-butene will form the radical CH2CH=CHCH3.
12
It was suggested in earlier studies that in the presence of 02 this radical
might subsequently react with 02 to form methyl vinyl ketone and hence a peak
at m/e 70.
°2 "
CH2CH=CHCH3 <—> CH2=CHCHCH3 ->—> CH2=CHCCH3 + other products (19)
Formation of methyl vinyl ketone by reaction 19 is supported by the fact that
it was below the gc detection limit in the cis-2-butene-03/N2 system but was
observed on 02 as diluent.
Ethylene and Isobutene Reactions. Formaldehyde fluorescence and OH
Meinel band emissions are common to all three olefin reactions studied.
Possible routes to the formation of electronically excited formaldehyde which
are consistent with the observed first-order kinetics involve a-H abstraction
. , 23,51
and 0.-H abstraction, respectively.
141
-------
H2C=CH2 + 03 —>
00 00
II II
CH2— CH2 < > CH2— CH2
a-H
abstraction
0
" * *
[HCCH2OOH] —> HCOOH + HCHO
(20)
CH3
CH3
>C=CH2 + 03
0 0
I I
CH3 — C i C — H ^
CH3 H
0«
o 6
I I
CHo C C H
I I
CH3 H
0-H
abstraction
0
CH3 —
)
: — CH2OH — >
3H2-
CH3
CH2OH
1
1
— C — CH2
1 1
1 1
0 — 0
0
CH3 — C — CH2OH + HCHO + hv
(21)
As in the cis-2-butene-ozone reaction, efficient H-atom production and
hence relatively strong OH Meinel band emission are anticipated in both of
these reactions at 2 torr from the decomposition of excited HCOOH formed by
the Criegee path.
HCHOO-
°
\ /
—> c
H 0-
[HCOOH] —> H- + -C02H (22)
—> C02 + H»
23 51
While in the isobutene reaction, a proposed ' route to the formation
of methylglyoxal is the a-H
142
-------
CH3
/
CH3
03
CH3 — C C — H
CH3 H
CHo —
0
I
c
C — H
a-H
abstraction
CH3 H
OOH 0
CH3 — C
CH3
CH
> CH3COCH3 + HCOOH
0 0
\U
> HCCCH3 + CH3OH
0-
I
> (CH3)2CCHO + -OH
(23)
abstraction, reaction 23b; the kinetics of the'emission (Table 11.2) suggest
a more complex mechanism than reaction 23b. The apparent involvement of 02
is also difficult to rationalize by this scheme, as is the observation of
excited methylglyoxal from the 2-methyl-2-butene and 2,3-dimethyl-2-butene
reactions. However, in the alternate mode of decomposition of the 1,2-di-
oxetane formed by g-hydrogen abstraction, analogous to that proposed above
for the cis-2-butene, reactions 9 and 10 may give electronically excited
methylglyoxal.
X1TT
CH3
CH2OH
C. . i. f^ii
Lil2
0. /\
(J J
*
rCH3C
L 0
+ 'CH2OH
(24)
CH3CCH2O
0
07 > H02 + [CH3CCHO
I
L o .
(25)
Reactions 24 and 25 would explain why 02 is necessary to produce excited
methylglyoxal. In addition, one anticipates a decreased ratio of the intensities
143
-------
of methylglyoxal phosphorescence to formaldehyde fluorescence in the iso-
butene-ds reaction, due to kinetic isotope effects, as is observed (Figure
11.511).
The effects of replacing the 02 diluent by % are expected to be similar
to those in the cis-2-butene reaction, consistent with experiment.
Atmospheric Imp1leations. In the formation of photochemical smog, the
rate of oxidation of NO to N02 in excess of that due to 0 atom and 63 reactions
77-79
is currently thought to be due at least in part to the H02 reaction
H02 + NO —> N02 + OH (26)
H02 may then be regenerated in a chain process involving the OH radical
OH + CO —> H + C02 (27)
H + 02 + M —> H02 + M (28)
Hence any reaction producing H atoms or OH in the atmosphere may contribute
to the NO to N02 oxidation. OH ha
lending support to this postulate.
80
to the NO to N02 oxidation. OH has recently been detected in ambient air,
144
-------
REFERENCES
1. L. Long, Che. Rev., 27, 437 (1943).
2. P. S. Bailey, Chem. Rev. 58, 925 (1958), and references therein.
3. P. S. Bailey, 161st National Meeting of the American Chemical Society,
Los Angeles, Calif., March 29-30, 1971, PETR No. 11; Advan. Chem. Ser.
No. 112 (1972).
4. R. W. Murray, Accounts Chem. Res., 1, 313 (1968), and references therein.
5. W. E. Scott, E. R. Stephens, P. L. Hanst, and R. C. Doerr, Proc. Amer.
Petrol. Inst., Sect. 3, 171 (1957).
6. P. L. Hanst, E. R. Stephens, W. E. Scott, and R. C. Doerr, "Atmospheric
Ozone-Olefin Reactions," The Franklin Institute, Philadelphia, Pa., 1958.
7. T. Vrbaski and R. J. Cvetanovic, Can. J. Chem., 38, 1053, 1063 (1960).
8. Y. K. Wei and R. J. Cvetanovic, Can. J. Chem., 41, 913 (1963).
9. W. B. DeMore, Int. J. Chem. Kinet., .1, 209 (1969).
10. R. A. Cox and S. A. Penkett, J. Chem. Soc., Faraday Trans. 1, 69, 1735
(1972).
11. B. J. Finlayson, J. N. Pitts, Jr., and H. Akimoto, Chem. Phys. Lett.,
12, 495 .(1972).
12. R. Atkinson, B. J. Finlayson, and J. N. Pitts, Jr., J. Amer. Chem. Soc.,
j>5, 7592 (1973).
13. E. Briner and P. Schnorf, Helv. Chem. Acta, 12, 154, 151 (1929).
14. E. Briner and R. Meier, Helv. Chem. Acta, J.2, 529 (1929).
15. G. W. Nederbragt, A. Van der Horst, and J. Van Duijn, Nature (London),
206, 87 (1965).
16. G. J. Warren and G. Babcock, Rev. Sci. Instrum., 41, 280 (1970).
S
17. J. A. Hodgeson, B. E. Martin, and R. E. Baumgardner, 160th National
Meeting of the American Chemical Society, Chicago, 111., Sept. 13-18,
1970, WATR No. Oil.
18. J. A. Hodgeson and R. K. Stevens, Abstracts, 161st National Meeting of
the American Chemical Society, Los Angeles, Calif., March 28-April 2, 1971,
Anal. No. 069.
19. J. A. Hodgeson, B. E. Martin, and R. E. Baumgardner, paper 77 presented
at the Eastern Analytical Symposium, New York, N. Y., 1970.
145
-------
REFERENCES (cont.)
20. W. A. Rummer, J. N. Pitts, Jr., and R. P. Steer, Environ. Sci. Technol.,
Jj, 1045 (1971).
21. J. N. Pitts, Jr., B. J. Finlayson, H. Akimoto, W. A. Rummer, and R. P.
Steer, Int. Symp. Identification Meas. Environ. Pollut., [Proc.], 32
(1971).
22. J. N. Pitts, Jr., W. A. Rummer, R. P. Steer, and B. J. Finlayson, Advan.
Chem. Ser., No. 113, 246 (1972).
23. H. E. O'Neal and C. Blumstein, Int. J. Chem. Rinet., _5, 397 (1973).
24. A. Fontijn, C. B. Meyer, and H. I. Schiff, J. Chem. Phys., ^0, 64 (1964).
25. A. N. Wright and C. A. Winkler, "Active Nitrogen," Academic Press, New
York, N.Y., 1968, pp 74-75, and references therein.
26. L. G. Anderson, C. S. Parmenter, H. M. Poland, and J. D. Rau, Chem.
Phys. Lett., 8, 232 (1971).
27. J. T. Yardley, G. W. Holleman, and J. I. Steinfeld, Chem. Phys. Lett.,
10, 266 (1971).
28. J. T. Yardley, J. Chem. Phys., .56, 6192 (1972).
29. A. B. Meinel, Astrophys. J., Ill, 555 (1950).
30. G. A. Cook, A. D. Riffer, C. V. Rumpp, A. H. Malik, and L. A. Spence,
Advan. Chem. Ser., No. 21, 44 (1959).
31. P. N. Clough and B. A. Thrush, Chem. Ind. (London), 1971 (1966).
32. R. Spence and W. Wild, J. Chem. Soc., London, 338 (1935).
33. G. W. Robinson, J. Chem. Phys., 22, 1147, 1384 (1954).
34. W. Holzer and D. A. Ramsay, Can. J. Phys., 48, 1759 (1970).
35. R. Y. Dong and D. A. Ramsay, Can. J. Phys., 51, 1491 (1973).
36. J. N. Pitts, Jr., and D. A. Hansen, unpublished data.
37. J. T. Yardley, A. V. Pocius, and C. Dykstra, personal communication, 1973.
38. J. D. McRinley, Jr., D. Garvin, and M. J. Boudart, J. Chem. Phys., 23,
784 (1955).
39. P. E. Charters, R. G. MacDonald, and J. C. Polanyi, Appl. Opt., 10,
1747 (1971), and references therein.
40. B. J. Finlayson, J. S. Gaffney, and J. N. Pitts, Jr., Chem. Phys. Lett.,
17, 22 (1972).
146
-------
REFERENCES (cont.)
41. R. Livingston in "Techniques of Organic Chemistry," 2nd ed., S. L.
Friess, E. S. Lewis, and A. Weissberger, Ed., Interscience, New York,
N.Y., 1961, Chapter II.
42. H. E. Smith and R. H. Eastman, J. Amer. Chem. Soc., 83, 4274 (1961).
43. R. D. Cadle and C. Schadt, J. Amer. Chem. Soc., 74, 6002 (1952).
44. E. A. Schuck, G. J. Doyle, and N. Endow, Air Pollution Foundation,
Report No. 31, San Marino, Calif., I960.
45. J. J. Bufalini and A. P. Altshuller, Can. J. Chem., 43_, 2243 (1965).
46. D. H. Stedman, C. H. Wu, and H. Niki, J. Phys. Chem., 77, 2511 (1973).
47. J. T. Herron and R. E. Huie, J. Phys. Chem., 78.* 2085 (1974).
48. S. M. Japar, C. H. Wu, and H. Niki, J. Phys. Chem., _78_, 2318 (1974).
49. R. E. Huie and J. T. Herron, Symposium on "Chemical Kinetics Data for
the Lower and Upper Atmosphere," Warrenton, Virginia, Sept. 16-18, 1974.
Int. J. Chem. Kinet., Symposium No. 1, 165 (1975).
50. K. H. Becker, U. Schurath and H. Seitz, Int. J. Chem. Kinet., j>, 725
(1974).
51. H. E. O'Neal, personal communication, 1973.
52. L. F. Phillips and H. I. Schiff, J. Chem. Phys., .37, 1233 (1962).
53. E. E. Daby and H. Niki, J. Chem. Phys., 51, 1225 (1969).
54. J. A. Cowfer, D. G. Keil, J. V. Michael, and C. Yeh, J. Phys. Chem.,
T5, 1584 (1971).
55. E. E. Daby, H. Niki, and B. Weinstock, J. Phys. Chem., 7j>, 1601 (1971).
56. R. N. Coltharp, S. D. Worley, and A. E. Potter, Appl. Opt., 10, 1786
(1971).
57. E. D. Morris, Jr., D. H. Stedman, and H. Niki, J. Amer. Chem. Soc., 93,
3570 (1971).
58. E. D. Morris, Jr., and H. Niki, J. Phys. Chem., 75.* 3640 (1971).
59. R. Atkinson and J. N. Pitts, Jr., J. Chem. Phys., 63, 3591 (1975).
60» AHf° was calculated using Benson's group additivity method.
147
-------
REFERENCES (cont.)
1'
61. S. W. Benson, "Thermochemical Kinetics," Wiley, New York, N.Y., 1968.
kr = A((E - E0)/E)s-l where A = 1013 sec"1, (s - l)eff - 24, E0 =
37 kcal/mol, and E = EoRt + 36 kcal/mol, where EQ^T is the vibrational
energy content of the OHt radical.
62. It is likely that diatomics (02 or N2) will not stabilize the excited
adduct molecule with unit efficiency, hence decreasing the calculated
value of ks. The existence of alternate decomposition paths for the
excited adduct will increase kg- Unfortunately, at the present time,
no information is available on the fate of the excited adduct.
63. D. Garvin, H. P. Broida, and H. J. Kostkowski, J. Chem. Phys., 32. 880
(1960).
64. N. Turro and P. Lechtken, J. Amer. Chem. Soc., 94, 2887 (1972).
65. J. H. Knox in "Photochemistry and Reaction Kinetics," P. G. Ashmore,
F. S. Dainton, and T. M. Sugden, Ed., Cambridge University Press, London,
1967, pp 250-286.
66. K. L. Demerjian, J. A. Kerr, and J. G. Calvert, Advan. Environ. Sci.
Technol., .4, 1 (1974).
67. D. B. Hartley, Chem. Commun., 1281 (1967).
68. N. Washida, R. I. Martinez, and K. D. Bayes, Z. Naturforsch. A, 29,
251 (1974).
69. H. P. Broida, J. Chem. Phys., _36, 444 (1962).
70. K. H. Becker, D. Kley, and R. J. Norstrom, 12th International Symposium
on Combustion, The Combustion Institute, Pittsburgh, Pa., 1969, p 405,
and references therein.
71. W. B. DeMore, J. Chem. Phys., 46, 1813 (1967).
72. R. J. Cvetanovic, Advan. Photochem., _!, 115 (1963).
73. N. Basco, D. G. L. James, and F. C. James, Int. J. Chem. Kinet., ^_, 129
(1972).
74. P. Gray, R. Shaw, and J. C. J. Thynne, Progr. React. Kinet., 4_, 63 (1967).
75. Reaction 18 may proceed via the excited CHsCOOH formed from the rearrange-
ment of the Criegee biradical CH3CHOO» since the pyrolysis of CH3COOH
leads to the formation of CH2CO and H20: P. G. Blake and G. E. Jackson,
J. Chem. Soc. B, 94 (1964), and references therein.
76. R. J. Cvetanovic, personal communication, 1973.
77. J. Heicklen, K. Westberg, and N. Cohen, publication No. 115-69, Center
for Air Environmental Studies, University Park, Pa., 1969.
148
-------
REFERENCES (cont.)
78. K. Westberg, N. Cohen, and K. W. Wilson, Science, 171, 1013 (1971).
79. D. H. Stedman, E. D. Morris, Jr., E. E. Daby, H. Niki, and B. Weinstock,
160th National Meeting of the American Chemical Society, Chicago, 111.,
Sept., 1970, WATR No. 26.
80. C. C. Wang and L. I. Davis, Jr., Phys. Rev. Lett., 32, 349 (1974).
149
-------
12. PHOTOrONIZATION MASS SPECTROMETER STUDIES OF GAS PHASE OZONE-OLEFIN REACTIONS
Ozone-olefin reactions are known to be involved in the formation of
photochemical smog. However, because of the paucity of data on the inter-
mediates and mechanisms of these reactions, little is known about the detailed
interaction of these intermediates and products with other atmospheric con-
stituents. While these reactions have been postulated to act as a source of
chain carrying OH radicals, it in only recently that any evidence for their
production from low pressure ozone-olefin reactions has been obtained.
Clearly, identification of the intermediates and labile as well as stable
products is necessary in order to permit the formulation of a detailed reaction
mechanism. Photoionization mass spectrometry offers several advantages in
1-3
this regard. First, the use of a low energy ionization source makes it
feasible to detect radical species produced in the reaction, free of the
fragmentation problems encountered wifch electron beam ionization. In addition,
variation of the photoionizing resonance lamps allows selective photoioniza-
tion of the species.
In this work it was therefore our purpose to detect and study the kinetics
of the intermediates and products of the reaction of ozone which ethylene,
cis-2-butene, and isobutene in order to clarify the chemiluminescent reaction
steps and their relationship to the overall mechanism.
Experimental. The exerimental system has been described in detail
previously. The fast flow system (i.d. = 11 mm) had a moveable inlet jet
(o.d. = 4 mm) through which the olefin was added axially. By throttling the
pumping system, the reaction time pouj.d be varied over the range -1-130 msec.
The pressure at the midpoint of the moveable inlet region was monitored by
an MKS Baratron pressure gauge.
The gas in the flow system was sampled via a pinhole into the photo-
ionization chamber of a quadrupole mass spectrometer. The ions were detected
by a Cu-Be electron multiplier and ion counts were displayed digitally with
provision for varying counting intervals. Xe, Kr, or Ar microwave powered
4
resonance lamps were used as photoionization sources. The ionizing energies
of these lamps are shown in Figure 12.1 together with the ionization poten-
tials of some stable molecules and free radicals of interest in this study;
selective photoionization can therefore be achieved by the appropriate use
of these lamps.
' 150
-------
All experiments were carried out at total pressures of approximately
2 torr. Typical partial pressures of reactants were 0.04 torr of 03 and
0.3 torr of the olefin. Approximately 2% 03 in 02 flows were obtained by
the passage of Matheson ultrahigh purity grade 02 (>99.99%) through a
Welsbach Model T-408 ozonator. The olefins were all of research grade
purity (>99.8%) and were used without further purification. Fully deuterated
2-butene (mixture of cis and trans) was obtained from Merck, Sharp and Dohme
with stated D atom purity levels >99%. A bypass system enabled the 03-02
gas flow to be replaced by a pure 02 stream.
Results. In all experiments, background counts were taken (typically
an average of several 10-sec counts) with 02-olefin mixtures, and these were
subtracted from the ion counts obtained in the presence of 03. All counts
were corrected for decay of the lamp intensity and small fluctuations in
olefin concentration by monitoring a suitable impurity peak or the parent
peak (generally C2Hit which occurred as an impurity in the cis-2-butene and
isobutene).
The m/e scale was corrected for any nonlinearity using parent peaks of
authentic compounds. The Kr lamp was used in the majority of experiments,
while the Xe resonance lamp was used to selectively photoionize and hence
identify the radicals C2H5 (m/e 29), CH3CO, and/or C3H7 (m/e 43). The Ar
lamp was used for the species H02, H202, HCHO, CH3OH, and HCOOH which are
not photoionized by either the Xe or Kr lamps (Figure 12.1).
A. cis-2-Butene. Table 12.1 shows the observed product mass numbers
together with the lamps with which they were observed at reaction times of
-130 msec. Column 3 of Table 12.1 gives the possible species for each mass
observed. Positive identification was obtained in some cases by a combination
of deuteration, kinetic behavior, and the use of various photoionization
lamps. For example, with the Kr lamp, a peak was observed at m/e 29- In
the present system this could be due to HCO or C2H5. On deuteration, the
m/e 29 peak shifted to m/e values of 30 and 34, indicating both HCO and C2H5.
Replacement of the Kr lamp by an Xe lamp showed a mass peak at m/e 29 confirm-
ing the presence of C2H5 (see Figure 12.1); furthermore, the kinetic time
dependence was that expected for a radical intermediate as shown in Figure
12.3b. Column 4 of Table 12.1 thus shows the species identified by procedures
similar to that described above.
For the peaks of m/e 29-47 inclusive, and for 104, their time dependent
151
-------
14
1 r
13
12
> H
-------
Table 12.1
Observed peak mass numbers and their assignments
Obsd mass
no. m/e
29
30
32
33
34
42
43
44
45
46
47
58
70
72
73
75
88
89
90
91
Lamp
Kr, Xe
Ar
Ar
Ar
Ar
Kr
Ar, Xe
Kr
Kr
Ar
Kr
Kr
Kr
Kr
Kr
Kr
Kr
Kr
Kr
Kr
A
Possible species
cis-2-Butene
HCO, C2Hs
HCHO
CH3OH, 02(1A)
H02
H202
CH2CO, C3H6
CH3CO, C3H7
CH3CHO
C2H50, HC02
HCOOH
CH302
(CHO)2,CH3COCH3
C2H5CHO
CH3COCH=CH2
CH3COC2H5 or
0
/ \
CH3CH-CHCH3
OH
1
1 .
CijHgO, CH3CH-CHCH3
C3H702, CH3C(0)00
CH3COCHOHCH3,
CH3CH(OCH3)CH2CH3
OOH
1
1
C4H902, CH3CH-CHCH3
OOH
1
1
CH3CHCH2CH3
o
Definite
identification
HCO and C2H5
HCHO
CH3OHC
H02
H202
CH2CO
d
CH3CHO
C2H50
HCOOH
CH302
(CHO)2
a
a
a
a
a
a
a
Means of ,
identification
D, L, K
L
K
L, K
L
D, K
D, K
D
L
D
D
v
104
107
30
43
30
42
Kr
Kr
Ar
Kr, Xe
Ar
Kr
0^0
CH3CH—CHCH3,
CH3CH CH-CH3,
0-0
0 OOH
II I
CH3C-CHCH3
C3H704, C2H305 Ethylene
HCHO
CH3CO, C3H7 Isobutene
HCHO
CH2CO, C3H6
HCHO
CH3COe
HCHO
L, K
L, K
L, K
153
-------
Table 12.1 (cont.)
Obsd mass
no. m/e
43
44
58
59
Lamp
Kr, Xe
Kr
Kr
Kr
Possible species
cis-2-Butene
CH3CO, C3H7
CH3CHO
CH3COCH3, (CHO)2
C2H5CHO
Definite
identification
d
CH3CHO
f
Means of ,
identification
3 For high mass numbers (m/e £70) there are several possible structures which cannot be
differentiated; typical structures are given which include those products expected
from the reaction scheme given in the text (see Discussion).
D = deuteration, L = selective use of lamps, K = kinetic behavior (see text).
C On the basis of published rate constants for Oa(1A) + 03 -»• 0 + 202 [F. D. Findlay and
D. R. Snelling, J. Chem. Phys., 54, 2750 (1971)] the kinetic behavior observed is
that expected of CH3OH rather than of 02(1A).
On use of Xe lamp with sapphire window (see text), this mass number was not observed.
would have to be formed by radical attack on CjHi,, and these reactions are
expected to be slow compared with radical reactions with 03 and 02>
e 6—8
On the basis of the known chemistry of isobutene-03 reactions, ~ acetone is by far
the most likely product.
154
-------
behavior and their dependence on olefin concentration were determined.
Figures 12.2 and 12.3 show the observed time dependence on these product
peaks. For the peak at m/e 104, deuteration showed the presence of eight
H atoms, hence the formula CHH803 (Table 12.1). The observed linear depen-
dence on both time and olefin concentration shown in Figure 12.4 indicated
that this was a stable molecule over this time range. In general, the
kinetic behavior confirms the identifications of Table 12.1; i.e., stable
products increase linearly with time, while radical intermediates show an
approach to a steady state. However, in certain cases, i.e., m/e ratios of
43, 45, and 47, anomalous behavior is observed as shown in Figure 12.2b.
Experiments using a Xe lamp with a sapphire window (transmitting only the
8.44-eV line) gave no evidence for an m/e 43 peak. This observation, in
conjunction with the observed kinetics, strongly suggests that this peak,
and by analogy possibly those at m/e 45 and 47, arises from fragmentation in
the mass spectrometer of more complex molecules whose ionization potentials
are in the range 9.57 > IP > 8.44eV.
It is seen from Figure 12.3a that the HCOOH and H202 signals tend to
level off at the longer time regions. This may be due to diffusion to and
subsequent absorption on the wall. Average diffusion times to the walls were
calculated to be ~6 msec.
The signals obtained here, although corrected for decay of lamp intensity
and fluctuations in the olefin pressure within a given series of runs, are
uncorrected for relative photoionization efficiencies of the species and
differences in photon intensities of the resonance lines used. Thus, for
example, although the signals from CH2CO and CH3CHO are similar (Figure 12.3a),
CH3CHO is photoionized only by the higher energy Kr line (Figure 12.1), which
is typically present to the extent of 22% of the combined 10.64- and 10.03-eV
Kr lines. Therefore, assuming equal photoionization efficiences, the yield
of CH3CHO becomes approximately five times that of CH2CO. Using a rate con-
stant of 9 x lo4 £ mol"1 sec"1 for the reaction of ozone with cis-2-butene
9
as measured under similar experimental conditions, the amount of ozone
reacted over the time interval of 130 msec will be -15% or 3 x 10~7 mol r"1.
With unit stoichiometry10'11 this corresponds to 3 x 10~7 mol JT1 cj.s-2-
butene reacted or -2% of the initial £is-2-butene concentration. From the
photoionization mass spectrometer sensitivity, it can be calculated that the
observed CH3CHO ion counts correspond (within a factor of 2) to 3 x 10~7 mol ST1
155
-------
400
300
1 200
o
a>
10
Q.
«n 100
o
o
0
- a.
80
60
40
20
0
- b.
CH3CHO
HCO+C2H5
0
6 8 10
TIME (msec)
12
14
Figure 12.2 Kinetic behavior of some species observed in the cis-2-
butene-ozone reaction using a krypton lamp: (a) HCO +
C2H5 (m/e 29), CH2CO (m/e 42), and CH3CHO (m/e 44);
(b) mass peaks m/e 43, 45, and 47.
156
-------
120
100
80
60
§ 40
- a,
TOT
-0.2-0.3 torr
0.05 torr
HCHO
CH3OH
40
60 80 100
TIME (msec)
120
140
Figure 12.3 Kinetic behavior of some species observed in the cis-2-butene-
ozone reaction: (a) HCHO (m/e 30), CH3OH (m/e 32), and HCOOH
(m/e 46) using an argon lamp; (b) C£H5 (m/e 29) using a xenon
lamp, H02 (m/e 33) and H^Z (ffi^e 3^) usin8 an argon lamp.
157
-------
Ul
00
CD
CO
10
8
c
o
o
o>
in
8. 6
in
I 4
O
0
PTOT
P
0
(b)
2 torr
0.2 - 0.3 torr
I 2 3
[OLEFIN] (mole liter"1)
20
40 60
TIME . (msec)
80
100
Figure 12.4 Dependence of signal at m/e 104 on (a) olefin concentration and (b)
time, using a krypton lamp in the cis-2-butene-ozone reaction.
-------
of CH3CHO after a 130-msec reaction. Hence, within these error limits, ~1 mol
of CH3CHO is produced per mole of £is-2-butene reacting, which is in agreement
with previous workers. '
B- Ethylene. The product mass peaks observed at reaction times of
-130 msec are shown in Table 12.1. Again, use of the various resonance lamps
provided confirmatory indentifications. Figure 12.5 shows the time-dependent
behavior of the m/e 43 peak (presumed to be CH3CO) using a Xe lamp and of
ECHO. The CH3CO shows the kinetic behavior expected for a radical intermediate,
attaining a steady state. Furthermore, the m/e 43 peak remained, within
experimental error (13 ± 11 counts above a background of -10 counts per 10
sec), on use of a Xe lamp with a sapphire window. Deuteration was not feasible
because of the generally very weak signals.
*-•• Isobutene. Table 12.1 shows the mass peaks observed at reaction
times of -130 msec, together with the lamps with which they were observed.
The kinetic time dependence of the mass peaks is shown in Figure 12.6. As
in the case of 03 + cis-2-butene the m/e 43 peak showed anomalous behavior,
increasing with time (Figure 12.6b). Use of a Xe lamp with a sapphire window
showed no evidence for m/e 43, again suggesting that it arises from fragmen-
tation of a more complex molecule in the mass spectrometer. Similarly, as
seen in Figure 12.6b, the time dependence of m/e 42 (CH2CO or C3Hg) suggests
that it also arises by fragmentation; stable products are expected to increase
linearly from a zero ordinate.
No identification could be made for m/e 59, and deuteration was not
attempted because of the low counts obtained and because of interference that
would have occurred from i-C^Ds at m/e 64.
Subsidiary experiments were carried out with acetone and acetaldehyde
to determine if CH3CO arose from fragmentation of these molecules in the mass
spectrometer. A small amount of fragmentation was observed. The ratio of
the signals at m/e 58:43 was approximately 30:1. Thus this cannot account
for the m/e 43 signals observed in this work.
Discussion. The majority of the data was obtained for the cis-2-butene-
ozone reaction because of its higher constant ' and therefore larger
observable percentage reaction as compared to the ethylene and isobutene
reactions.
The stable products observed at these short reaction times are compared
in Table 12.2 with selected previous product studies reported in the literature.
X
159
-------
T3
£±
O
O
-------
o
0 m/e = 42
40 60 80 100 120 140 160
TIME (msec)
Figure 12.6 Kinetic behavior of some species in the isobutene-ozone reaction:
(a) HCHO (m/e 30) (argon lamp) and CH3COCH3 (m/e 58) (krypton
lamp); (b) mass peaks m/e 42, 43, and 59 using a krypton lamp.
161
-------
Table 12.2
Stable products observed in the reaction of ozone with ethylene, isobutene, and cis-2-butene
I—1
ON
ro
Olefin ECHO CH3CHO CH3COCH3 CH2CO
C2H6 0.019
0.53
Major
i-CitHg 0.53
0.008 0.52
Major Major
ci8-Ci»H8-2 1.02
0.98
Major Major Major
i-C3H7CHO MFb (CHO)2 C3H6 C3He HCOOH
0.25
<0.03
0.005 0.005 0.005 0.009 0.66
0.023 0.20
Minor Minor
CU3OH H202 CO 2
0.13
0.013 0.37
0.35
0.152 0.42
0.35
Minor Minor
CO H20 Ref
6
0.88 0.88 12
This work
6
7
This work
6
7
This work
a Product yields are given as the number of moles of product per mole of ozone consumed.
b MF - methyl formate.
-------
It is seen that there is good agreement with the major products observed:
formaldehyde from ethylene, acetone from isobutene, and acetaldehyde from
cis-2-butene. Also included in Table 12.2 are qualitative estimates from
this work of the product yields using the observed ion counts as a measure
of concentration. In certain cases minor amounts of products not listed
in Table 12.1 may have been present. For example, in the cis-2-butene-
ozone reaction, propylene could have been present in small amounts along
with much larger quantities of ketene. In this work, several of the
anticipated products such as CO, C02, H2, and H20 cannot be detected with
photoionization mass spectrometry (Figure 12.1). In addition, it was not
13
possible to detect the OH radical which is known to be present in the low
pressure 03-olefin reactions.
On the basis of the observed products and their kinetics in the cis-2-
butene oxidation, some of the reaction steps can be tentatively identified
as shown below. Reactions which are slow under these experimental conditions
compared to other available reaction paths have been omitted; for example,
the three-body reaction. '
H + 02 + M —> H02 + M
has been omitted as it is slow compared to H atom removal by ozone and cis-2-
butene. The rate constants used to evaluate the relative importance of
various reaction paths were the estimated or experimentally determined ones
listed in the references given after each reaction. All the reactions
postulated are exothermic and the observed mass numbers are consistent with
the products expected from kinetic considerations.
The primary reaction step is assumed to involve the initial attack of
ozone on the olefin.
A
0 0
1 ' m
CH3CH=CHCH3 + 03 —> CH.3 - CH - CH - CH3 U)
A
The m/e 104 peak may correspond to any of the structures shown in Table
12.1. However, the hydroperoxide shown in reaction 12.2a is considered to
163
-------
be the
0 OOH
i i
A —> CH3 - C - CH - CH3 (2a)
m/e 104
I
most likely product for the following reasons.
(1) The concentration of the species increases linearly with time as
shown in Figure 12.4, while the primary "molozonide" A is expected to show
steady state kinetics. Application of simple RRK theory to the initially
formed 1,2,3-trioxolane (A) which is assumed to contain the 45 kcal/mol
exothermicity from reaction 1 predicts a lifetime of -lO"11 sec under these
experimental conditions.
(2) There is some infrared absorption evidence for production of the
11 18 19
secondary ozonide in gas phase ozone-olefin reactions. ' ' Recent infra-
red studies in the condensed phase at low temperatures have shown that the
20
secondary ozonides are major reaction products. However, formation of the
secondary ozonide must occur either by rearrangement of the initial ozone-
*
olefin adduct or by recombination of the Criegee "zwitterion" (CH3CHOO') with
21
aldehyde, both of which seem unlikely under our room temperature, low
pressure conditions.
This direct observation of an adduct or rearranged product from an ozone-
olefin reaction with a half-life £.0.2 sec may be related to previous experi-
mental observations. For example, while highly speculative, such a species
could be the phytotoxicant in ozone-olefin reactions which causes plant
22
damage and may be involved in the rapid oxidation of S02 reported by Cox
and Penkett in 03-olefin-S02 systems.
Reaction steps 2b-e involve the decomposition of the intermediate A,
formed in the primary step 1.
A —> CH3CHO + CH2CO + H20 (2b)
—> CH3CHO + CH3OH + CO (2c)
?
—> HCO + C3H7 + 02 (2d)
—> C2H5 + CH3CO + 02 (2e)
164
-------
Although the products other than acetaldehyde in eq 2b and 2c can be
rationalized on the basis of the decomposition of the Criegee "zwitterion"
(or biradical in the gas phase), no direct evidence for its existence was
found .
CH3CO and C3H7 may be produced in reactions 2d and 2e. Both of these
species are expected to react rapidly with 02 (reactions 3 and 4)23 giving
rise to a
C3H7 + 02 — > C3H702 (3)
CH3CO + 02 — > CH3C03 (4)
k - 4 x io9 i mol"1 sec"1
mass peak at m/e 75, consistent with the observations in Table 12.1. Similarly,
C2H5 radicals, formed via reaction 2e would be expected to react with either
02 or 03.
C2H5 + 02(+M) — > C2H502(4M)
23
k ~ 2.5 x i()9 £ mol'1 sec"1 (5)
C2H5 + 03 — > C2H50 + 02 (6)
C2Hs appears to attain a steady state slowly (Figure 12. 3b) from which
estimates of k5 and kg can be obtained using the simplified kinetic expression,
which assumes a constant rate of formation of the intermediates where C
C = B(l - e"Rt)
= concentration of the intermediate, B = constant, and R = the rate loss of
the intermediate which, in general, will be governed by reaction with 02,
03, or cis-2-butene. If C2H5 reacts only with 02, k5 ~ 5 * io5 H mol sec ,
or, if in the third order pressure region, ks ~ 5 x 10 &2 mol sec .
If C2H5 reacts only with 03, then k6 - 2 * IO7 A mol" sec . A possible
explanation for the lack of agreement of k5 with literature values is that
C2H5 is not produced exclusively in reaction 2e but could also arise through
a much more complex reaction sequence.
165
-------
The reactions
HCO + 02
H02 + CO
k ~ 3.4 x 109 £ mol"1 sec"
25
(7)
HCO + 02
[HC02]* + 02
H + C02 + 02
(8)
are expected to be the main removal processes for the HCO radical. While
reaction 8 has not been experimentally verified, it may be one of the H atom
sources responsible for the production of the Meinel band emission of the OH
13
radical. In this work, the very rapid rise of HCO to a steady state con-
centration means that only lower limits of the rate constants for reactions
7 and 8 can be obtained. If HCO reacts only with 02, then k7 > 1 x 107 £ mol
sec , or, if it reacts only with 03, then k8 £ 5 x 1Q8 U mol sec .
From previous investigations, the reactions of H02,
H,
and
35-39
OH expected in the present experimental system are given by reactions
9-16.
H02 + CH3CH=CHCH3
CH3CH(02H)CHCH3
m/e 89
k ~ 3.3 x io6 £ mol sec
-1 -1
26,27
(9) -
CH3CH(02H)CHCH3 + CH3CH=CHCH3
CH3CH(OOH)CH2CH3 + -CH2CH=CHCH3 (lOa)
m/e 90
•CH2CH=CHCH3
CH2=CHCHCH3
0
I
CH2=CHCCH3 + OH
m/e 70
(10b)
HO,
Wall
(11)
H + 03
OH
v<9
k = 1.6 x io10 £ mol"1 sec"1
29-31
(12)
—> OH + hv (Meinel bands)
k = 3.4-16 sec
166
-1
35
(13)
-------
°HV<9 + °3 ~" H°2 + °2
—> 0 + 02 + OH (U)
—> H + 202
i ,28,35
k = 1.2 x 108(V - 0) to 5 x lQ9(v - 9) Si mol sec
OH + CH3CH=CHCH3 —> CH3CH(OH)CHCH3 —> CH3COC2H5
m/e 73 m/e 72
_i37
k = 3.7 x io10 l
H + CH3CH=CHCH3 —>
32-34
k = 5 x IQ8 i mol"1 sec"1 (16)
For H02, reaction with 03 is unimportant (k « 1.8 x 1Q6 A mol"1 sec"1),28
and thus assuming H02 reacts with cis-2-butene. kg can be estimated from the
rise of H02 to its steady state to be approximately 1 x 1Q7 £ mol" sec" ,
96 97
which is consistent with literature estimates. '
37
Morris and Niki have determined by time of flight mass spectrometry
that at least part of the reaction of OH radicals with cis-2-butene proceeds
by addition; transient mass peaks were observed corresponding to the OH-olefin
adduct. Also, by analogy with the products observed by Morris, et al., for
OH + C2H4, C3He, it is expected that methyl ethyl ketone will be the major
stable product in the case of OH + cis-2-butene.
The vibrationally excited C^Hg radical can undergo reactions 17a and 17b
40
in the present system. The
+ M —> Ci+Hg + M (17a)
* —> CH3 + C3H6 (17b)
CH3 produced in reaction 17b can react with either oxygen or ozone.
CH3 + 02 + M —> CH302 + M
-2 -I24
k ~ IO11 £2 mol sec (18)
167
-------
CH3 + 03 --> CH30 + 02
—> HCHO + H02
1 I41
k - 1010 A mol'1 sec (19)
If reactions 16-19 become important at longer reaction times, the anoma-
lous rise in CH302 (We 47) at longer reaction times may be explained; how-
ever, fragmentation in the mass spectrometer to produce CH302 cannot be
excluded at this time.
This latter possibility is supported by the fact that, although reaction
17b is postulated as the source of CH3 in this system, substantial quantities
of C3Hg were not observed.
In similar manner, stabilized C^Hg radicals are expected to react with
02 and 03.
C\H9 + 02 —> CitHgC^ (20)
m/e 89
CijHg + 03 > CijHgO + 02 (21)
m/e 73
The results of this study do not permit detailed formulation of a reaction
scheme for the ethylene and isobutene oxidations. The observed major stable
products are expected to arise from the decomposition of the initial ozone-
olefin adduct.
While the mechanism presented for the cis-2-butene-ozone reaction is
obviously tentative and incomplete, the observation of radical species, such
as HCO and H02, provides evidence for alternative reaction pathways to the
Criegee mechanism which has previously been applied.
168
-------
REFERENCES
1. I. T. N. Jones and K. D. Bayes, Symp. (Int.) Combust., [Proc], 14th
(1972).
2. I. T. N. Jones and K. D. Bayes, J. Amer. Chem. Soc., JJ4, 6869 (1972).
3. J. R. Kanofsky and D. Gutman, Chem. Phys. Lett., 15, 236 (1972).
4. H. Okabe, J. Opt. Soc. Amer., 54, 478 (1964). .'
5. R. W. Riser, "Introduction to Mass Spectrometry and Its Applications,"
Prentice-Hall, Englewood Cliffs, N. J., 1965.
6. T. Vrbaski and R. J. Cvetanovic, Can. J. Chem., 38, 1053, 1063 (1960).
7. Y. K. Wei and R. J. Cvetanovic, Can. J. Chem., 41, 913 (1963).
8. J. J. Bufalini and A. P. Altshuller, Can. J. Chem., 43_, 2243 (1965).
9. B. J. Finlayson, J. N. Pitts, Jr. and R. Atkinson, J. Amer. Chem. Soc.,
j?6, 5356 (1974).
10. D. H. Stedman, C. H. Wu, and H. Niki, J. Phys. Chem., 77, 2511 (1973).
11. R. A. Cox and S. A. Penkett, J. Chem. Soc., Faraday Trans. 1, 68, 1735
(1972).
12. W. E. Scott, E. R. Stephens, P. H. Hanst, and R. C. Doerr, Proc. Amer.
Petrol. Inst., _3 (37), 171 (1957).
13. B. J. Finlayson, J. N. Pitts, Jr., and H. Akimoto, Chem. Phys. Lett.,
12, 495 (1972).
14. M. J. Kurylo, J. Phys. Chem., 76, 3518 (1972), and references therein.
15. D. L. Baulch, D. D. Drysdale, D. G. Home, and A. C. Lloyd, "Evaluated
Kinetic Data for High Temperature Reactions. Vol. 1: Homogeneous Gas
Phase Reactions of the H2-02 System," Butterworths, London, 1972.
16. L. S. Kassel, "The Kinetics of Homogeneous Gas Reactions," Chemical
Catalogue, New York, N. Y., 1932.
17. S. W. Benson, "Thermochemical Kinetics," Wiley, New York, N. Y., 1968,
p. 173.
18. W. B. DeMore, Int. J. Chem. Kinet., .1, 209 (1969).
19. P. L. Hanst, E. R. Stephens, W. E. Scott, and R. C. Doerr, Symposium on
Air Pollution Research, Division of Petroleum Chemistry, Vol. 4, Sept.
1959.
169
-------
REFERENCES (cont.)
20. L. A. Hull, I. C. Hisatsune, and J. Heicklen, J. Amer. Chem. Soc., 94,
4856 (1972).
21. H. Niki, E. E. Daby, and B. Weinstock, Advan. Chem. Ser., No. 113, 16
(1972).
22. W. N. Arnold, Int. J. Air Pollut., 2^ 167 (1959), and references therein.
23. G. R. McMillan and J. G. Calvert, Oxid. Combust. Rev., _!, 83 (1965).
24. N. fiasco, D. G. L. James, and F. C. James, Int. J. Chem, Kinet., 4^,
129 (1972).
25. N. Washida, R. I. Martinez, and K. D. Bayes, Z. Naturforsch., 29A, 251
(1974).
26. K. L. Demerjian, J. A. Kerr, and J. G. Calvert, Advan. Environ. Sci.
Technol., ^ (1973).
27. A. C. Lloyd, "Estimated and Evaluated Kinetic Data for Gas Phase Reactions
of the Hydroperoxyl Radical," National Bureau of Standards Report 10447,
July 1971; Int. J. Chem. Kinet., j6, 169 (1974).
28. W. B. DeMore, Science, 180, 735 (1973).
29. J. D. McKinley, Jr., D. Garvin, and M. J. Boudart, J. Chem. Phys., 23,
784 (1955).
30. L. F. Phillips and H. I. Schiff, J. Chem. Phys., 3_7_, 1233 (1962).
31. P. E. Charters, R. G. MacDonald, and J. C. Polanyi, Appl. Opt., 10,
1747 (1971).
32. E. E. Daby and H. Niki, J. Chem. Phys., 51, 1255 (1969).
33. J. A. Cowfer, D. G. Keil, J. V. Michael, and C. Yeh, J. Phys. Chem., 75,
1584 (1971).
34. E. E. Daby, H. Niki, and B. Weinstock, J. Phys. Chem., _7_5, 1601 (1971).
35. R. N. Coltharp, S. D. Worley, and A. E. Potter, Appl. Opt., 10, 1786
(1971).
36. E. D. Morris, Jr., D. H. Stedman, and H. Niki, J. Amer. Chem. Soc., 93,
3570 (1971).
37. E. D. Morris, Jr. and H. Niki, J. Phys. Chem., 7_5, 3640 (1971).
'
-------
REFERENCES (cont.)
40. B. S. Rabinovich and D. W. Setser, Advan. Photochem., ^3, 1 (1964)
41. J. G. Calvert and P. L. Hanst, Can. J. Chem., J7., 1671 (1959).
171
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-77-014a
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
MECHANISMS OF PHOTOCHEMICAL REACTIONS IN URBAN AIR
Volume I. Chemistry Studies
5. REPORT DATE
February 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
James N. Pitts, Jr.
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
University of California
Statewide Air Pollution Research Center
Riverside, California 92502
10. PROGRAM ELEMENT NO.
1AA605
11. CONTRACT/GRANT NO.
Grant No800649-15.14,15
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD;CO,VERED
Final Report 12/1/71-12/1/74
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT Resuits are presented of a research program concerned with selected aspects
of the kinetics, mechanisms and products of reactions involved in photochemical air
pollution.
Rate constants were determined, using competitive and modulation-phase shift
techniques, for fche gas phase reaction of 0( P) atoms wjth a variety of organic
and inorganic species over the temperature range 296-423 K. Products for the gas
phase reaction of 0( P) atoms with toluene and 1-methylcyclohexene were also studied.
The products and mechanisms of the reaction of nitric oxide with methyl peroxy
radicals were investigated at 296°K using long path infrared spectroscopic and gas
chromatographic techniques.
The reactions of peroxyacetyl nitrate were investigated in the gas phase with
selected constituents of polluted atmospheres, and in the liquid phase with a variety
of organics. Chemiluminescence from the reaction of peroxyacetyl nitrate with a
series of amines was studied in the liquid phase. The mechanism and products of the
gas phase reactions of ozone with a variety of organics was investigated in low
pressure flow systems using chemiluminescent and photoionization mass spectrometric
techniques.
The NO^-catalyzed geometric isomerization of 2-butenes and 2-pentenes was studied
over the temperature range 298~400°K while an investigation of the NO -propylene
^^ ^_ * _* _^ ,^_ _ __ _ ^*
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air pollution
Photochemical reactions
Reaction kinetics
Hydrocarbons
Ozone
Nitrogen oxides
13B
07E
07D
07C
07B
8. DISTRIBUTION STATEMEN1
RELEASE TO PUBLIC
19. SECURITY CLASS (TMsReport)
21. NO. OF PAGES
184
22. PRICE
EPA Form 2220-1 (9-73)
172
-------
|