United States
Environmental Protection
Agency
Office of Toxic Substances
Washington, DC 20460
Toxic Substances
Atmospheric
Products
of Organic
Compounds
June 1979
EPA-560/12-79-001
Final
Report
-------
EPA-560/12-79-001
June 1979
ATMOSPHERIC REACTION PRODUCTS OF
ORGANIC COMPOUNDS
by
D. 6. Hendry and R. A. Kenley
%
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
Contract No. 68-01-5123
Project Officer
Stuart Z. Cohen
Office of Chemical Control
Washington, D. C. 20460
OFFICE OF CHEMICAL CONTROL
OFFICE OF TOXIC SUBSTANCES
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
-------
DISCLAIMER
This report has been reviewed by the Office of Toxic Substances, U.S.
Environmental Protection Agency, and approved for publication. 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
A general procedure has been developed to predict the products resulting
from reaction of various compounds in the atmosphere. The procedure is
designed to be used in unreasonable-risk evaluations that include assessing
new chemicals for persistence and exposure in the environment.
In this procedure, the relative importance of the three dominant reaction
pathways—photolysis, reaction with OH radical, and reaction with ozone—are
first determined for each compound. Then the products from each major pathway
are estimated using the techniques that have been developed.
The methods are applicable to a wide variety of compounds; however in
cases where the structure of the compounds differ in type from the structures
on which the procedures were based, the conclusions may be less certain.
Although it is not possible to anticipate the new structures to which the
procedure will be applied, only a minor fraction of the cases are expected
to fall into this latter category.
iii
-------
CONTENTS
Page
ABSTRACT ill
FIGURES vii
TABLES viii
1. INTRODUCTION 1
2. CONCLUSIONS AND RECOMMENDATIONS 3
3. BACKGROUND 4
General Considerations 4
Importance of OH and 03 Reactions 5
Basis for Prediction of Reaction Products 13
4. ESTIMATING RATES OF ATMOSPHERIC REACTIONS 14
Hydroxyl Radical Reactions 14
Ozone Reactions 21
Photochemical Transformations 22
5. PRODUCTS FROM REACTIONS OF OH 30
Products Resulting from Hydrogen Abstraction 31
Atmospheric Reaction of Peroxy Radicals 31
Chemistry of Alkoxy Radicals 32
Bimolecular Reactions 32
Unimolecular Reactions 33
Products Resulting from Addition of OH to C=C Double
Bonds 40
Products from Addition to Aromatic Compounds 42
6. PRODUCTS FROM REACTIONS OF OZONE 49
7. PHOTOCHEMICAL REACTION PRODUCTS 53
General 53
Limitations 53
Photochemical Reactions 57
Carbonyl Compounds 57
Olefins 62
Nitrites, Nitrates, Nitro and Nitroso Compounds . . 63
Halides 65
Procedures 65
8. PREDICTIVE SCHEME 69
Step 1. Estimation of Rates of Environmental
Processes 69
Estimation of the Rate Constant of Reaction with OH
(koH) and TOH 69
Estimation of the Rate Constant for Reaction with
Ozone (kQ3) and TQ9 ^ . 70
Estimation of the Rate Constant for Photolysis (k )
and Tp p. 71
Step 2. Estimation of Products from OH Reactions ... 71
-------
C-H Bond Reactions 71
C=C Bond Reactions 72
Aromatic Ring Reactions 72
Relative Concentration of Products 72
Step 3. Estimation of Products from 03 Reactions ... 73
Step 4. Estimation of Products from Photolysis .... 73
INDEX 74
REFERENCES 76
vi
-------
FIGURES
Number Page
1 Correlation of Rate Constants for Addition of OH to
Substituted Benzenes with Substituent Constants £0 20
P
2 Prediction of Products for 1,1-Dichloroethane 39
3 Prediction of Products for 3-Chloro-4-Hydroxy-cis-butenoic Acid
Methyl Ester 43
4 Prediction of Products for 3-Chloro-Ethylbenzene 49
vii
-------
TABLES
Number Page
1 Apparent First-Order Rate Constants for Reactions of
Atmospheric Oxidants with Propene, n-Butane, and Toluene .... 6
2 Rate Constants and Environmental Half-Lives for Reactions of
Hydroxyl Radicals near 300°K 7
3 Rate Constants and Environmental Half-Lives for Reactions of
Ozone at 300°K 10
4 Abstraction Rate Constants (kg) for Reactions of OH with
Generalized Groups and Values of Induction Factors (OL and B ) . 17
5 Addition Rate Constants (kg) for Reaction of OH with
Carbon-Carbon Bond and Values of Induction Factors (a.) .... 18
6 Addition Rate Constants (kA) for Reaction of OH with Aromatic
Rings and Values of Induction Factors (or.) 19
7 Rate Constants for Reaction of Ozone with Generalized
Structures 23
8 JA' Values at 10°H Latitude 25
9 Jx' Values at 30°N Latitude 26
10 J. Values at 50°N Latitude 27
/.
11 Examples of Hand Computation of Photochemical Rate Constants . . 28
12 Estimated Rate Constants for Reactions of Primary, Secondary,
and Tertiary Alkozy Radicals 35
13 Estimated Rate Constants for Reaction of Methoxy Substituted
Alkoxy Radicals 37
14 Selected Values of Hammett a 46
P
15 Products from Reaction of Ozone and Alkenes under Atmospheric
Conditions 51
viii
-------
Number Page
16 Approximate Absorption Regions for Organic Molecules 54
17 Bond-Dissociation Energies for Some Organic Molecules R'-R" . . 56
18 Products of Photochemical Reactions 67
ix
-------
SECTION 1
INTRODUCTION
Both chemical and physical processes compete in determining the fate of
chemicals in the atmosphere. The more reactive compounds react rapidly within
the boundary layer immediately adjacent to the surface of the earth where they
enter the troposphere. Less reactive compounds will be distributed by diffusion
and turbulent mixing into the remainder of the troposphere where reaction will
occur over a longer period of time. Compounds that are stable in the troposphere
and not removed by physical processes will gradually diffuse into the
stratosphere where additional loss mechanisms exist—the more important being
reaction with 0(XD) and photolysis in the 150-300 run region of the solar
spectrum above the ozone layer.
To understand the effects of industrial chemicals on the environment, it
is important to be able to anticipate not only what the rates of the possible
processes are for each chemical but also what the products will be from these
processes. This is especially important because the effect of introducing a
chemical into the troposphere on the environment is not limited to the chemical,
physical, and biological interactions of that chemical but also involves the
effect of products derived from various reactions of that chemical.
Under EPA Contract No. 68-03-2227, SRI International has prepared proce-
dures to assess the rates of dominate atmospheric processes for chemicals
(Hendry et al., 1979). These procedures involve two levels: first, a screening
level where rates are approximated based on the generalized chemical structures
in the literature data, and second, an experimental level where rate constants
for processes that are potentially critical are determined. Using the proce-
dures in the first level, we can readily approximate the rate of individual
processes for many chemicals without further experimental study and thereby
predict the lifetimes of the chemicals in the environment.
The objective of this study is to review concepts and data on product
formation from tropospheric reactions and to develop procedures for predicting
transformation products of new chemical substances in the atmosphere. The
methods should be suitable for use in unreasonable-risk evaluations that in-
clude assessing new chemicals for persistence and exposure in the environment.
The general approach is to identify structure-reactivity relationships
that correlate existing rates of reaction and that will be useful in predicting
rates of reaction of new structures. The success of this approach depends not
only on the reliability of the correlations between structure and reactivity
but also on the application of new structures to that correlation. To the
extent that a new compound introduces a new structural variation, which will
-------
often be the case for a new industrial chemical, it is difficult to know what
reliability a correlation will have, if it can be applied at all, for such a
compound. Therefore, these correlations cannot be totally successful in pre-
dicting the reaction rates and products of new structures. However, they will
be very helpful in many cases, especially when the structure is consistent with
those structures where a correlation has been demonstrated.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
A predictive procedure has been set forth to anticipate the products
resulting from the reaction of compounds in the atmosphere. The methods should
find application to a wide variety of compounds and be suitable as part of the
persistence and exposure assessment of new chemicals. The ultimate goal has
been to use structure reactivity relationships to assist in the predictions;
however, the cases where such relationships have been applied to gas-
phase reactions of importance to atmospheric reactions are very limited, and
in general, the study of the effect of structure, as has been investigated in
the solution phase, has not been explored.
An investigation of methods of correlating atmospheric reactions of OH
and 03 should yield useful results. The success in correlating the rate
constants for addition of OH to aromatic rings with 0-type substituent constants
and the rate constants of cleavage of alkoxy radicals with the heats of reaction
indicate these methods should be extended. There are two levels of this general
problem that deserve further consideration. First, using the current data base
extend our initial effort to investigate types of structural parameters that
best correlate the data. For example, do solution phase parameters of the
Hammett type correlate OH and 03 reactions adequately, or do parameters like
electron affinity or ionization potential apply more satisfactorily? The
second level would be to enlarge the experimental data base on which to test
various methods of correlation. The types of structures that we have OH and
03 rate constant data for are limited and determination of rate constants for
a wider range of structures would be very helpful.
Finally, our knowledge of atmospheric gas-phase photo reactions is very
limited. One major problem centers around our anticipating whether or not
oxygen quenches the photo-excited states of the compounds of interest and there-
by prevent product formation. In many cases, this factor can mean the
difference between a rapid reaction due to photolysis or no reaction at all.
A detailed review of the literature of oxygen quenching would be helpful in
determining what generalizations regarding quenching are possible.
-------
SECTION 3
BACKGROUND
GENERAL CONSIDERATIONS
The detailed atmospheric chemistry by which compounds containing carbon-
and hydrogen are largely converted to C02 and H20 and by which organic nitrogen
and sulfur are converted to nitrates and sulfates is quite complex (Leighton,
1961; Demerjian et al., 1974; Pitts and Finlayson, 1975; Baldwin et al., 1977;
Hendry et al., 1978; Carter et al., 1978). We have spent considerable effort
studying these processes, and our understanding is reasonably good, although
semi-quentitative. Currently we are developing chemical mechanisms that
describe the chemistry of individual hydrocarbons to simulate the chemistry
occurring in smog chambers. These simulations are reasonably effective in
predicting the smog chamber results.
Our understanding of atmospheric chemistry is sufficient to allow
identification of the critical chemical processes that will affect the lifetime
of chemicals in the atmosphere. These processes involve oxidation reactions
and photochemical transformations. These two basic processes are defined as
follows.
Oxidation reactions involve interaction of a compound in its electronic
ground state with a reactive species such as oxygen, ozone, 0 atom, or OH
radical. In this case, the rate determining step is the bimolecular reaction
of compound and reactive species
Rate of Oxidation = k [OX][C]
UA.
where k0X is the rate constant for the process, [OX] is the concentration of
reactive species, and [C] is the concentration of chemical.
Photochemical transformations are those reactions wherein the compound
absorbs solar energy directly to form an electronically excited intermediate
that undergoes further reactions by either first-order or second-order processes.
In all cases the rate determining step is the absorption of solar energy and
follows the relation
Rate of Phototransformation » I /h
a
where Ia is the light absorbed (in photons cm"2 s"1), h is the depth (in cm)
over which the light is absorbed, and , quantum yield, is the fraction of
-------
excited species that reacts.
IMPORTANCE OF OH AND 03 REACTIONS
Except for some compounds that photolyze rapidly, the most important
reactions of organic compounds in the atmosphere are with OH radical and ozone.
Table 1 summarizes approximate daytime concentrations of these species along
with several other oxidants also believed to be present in the atmosphere.
Included in the table are rate constants for reaction of these oxidants with
propene, n-butane, and toluene, which are common atmospheric pollutants. The
product of the rate constant and the concentration of oxidant is the apparent
first-order rate constant (k^) for that process.
Rate = kQX[OX][C] = k^[C]
Clearly, in the case for propene the reaction with OH is most important, while
the reaction with 03 is second most important. Together, OH and 03 account
for more than 99% of the reaction; ground state oxygen atom (0(3P) accounts
for only 0.1% of the reaction. For both n-butane and toluene, OH alone accounts
for greater than 99% of the reaction. These conclusions obviously are dependent
on the concentrations that are assigned for each oxidant. However, the con-
centrations of OH and 03 are not expected to vary significantly from the values
in the table.
In some cases a compound might have unusually high reactivity to one of
the other oxidants, although this should be the exception. One important case
is the oxidation of NO by HOO- (or R00») . The reaction
HOO- + NO — »~ H0» + N02
is the major route by which NO is oxidized in the atmosphere. The reaction
competes favorably with the reaction with ozone. However, to our knowledge,
there are no cases involving organic compounds where reactions with oxidants
other than OH and 03 are important.
Thus, to estimate the lifetime of chemicals in the atmosphere as deter-
mined by oxidation processes, we need to determine only the rate constants for
reactions with OH and 03. The rate constants for reaction of numerous compounds
with OH and 03 are summarized in Tables 2 and 3.
The lifetime (T) for reaction of substrate (S) with an oxidant is defined
T = = 1/ko*[oxl
where kox is the bimolecular rate constant and [OX] is the approximate concen-
tration of the reactive species. The time to deplete one-half of the chemical
is referred to as the half-life (t,) and is defined
-------
TABLE 1. APPARENT FIRST-ORDER RATE CONSTANTS FOR REACTIONS OF ATMOSPHERIC OXIDANTS WITH PROPENE, n-BUTANE, AND TOLUENE
Oxidant
OH
0,
0(SP)
0('D)
H0a*
Oa('Aa)
NO,
Concentration
(molec cc"1)
2 x 106
1 x 1012
3 x 10*
2 x 1CT1
1 x 1010
1 x 10'
2 x 10"
Propene
k-x, cc molec"1 s"1
2.5 x 10"11
1.0 x 10"17
3.5 x 10~12
•v. 1 x 10" ' "
< 2 x 10"18
<; 1.7 x 1CT20
5.3 x 10~>s
kox[OX], s"1
5 x 10"s
1 x 10" *
1 x 10" 7
2 x 10" ' 2
< 2 x 1CT8
5 2 x 10T11
1 x 10" 6
n-Butane
k__, cc molec"1 s-1
r
2.7 x lO"12
% 0
5.2 x 1(T9
'v- 1 x 10" 1 1
^ 0
•v 0
-v 0
kox[OX], s"1
5 x lO"6
•v 0
2 x 10-'
2 x 10- 12
* 0
•x. 0
-v 0
Toluene
kQ , cc molec"1 s"1
6.4 x 10"12
•v 0
7.5 x 10" 16
* 1 x 10" * l
•v. 1 x 10"18
•v 0
^ 0
kox(OX), s-1
1 x 10"3
•v 0
2 x 10"'
2 x 10" * 3
1 x 10" 8
•v 0
•v 0
Kate constants from references shown below; concentrations are estimated based on smog chamber stimulation data from Hendry (1978);
rate constants were taken from Hampson and Garvin (1978), Hendry and Mabey (1978), Demerjian, Kerr, and Calvert (1974); Japar and Niki
(1975).
To convert to ppm multiply by 4.1 x 10~1*; to convert to M, multiply by 1.7 x 10"ai.
-------
TABLE 2. RATE CONSTANTS AND ENVIRONMENTAL HALF-LIVES FOR REACTIONS OF HYDROXYL RADICALS NEAR 300*K*
Compound
Alkanes
Methane
Ethane
Propane
Methyl
Dimethyl
n-Butane
Methyl
2,3-Dimethyl
2,2,3-Trlmethyl
2,2,3,3-Tetramethyl
n-Pentane
2-Methyl
3-Methyl
2.2,4-Trlmethyl
n-Hexane
n-Octane
Cycloalkanea
c-Butane
c-Pentane
c-Hexane
Haloalkanea
Methane
Fluoro-
Difluoro-
Trlfluoro-
Tetrafluoro-
Chloro-
Dichloro-
Trlchloro-
Tetrachloro-
Bromo
Ethane
Chloro
1,1-Dtchloro
1,2-Dichloro
1,1,1-Trichloro
1,1, 1-Trif luoro-2-chloro
l,l,l-Trlfluoro-2,2-dichloro
1,1,1,2-Tetraf luoro-2-chloro
1,2-Dibromo
Alkanone
Butanone
2-Methylpentanone
2 , 6-Dlmehtylhep tanone
Alkanola
Methanol
Ethanol
Propanol
2-Propanol
Butanol
4-Methyl-2-pentanol
O.N.S Substituted alkanea
Methyl ether
Ethyl ether
n-Propyl ether
101> "OH'
cm1 molec"1 s~*
0.0079
0.29
2.2
2.2
0.81
2.7
3.3
5.1
3.8
1.1
6.5
5.3
7.1
3.8
6.0
8.5
1.2
6.1
7.0
0.016
0.0078
0.0002
< 0.0004
0.04
0.14
0.11
< 0.0004
0.04
0.39
0.26
0.22
0.015
0.010
0.028
0.012
0.25
3.3
14.9
24.9
0.95
3.0
3.8
7.1
6.8
7.1
3.5
9.3
17.3
t^, days'
1,000
28
3.6
3.6
9.9
3.0
2.4
1.6
2.1
7.3
1.2
1.5
1.1
2.1
1.3
0.9
6.7
1.3
1.1
500
1,030
40,000
> 20,000
200
57
73
> 20,000
200
20
31
36
530
800
280
670
32
2.4
0.54
0.32
8.4
2.7
2.1
1.1
1.2
1.1
2.3
0.86
0.46
-------
TABLE 2 (Continued)
Compound
O.N.S Substituted alkanes
Tetrahydrofuran
1-Propylacetate
2-Butylacetate
Methylaoine
Methyl sulflde
Formaldehyde
Acetaldehyde
Proplonaldehyde
Benzaldehyde
Alkenes
Ethene
Propene
Me thy 1-
1-Butene
2 -Methyl
3,3-Dlmethyl
2-Butene
cis-
trans-
2-Hethyl
2,3-Dimethyl
1-Pentene
cis-2-Pentene
1-Hexene
1-Heptene
Cycloalkenes
c-Cyclohexene
1-Methyl
Haloalkenes
Ethene
Fluoro
1,1-Difluoto
Chloro
Trichloro
Tetrachloro
Chlorocrlfluoro
Brooo
0-Substituted alkene
Hethoxy
Alkadienes
Propadiene
1,3-Butadlene
2-Methyl
Terpenes
p-Menthane
o-Plnene
0-Plnene
3-Carene
Carvomenthane
B-Phellandrone
d-Lloonene
Dlhydromyrcene
Myrcene
cls-Oc inane
10" kOH. can1 -olec- ' f '
14.6
4.5
5.6
21.9
33.9
15
16
21
13
7.9
24.8
50.6
35.4
58.1
28.2
53.6
69.9
79.7
153
29.9
64.8
31.5
36.5
71.4
96.3
5.6
2.0
6.6
2.0
0.17
7.0
6.8
33.5
4.5
77
78
6.6
25.7
21.6
28.2
41.5
38.2
48.1
59.7
74.7
105
tjj, days
0.55
1.8
1.4
0.37
0.24
0.53
0.50
0.38
0.61
1.5 0.85
0.5
0.2
0.3
0.2
0.4
I
0.2
0.2
0.1 j
0.1
0.4
0.2
0.4
0.3
i
i
0.2
0.1
1.4
4.0
1.2
4.0
47
1.1
1.2
0.24
2.6
0.1
0.1
1.2
0.31
0.37
0.28 '
0.19
0.21
0.17
0.13
0.11
0.076
\
8
-------
TABLE 2 (concluded)
Compound
Alkynes
Ethyne
Methyl
Aralkanes
Benzene
Methyl
1,2-Dlmethyl
1,3-Trlmethyl
1,4-Trimethyl
1,2,3-Triaethyl
1,2,4-Trimethyl
1,3,5-Trijnethyl
Ethyl
10 ia kOH, cm8 molec"1 s" »
0.16
0.95
1.4
5.9
13
20
10
24.7
33.2
49.3
ty daysb
50
8.4
5.7
1.3
0.62
0.40
0.80
0.32
0.24 '
0.16
7.5 : 1.1
1,2-Ethylmethyl 13.6 0.59 \
1,3-Ethylmethyl 19. A ! 0.41
1,4-Ethylmethyl i 12.9 0.62
propyl : 6.0 1.3
2-propyl 7.8 ! 1.0
1 , 4-raethy lpropyl-2-
hexafluoro
propylpentafluoro
Substituted Aralkanes
Methoxybenzene
o-Cresol
15.2
0.22
3.0
19.6
34.0
0.53
36
2.7
0.41
0.24
aCompiled from data and references in Pitts et al., 1977, and Hampson
and Garvin, 1978.
Environmental half-lives in 24-hour days, assuming an OH concentration
equal to 1 x 10* radicals cm"'.
9
-------
TABLE 3. RATE CONSTANTS AND ENVIRONMENTAL HALF-LIVES FOR REACTIONS OF OZONE AT
300 °K
Compound
Alkanes
Methane
Ethane
Propane
Methyl
n-Butane
Alkenes
Ethene
Propene
Methyl-
1-Butene
2-Butene
cis-
trans-
2-Methyl
2 , 3-Dimethyl
1-Pentene
2-Pentene
cis-
trans-
1-Hexene
1-Heptene
1-Octene
1-Decene
Cyclohexene
Conjugated Alkenes
1,3-Butadiene
Phenylethene
1018 k. , cm3 molec-1 s"1
Oa
1.4(-6)
1.2 (-6)
6. 8 (-6)
2.0 (-6)
9. 8 (-6)
1.9
13
15.1
12.3
161
260
493
1510
10.7
456
563
11.1
8.14
8.14
1.08
169
8.4
171.0
Reference
a
b
a
a
a
h
h
h
h
h
h
h
*
h
h
h
h
e
e
e
h
h
i
ty days
5.73 (6)
6.68 (5)
1.18 (6)
3.95 (5)
8.18 (6)
4.22
0.617
1.30
0.652
0.05
0.031
0.016
0.005
0.75
0.018
0.014
0.723
0.99
0.99
7.43
0.047
0.95
0.05
10
-------
TABLE 3 (continued)
Compound
Halogenated Alkenes
Ethene
Chloro
1,1-Dichloro
1
1,2-Dichloro
• cis
trans
Trichloro
• Tetrachloro
; Tetrafluoro
Propene
: 3-Chloro
Hexafluoro
Terpenes
o-Pinene
Alkynes
Ethyne
i
Aromatic Hydrocarbons
Benzene
Methyl
1,2-Dimethyl
1,3-Dimethyl
1,4-Dimethyl
1,3,4-Trimethyl
; 1,3,5-Trimethyl
i
; 1,2,4,5-Tetramethyl
I Pentamethyl
i Hexamethyl
10 18 kQ , cm3 molec-1 s"1
1.96
3.67 (-2)
6.14 (-2)
3.82 (-1)
5.98 (-3)
1.66 (-3)
134
18.3
21.6
164
7.8 (-2)
4.65 (-5)
2.76 (-4)
1.58 (-3)
1.30 (-3)
1.58 (-3)
4.65 (-3)
6.97 (-3)
1.78 (-2)
8.30 (-2)
0.407
1 !
Ethyl j 5.65 (-4)
( i
Reference
i
i
g
g
i
i
c
i
c
j
f
f
f
f
f
f
f
f
f
f
f
t^, days
i
4.09
2.19 (2)
1.31 (2)
2.10 (1)
1.34 (3)!
4.83 (3)
.06
.438 ;
.37
.05
102.8
i
1.72 (5)|
2.91 (4)
5.09 (3)|
6.19 (3)|
5.09 (3)|
1.76 (3)|
1.15 (3)
4.51 (2)
96.6
19.71
1.42 (4)
11
-------
TABLE 3 (concluded)
Compound
Aromatic Hydrocarbons
Benzene (cont . )
1,3-Diethyl
1018 kn , cm3 mole
-------
= ln2/kox[OX] = Tln2
BASES FOR PREDICTION OF REACTION PRODUCTS
Because photolysis and reactions with OH and ozone are the dominant re-
actions of organic compounds in the atmosphere, we must be able to predict the
products of these reaction. If, from estimates of the rate constants, we can
determine that one process dominates the other two, then we know the products
from that dominant process will be the atmospheric products. If more than
one process is dominant, then the products will reflect the contribution of
those reactions. Thus, the reaction of compound C is expressed
d[C]/dt = kOR[OH][C] + k03[03][C] + kp[C]
= {kOR[OH] +k03[03] + kp}[C]
The fraction of C reacting during each process and thus the contribution of
each process to the products is
kroHl[OH]
%Rx by OH reaction -
%Rx by 03 reaction =
[0,]
%Rx by photolysis - k^OH] + [0,] + kp
Therefore, to determine relative importance of each process we must know
the constants [OH] and [03]. We previously have shown that the annually
averaged ground level concentration of OH is best estimated by 106
molecules/cm3, and the average ozone concentration is best estimated by 10 12
molecules/cm3 (Hendry, 1979). Thus, the ability to estimate the constants
for the three basic processes is the key to predicting what products are
important.
13
-------
SECTION 4
ESTIMATING RATES OF ATMOSPHERIC REACTION
HYDROXYL RADICAL REACTIONS
From the data in Table 2 it is possible to make several generalizations
about OH reactivity that can be useful in estimating the reactivity of compounds.
Most alkanes, except methane and ethane, react toward OH with a rate constant
of 1 to 10 x 10" 12 cm3 moled"1 s~l. The reaction with alkanes entails abstrac-
tion of hydrogen to form water and an alkyl radical
Substituents can have many different effects. Single halogen atoms in the
a-position activate the carbon-hydrogen bond, but in the 3-position they tend
to deactivate the bond. Thus, single halogen atoms do not have a large effect
on the reactivity of any alkane except methane. Multiple halogens in a
molecule generally deactivate the molecule relative to the parent alkane.
Thus, 1,1,1-trichloroethane is about 1/20 as reactive as ethane; one-half of
the reduction is due to the reduction of the number of available hydrogens,
and then the remaining hydrogens are reduced in reactivity by 1/10.
In ketones, alcohols, ethers, and esters, activation occurs relative to
the parent hydrocarbon after correction for the number of hydrogens. Amines
and sulfide appear to show even larger rate enhancement due to activation of
adjacent C-H groups.
The carbon-carbon double bond is much more reactive than the carbon-
hydrogen bond and reacts by addition of OH, which generally causes cleavage
of the double bond (Niki et al., 1978).
Rate constants for these processes are sensitive to the degree of alkyl sub-
stitution on the double bond and vary from 20-200 x 10~12. Consequently, al-
kenes have a half-life in the environment one-tenth of that for most alkanes.
Halogens tend to reduce alkene reactivity somewhat, whereas methoxy has the
same effect as an alkyl group. Conjugated dienes have reactivities equal to
two double bonds. The reactivities of terpenes are consistent with the
similarly substituted alkenes. Alkynes appear to be significantly less re-
active than the. corresponding alkenes.
14
-------
Aromatic hydrocarbons have reactivities about the same as alkenes as a
result of addition of OH to the ring (Kenley et al., 1978). For examples,
85% of the time, the reaction of toluene with OH adds to the ring giving mainly
CH3
(+ isomers) + H02-
Hydrogen abstraction (shown below) occurs only 15% of the time.
OH +
+ HaO
It is desirable to be able to estimate rate constants for reaction
of various kinds of organic structures with OH. A reasonably reliable pre-
dictive scheme can be developed on the basis that the rate constant or total
molecular reactivity of a molecule is the sum of rate constants for reactivi-
ties of each portion of the molecule, and that little or no effect is exerted
on the reactivity of a specific site by a substituent more than two atoms
away. Thus, the same group in different molecules is assumed to have the same
reactivity. For example, the methyl groups in all alkanes are assumed to have
the same reactivity; however, methyl groups in compounds like toluene, acetone,
and methyl chloride are assumed to have potentially different reactivities
because the second atom from the hydrogen which is abstracted is different.
The chemistry of OH reactions is largely limited to the three kinds of
reactions discussed above: abstraction of a hydrogen atom, addition to a
carbon-carbon double bond, and addition to an aromtic ring. The rate constant
for OH attack is the sum of the rate constant for each of these processes:
OH
k(H-abstraction) + k(C=C addition) + k(ring addition)
The rate of hydrogen atom abstraction is affected by substitution on the
same and adjacent atoms. The total rate constant for abstraction (kg) may be
expressed as the summation of the rate constants for each reactive hydrogen
atom as suggested by Greiner (1970)
where kg^ is the reactivity of the ith hydrogen atom and depends on the degree
of substitution on the adjacent atom and on whether a vinyl or phenyl group
is attached. The terms OH and (Jjj account for the effect of substituents other
15
-------
than hydrogen; aH is the constant for the substituent in the a-position and
0H is the substituent in the 0-position. The term m is the number of times
the same type of hydrogen group with the same a and 0 substituents appears in
the molecule. Greiner (1970) and Darnall et al. (1978) have applied this
type of expression to simple alkane reactivity where a and 0 = 1. Values of
kH, otjj , and PH are given in Table 4.
The rate constant for addition to a carbon-carbon double bond (k£) is
expressed by
j
where kg. is the reactivity of the jth carbon-carbon double bond and depends
on the degree of substitution by carbon, oxygen, nitrogen, or sulfur but not
halogen. The erg. term is unity unless a halogen is Immediately attached to
the double bond. Values of k and o_, are given in Table 5.
Similarly, the rate constant for addition to aromatic rings (k.) is ex-
pressed by
£
kA = Zo. 0kAO
A Oal™' "*
where kAA is the reactivity of the ith aromatic ring toward OH addition, which
depends on the degree of alkyl substitution. The a^ is a factor to account
for the effect of halogen atoms substituted in the ring. Table 6 summarizes
values of k^ and a^.
In Figure 1 we have plotted the report value of the rate constants for
addition to aromatic rings versus the summation of ap values for the substituent
on the benzene ring. Monosubstituted and polysubstituted ring data appear to
fall on two separate lines. These expressions can be used to predict k^ for a
much wider variety of structures, although it has not been tested for strong
electrons with drawing substituents.
Thus the total molecular rate constant (km) can be expressed by the sum
of these three processes:
i j I
kOH - ^i-BAAi + ^jEj^j + ,f°A*kA*
The reactivities of four representative compounds are tabulated on page 21.
Upper and lower limits are given in parentheses following the best estimates.
In these examples the uncertainty in the estimated value of kQH is about a
factor of 2, the range being one-half the estimate to twice the estimate. Rate
constants have been measured for the first two examples and the data are given.
16
-------
TABLE 4. ABSTRACTION RATE CONSTANTS (kH) FOR REACTION OF OH WITH GENERALIZED
GROUPS AND VALUES OF INDUCTION FACTORS (O AND
«
Group
Cg CX2—H
(Cg)2CX-H
(Cg)3C-H
CD-H
CDCXa-H
(Cg)(CD)CX-H
(Cs)2(CD)C-H
-in
S-H
CD0-H
10 ia kg (per hydrogen) jSubstituenti Xc
0.065 ± 0.013
0.55 ± 0.07
2.9 ± 0.58
0.01 ± 0.002 i
!
0.3 ± o.i :
2.5 ± 1.0 j
•
4.0 ± 1.5 j
i
i
i
!
17 ± 4
2.6 ± 1.3 i
j
1.7 ± 0.8 ,
'
H
Cl, Br
F
OH
0-alkyl
-
C-
N<
S-
all cases
1
i
1
1
I
I !
c
1.0
2.4 ± 0.6
1.0
2.0 ± 0.05
6.0 ± 2.0
1.0
1.3 ± 0.2
100 ± 50
200 ± 100
1.0
'Hc :
1.0
0.4 ± 0.1
i
0.3 ± o.i :
i.o :
i
1.0 i
1.0
1.0
1.0
1.0
1.0
1
'
1
aCs " saturated carbon, -0, -C«0, or S; CQ = unsaturated carbon as in vinyl or
pnenyl groups; X » H, F, Cl, Br, or other groups listed in third column.
Rate constant expressed as cm3 molec'1 s"1.
See text for application.
17
-------
TABLE 5. ADDITION RATE CONSTANTS (kg) FOR REACTION OF OH WITH CARBON-CARBON
BOND AND VALUES OF INDUCTION FACTORS (a_)
Substituent
none (ethene)
1-alkyl
1,1-dialkyl
1,2-dialkyl
cis
trans
t
! trialkyl
tetraalkyl
, vinyl or phenyl
: OMe
10:
(per double bond)
7.9
27 ± 5
50 ± 10
60 ± 12
70 ± 14
80 ± 16
150 ± 30
80 ± 20
33
Substituent aT
H - 1.0
F - 0.5 ± 0.3
Cl,Br = 0.7 ± 0.3
Kate constant expressed as cm3 molec"1
See text for application.
18
-------
TABLE 6. ADDITION RATE CONSTANTS (kA) FOR REACTION OF OH WITH AROMATIC RINGS
AND VALUES OF INDUCTION FACTORS (a )
A
Sub&tituent
H
alkyl
. dialkyl
j 1012ka
1
i
I 5.0 ± 2
\
j 12 ± 4
Substituentb| a b
,
H ! 1.0
Cl,F,Br 1 < 1.0
i
!
1,2,3-trialkyl j 10 ± 5
1,2,4-trialkyl ! 25 ± 5
1,3,5-trialkyl I 49 ± 5
methoxy 17 ± 5
OH plus alkyl . 34 ± 10
i
-CH < 1.0
aRate constant expressed as cm3 molec"1 s~x.
See text for application. Effect of substituent may also be
estimated from Figure 1.
19
-------
Best Line for Monosubstituted Benzenes
log k =-11.847-4.193 ap
_ 1.0 -
n
u
o
I
o>
CM
Best Line for Mono- and Poly-
substituted Benzenes
log k = -11.578 -2.120 an
-0.75
0.5
SA-S395-1
FIGURE 1 CORRELATION OF RATE CONSTANTS FOR ADDITION OF OH TO SUBSTITUTED
BENZENES WITH SUM OF SUBSTITUENT CONSTANTS 2an
20
-------
1,2-Dichloroethane: C1CH2CH2C1
H2CC1CC1: 4(0.065 ± 0.013)(2.4 ± 0.6)(0.4 ± 0.1) = 0.25(0.11
0.47)
1,1-Dichloroethane:
HCClaC:
H3CCC12:
Measured value (Table 3) = 0.22
C12CHCH3
(0.065 ± 0.013)(2.4 ± 0.6)* = 0.37(0.17 - 0.70)
3(0.065 ± 0.013)(0.4 ± O.I)2 = 0.03(0.015 - 0.06)
Z = 0.40(0.17 - 0.72)
Measured value (Table 3) = 0.26
3-Chloro-4-hydroxy-cis-butenoic acid methyl ester:
HOCH2 C(0)OCH3
^^f*--fr
Jw**v
cr ^
cis-C-C: (60 ± 12) (0.7 ±0.3) = 42(15 - 72)
-CH2-: 2(2.5 ± 1)(2.0 ± 0.5) = 10(4.4 - 17.6)
CH3-: 3(0.55 ± 0.07) » 1.6(1.4 - 1.9)
C-C-H: (0.01 ± 0.002) - 0.01(0.008 - 0.012)
Z - 54 (21 - 91)
CH2GH9
3-Chloroethylbenzene
ring:
-CH2-:
CH3-:
ring-H:
0.6 (figure 1)
2(2.5 ± 1)
3(0.065 ± 0.013)
4(0.01 ± 0.002)
£
0.6(0.5 - 0.7)
5(3 - 7)
0.2(0.16 - 0.23)
0.04(0.03 - 0.05)
5.84(3.-5 - 7.9)
21
-------
OZONE REACTIONS
The only organics that react with ozone fast enough for the reaction to
be environmentally significant are the alkenes and some of the aromatic hydro-
carbons. Table 7 summarizes the effect of structure on the rate constants.
The substituent increases the reactivity of the carbon-carbon double bond, but
the effects are small. Most all alkenes appear to have rate constants in the
range of 1.6-22 x 10" 18 cm3 molec"1 s"1. These rate constants are about 107
times larger than those for reaction with OH radical. However, the ratio of
the average atmospheric concentrations of ozone to OH is about 10 6 or slightly
larger, so that, on the average, reactions of alkenes with ozone are about
one-tenth as fast as the reactions with OH. In some cases, because of varia-
tions of OH and ozone concentrations in the atmosphere, the ozone reaction tiould
predominate .
The rate constants for reaction of aromatic hydrocarbons with ozone vary
over a wide range, from less than 10~ai to 10" 18 cm"3 molec" * s"1. Because
ozone concentrations range from 0.3-3 x 10 ia molec cm"3, only under the most
favorable conditions can the half-lives be less than a few days, in which case
the ozone reactions compete favorably with the OH reactions. However, for most
aromatic compounds, ozone reactions are unimportant relative to OH reactions.
PHOTOCHEMICAL TRANSFORMATIONS
The rate of photochemical reaction depends on the absorption of solar
energy, which is determined by Beer-Lambert law:
_ ,_ a,C£-
W1* = e x
where Io\ (in photons cm"2 sec-1) is the intensity of incident light at a given
wavelength entering a layer of atmosphere, 1^ (also in photons cm"2 sec"1) is
the intensity of transmitted light, a\ is the cross section (in cm2) at wavelength
X, C is the concentration of chemical (in molecules cm"3), and £- is the path
length of light. The light absorbed at each wavelength by the chemical (I&^)
is expressed
where ax is t*16 absorption cross section in cm2 molec"1 at wavelength X, C is
the concentration and i the pathlength. Under atmospheric conditions equation
(1) may be simplified
ZaX = 'oX'x™
The rate of photochemical reaction is the light absorbed by the compound in the
solar spectrum (Ia) divided by the height (h) of the layer times the quantum
efficiency (<|>) .
22
-------
TABLE 7. RATE CONSTANTS FOR REACTION OF OZONE WITH GENERALIZED STRUCTURES*
Structure
Ethenes
alkyl
1,1-dialkyl
1,2-dialkyl
cis-
! trans-
! trialkyl
' tetraalkyl
Cycloalkenes
C3
c«
Alkadienes
i
1,3-butadiene
Benzene
alkyl
dialkyl
trialkyl
tetraalkyl
i
hexaalkyl
i
hydroxy
cm3 mo lee l s~ l
1.9
12
14
160
260
500
1500
800
170
8.4
0.00005
0.003 - 0.0006
0.001 - 0.002
0.006 - 0.008
0.02
0.4
1-2
ti , daysb
4.2
0.7
0.6
0.06
0.03
0.01
0.006
0.01
0.05
1.0
2 x 10s
2 x 10*
7 x 103
i
2 x 103
I
400
i
20
! 3-7
from Table 3.
24-hour days.
23
-------
I
Rate =
where J^ is the Iox corrected for zenith angle 0 (Cos 0 = h/fc) and is referred
to as the actinic flux.
The apparent first-order rate constant (k ) is thus
k = AZa.J,
p A A
which is conveniently independent of C and h (concentration and depth of the
atmospheric layer). Thus the photochemical reaction rate constant may be
calculated from the measured cross section of a compound, J^ and . Schere and
Demerjian (1977) have published a computer routine to perform this integration
over 10-mm wavelength intervals. The routine calculates momentary photochemical
rate constants throughout the day for any day of the year at any latitude and
longitude. The J^ values of Peterson (1976) are used in the calculations.
Because most compounds require more than one day to photolyze, it is necessary
to integrate the output from the Schere and Demerjian routine over time to
obtain rates on a day basis.
To facilitate the computation of photochemical rate constants, we have
prepared in Tables 8 through 10 day-averaged light intensity values as a
function of wavelength at latitudes of 10, 30, and 50°N for the summer and
winter solstices and the equinox. In addition, the tables include average rates
for spring-summer and fall-winter light intensity values. The light intensity
values in the tables refer to J^' values and represent the total available
actinic light flux in a given wavelength region for a specific day and at a
specific latitude. The wavelength ranges in Tables 8 through 10 are divided
into 31 intervals over the total range 290-800 nm to maintain reliability of
the calculations and yet to allow them to be performed rapidly.
The data in Tables 8 through 10 are multiplied by the average absorption
coefficients over the wavelength ranges in the appropriate table with cor-
responding J,' values and summing
*P • EaxY
where Lj = kp/<(> and is an upper limit of k«; it will equal k- if ^ = 1. The
value of kp corresponds to the latitude ana time of year of the Jx' values used.
To demonstrate the hand computation, Table 11 shows calculations of the
photolysis rates for benzaldehyde and biacetyl.
Estimating the photo rate imposes a major limitation in estimating the
rate constants for the individual processes. The difficulty is caused by the
assumption that photo reaction occurs with a unity quantum yield. This assump-
tion gives a maximum rate photo transformation. Although there are several
photochemical and photophysical reasons why the quantum yield need not be unity,
we will briefly discuss only two of the more important reasons why assuming
24
-------
TABLE 8. JA' VALUES AT 10° N LATITUDE8
Wavelength
Range, nm
285-295
295-305
305-315
315-325
325-335
335-345
345-355
355-365
365-375
375-385
385-395
395-405
405-415
415-425
425-435
435-445
445-455
455-465
465-475
475-485
485-495
495-515
515-535
535-555
555-575
575-595
595-635
635-675
675-715
715-755
755-795
Solstice
Summer
0
6.39 17
9.701 18
2.451 19
4.474 19
4.945 19
5.666 19
5.85 19
7.301 19
7.157 19
7.329 19
1.004 20
1.275 20
1.331 20
1.349 20
1.491 20
1.718 20
1.857 20
1.914 20
1.941 20
1.945 20
3.970 20
4.013 20
3.995 20
4.053 20
4.197 20
8.529 20
8.976 20
9.077 20
8.744 20
8.460 20
Winter
0
3.31 17
6.679 18
1.852 19
3.518 19
3.957 19
4.605 19
4.758 19
5.976 19
5.891 19
6.065 19
8.341 19
1.063 20
1.114 20
1.133 20
1.256 20
1.449 20
1.568 20
1.619 20
1.645 20
1.652 20
3.370 20
3.41 20
3.398 20
3.446 20
3.573 20
7.235 20
7.658 20
7.778 20
7.514 20
7.287 20
Equinox
0
6.38 17
9.522 18
2.386 19
4.341 19
4.789 19
5.479 19
5.656 19
7.655 19
6.913 19
7.075 19
9.687 19
1.23 20
1.283 20
1.30 20
1.436 20
1.655 20
1.789 20
1.843 20
1.869 20
1.873 20
3.823 20
3.863 20
3.846 20
3.902 20
4.040 20
8.211 20
8.637 20
8.732 20
8.410 20
8.135 20
Season Average
Spring/Summer
0
6.63 17
9.815 18
2.457 19
4.468 19
4.828 19
5.638 19
5.820 19
7.226 19
7.125 19
7.278 19
9.966 19
1.265 20
1.320 20
1.338 20
1.478 20
1.702 20
1.840 20
1.896 20
1.922 20
1.926 20
3.932 20
3.974 20
3.956 20
4.014 20
3.659 20
8.446 20
8.884 20
8.961 20
8.650 20
8.367 20
Fall/Winter
0
4.6 17
7.91 18
2.685 19
3.877 19
4 . 321 19
4.735 19
4.925 19
6.357 19
6.343 19
6.663 19
8.927 19
1.135 20
1.187 20
1.205 20
1.339 20
1.538 20
1.664 20
1.716 20
1.742 20
1.747 20
3.565 20
3.605 20
3.591 20
3.692 20
3.774 20
7.652 20
8.077 20
8.186 20
7.897 20
7.648 20
aSecond number in column is the power of ten by which the first number is
multiplied.
bUnits are in photons cm"* day"l as discussed in text.
25
-------
TABLE 9. J,' VALUES AT 30"N LATITUDE'
,a,b
Wavelength
Range , run
285-295
295-305
305-315
315-325
325-335
335-345
345-355
355-365
365-375
375-385
385-395
395-405
405-415
415-425
425-435
435-445
445-455
455-465
465-475
475-485
485-495
495-515
515-535
535-555
555-575
575-595
595-635
635-675
675-715
715-755
755-795
Solstice
Summer
1.0 16
7.40 17
1.09 19
2.74 19
4.98 19
5.49 19
6.28 19
6.49 19
8.09 19
7.93 19
8.12 19
1.11 20
1.41 20
1.47 20
1.49 20
1.65 20
1 . 90 20
2.05 20
2.12 20
2.15 20
2.15 20
4.39 20
4.44 20
4.42 20
4.48 20
4.64 20
9.44 20
9.92 20
1.00 21
9.67 20
9.35 20
Winter
0
6 . 80 16
2.83 18
1.00 19
2.11 19
2.48 19
3.11 19
3.10 19
3.95 19
3.95 19
4.12 19
5.72 19
7.36 19
7.80 19
7.99 19
8.93 19
1.04 20
1.13 20
1.17 20
1.19 20
1.20 20
2.45 20
2.49 20
2.49 20
2.52 20
2.62 20
5 . 39 20
5.72 20
5.85 20
5.70 20
5.57 20
Equinox
0
3.98 17
7.49 18
2.03 19
3.82 19
4.28 19
4.96 19
5.13 19
6.43 19
6.33 19
6.51 19
8.94 19
1.14 20
1.19 20
1.21 20
1.34 20
1.55 20
1.67 20
1.73 20
1.75 20
1.76 20
3.59 20
3.63 20
3.62 20
3.67 20
3.80 20
7.70 20
8.15 20
8.27 20
7.98 20
7.74 20
Season Average
Spring/Summer
1.18 15
6.24 17
9.77 18
2.50 19
4.58 19
5.07 19
5.82 19
6.01 19
7.51 19
7.36 19
7.55 19
1.03 20
1 . 31 20
1.37 20
1.39 20
1.54 20
1.77 20
1.92 20
1.98 20
2.01 20
2.01 20
4.10 20
4.15 20
4.13 20
4.19 20
4.34 20
8.81 20
9.28 20 !
Fall/Winter
0
1.78 17
4.60 18
1.41 19
2.79 19
3.20 19
3.68 19
3.92 19
4.96 19
4.92 19
5.09 19
7.04 19
8.64 19
9.37 19
9.67 19
1.08 20
1.25 20
1.35 20
1.40 20
1.42 20
1.43 20
2.92 20
2.96 20
2.95 20
2.99 20
3.11 20
6.34 20
6.72 20
9.39 20 6.84 20
9.05 20 6.64 20
8.76 20 6.46 20
Second number in column is the power of ten by which the first number is
multiplied.
Units are in photons cm"3 day"1 as discussed in text.
26
-------
TABLE 10. Jx' VALUES AT 50°N LATITUDE3'b
Wavelength
Range , nm
285-295
295-305
305-315
315-325
325-335
335-345
345-355
355-365
365-375
375-385
385-395
395-405
405-415
415-425
425-435
435-445
445-455
455-465
465-475
475-485
485-495
495-515
515-535
535-555
555-575
575-595
595-635
635-675
675-715
715-755
755-795
Solstice
Summer
0
5.7 17
1.005 19
2.680 19
5.0 19
5.577 19
6.446 19
6.659 19
8.341 19
8.204 19
8.427 19
1.157 20
1.472 20
1.540 20
1.564 20
1.732 20
1.996 20
2.159 20
2.227 20
2.261 20
2.268 20
4.629 20
4.682 20
4.664 20
4.732 20
4.903 20
9.935 20
1.048 21
1.061 21
1.023 21
9.902 20
Winter
0
6.0 14
2.981 17
1.857 18
5.429 18
7.118 18
8.513 18
9.536 18
1.248 19
1.279 19
1.070 19
1.951 19
2.569 19
2.792 19
2.915 19
3.326 19
3.91 19
4.30 19
4.533 19
4.685 19
4.769 19
9.789 19
1.001 20
1.002 20
1.014 20
1.063 20
2.227 20
2.463 20
2.610 20
2.644 20
2.648 20
Equinox
0
1.12 17
3.945 18
1.317 19
2.714 19
3.163 19
3.86 19
3.923 19
4.985 19
4.967 19
5.165 19
7.161 19
9.194 19
9.715 19
9.938 19
1.109 20
1.283 20
1.392 20
1.443 20
1.471 20
1.483 20
3.026 20
3.069 20
3.062 20
3.104 20
3.228 20
6.631 20
7.015 20
7.165 20
6.977 20
6.809 20
Season Average
Spring/Summer
0
3.763 17
7.724 18
2.173 19
4.162 19
4.700 19
5.495 19
5.670 19
7.130 19
7.038 19
7.255 19
9.987 19
1.274 20
1.337 20
1.361 20
1.510 20
1.743 20
1.886 20
1.948 20
1.981 20
1.983 20
4.060 20
4.109 20
4.095 20
4.153 20
4.308 20
8 . 749 20
9.253 20
9.40 20
9.089 20
8.821 20
Fall/Winter
0
2.55 16
1.398 18
5.699 18
1.314 19
1.597 19
1.984 19
2.057 19
2.648 19
2.646 19
2.816 19
3.949 19
5.132 19
5.478 19
5.652 19
6.366 19
7.413 19
8.087 19
8.444 19
8.659 19
8.758 19
1.791 20
1.823 20
1.822 20
1.845 20
1.926 20
3.983 20
3.428 20
4.431 20
4.370 20
4.311 20
aSecond number in column is the power of ten by which the first number is
multiplied.
bUnits are in photons cm"2 day"1 as discussed in text.
2:7
-------
TABLE 11. EXAMPLES OF HAND COMPUTATION OF PHOTOCHEMICAL RATE CONSTANTS0'
285-295
295-305
305-315
315-325
325-335
335-345
3A5-355
355-365
365-375
285-295
295-305
305-315
315-325
325-335
335-345
345-355
355-365
365-375
375-385
385-395
395-415
415-435
Avg. Cross-
Section (o.)
A
Benzaldehyde Day
7.10 -20
3.33 -20
3.84 -20
4.30 -20
4.22 -20
3.42 -20
2.76 -20
1.32 -20
7.48 -21
Biacetyl Day Rat
1.03 -20
4.98 -21
2.04 -21
1.11 -21
9.61 -22
2.19 -21
3.94 -21
6.10 -21
8.68 -21
1.24 -21
1.88 -20
2.37 -20
2.96 -20
j'Value
A
Rates at 30°N
1.0 16
7.4 17
1.09 19
2.74 19
4.98 19
5.49 19
6.28 19
6.49 19
8.09 19
es at 30°N on
1.0 16
7.4 17
1.09 19
2.74 19
4.98 19
5.49 19
6.28 19
6.49 19
8.09 19
7.93 19
8.12 19
1.11 20
1.41 20
VI
on Summer Sol
7.10 -4
2.46 -2
4.19 -1
1.18
2.10
1.88
1.73
8.57 -1
6.05 -1
IavY- fl
stice
7.10 -4
2.54 -2
4.44 -1
1.62
3.72
5.60
7.33*
8.19
8.80
k - 8.80 (T1
Pl
Summer Solsti
1.03 -4
3.69 -3
2.22 -2
3.04 -2
4.79 -2
1.20 -1
2.47 -1
3.96 -1
7.02 -1
9.83 -1
1.53
2.63
4.17
1.03 -4
3.79 -3
2.60 -2
5.64 -2
1.04 -1
2 -.25 -1
4.72 -1
8.68 -1
1.57
2.55
4.08
6.71
10.88
k • 10.88 cT1
P
Cross-section data from Berger, 1973.
Cross-section data from Calvert and Pitts, 1966.
*"Second number in column is the power of ten by which the first number is
multiplied.
Units are in photons cm"a day"1 as discussed in text.
28
-------
unity quantum yield will often give high photo rates.
First, the reaction may have a thermodynamic cutoff at a wavelength shorter
than the wavelength where all or partial absorption occurs. In an extreme case,
the thermodynamic cutoff is at a wavelength beyond the solar region where,
if absorption occurs, the energy will be insufficient for any reaction.
Second, many photochemical reactions may be extremely sensitive to quenching.
For example, reactions that proceed via a triplet state very often are rapidly
quenched by oxygen. The interaction of ground state oxygen, which is a triplet
having a low-lying singlet excited state, with excited organic triplet to give
the ground state singlet is an electronically allowed transition. Thus the
reaction for a compound C with a singlet ground state would be
C + hv *- C (singlet) »- C (triplet)
C (triplet) + 02 —*- C + 02 (singlet)
In some case enhanced intersystem crossing occurs without formation of excited
oxygen:
C (triplet) + 02 *~ C + 02
The effect of oxygen is not only on triplet states. For example, the singlet
states of many aromatic hydrocarbons are believed to be readily quenched by
oxygen. In these cases the process is referred to as catalyzed intersystem
crossing (Turro, 1978).
The interaction of excited states with oxygen need not always lead to the
ground state. Oxygen may react with excited states which otherwise would not
have enough energy to react and which would have returned to their ground state
in the absence of oxygen. This has been observed in the photolysis of biacetyl
(Porter, 1960; Bouchy and Andre, 1977); yet the number of examples where this
has been reported is small.
29
-------
SECTION 5
PRODUCTS FROM REACTION WITH OH
As discussed above the reactions of OH with organic molecules are composed
of three basic types: (1) abstraction of hydrogen-atoms,
-C-H + OH
-A. +
I
OH
X
(2) addition to carbon-carbon double bonds,
+ OH —>- HO-C-C*
and addition to aromatic rings
H OH
OH
The first step in predicting the stable products that result from these
reactions in the atmosphere is to determine the relative amounts of these three
pathways. For compounds with only one of these pathways open to reaction, the
problem is trivial. For compounds having more than one pathway, it is necessary
to estimate the rates of each pathway and the relative importance of each. The
rate constant for OH attack is estimated by calculating the rate constant for
each process. Thus
OH
Ctotal
OH OH OH
= k (abstraction) + k (C=C addition) + k (ring addition)
The actual constants are estimated as discussed in Section 4. The relative
importance of three terms is equal to the relative importance of three terms
in the above equation. In the examples on page 21, for 1,2-dichloroethane and
1,1-dichloroethane, only H-abstractions are important because they contain no
carbon-carbon double bonds on aromatic rings. For the third example,
3-chloro-4-hydroxy -cis-butenoic acid methyl ester, the addition to the C=C
double bond is 42/54 or about 78%.' The remainder is abstraction at the -CH2-
group (10/54 or 19% and the -CH3 group (1.6/54 or 3.0%) with less than 0.1%
involving abstraction of the vinyl hydrogens.
In the fourth example on page 21, there is
30
-------
86% -CH2- abstraction.and 10% of aromatic ring addition. There is
also about 3% abstraction from the CH3- and about 0.7% abstraction of the
aromatic rings hydrogens.
Once the relative importance of each pathway is estimated, the products
and their relative importance must be predicted from each pathway.
PRODUCTS RESULTING FROM HYDROGEN ABSTRACTION
Abstraction of hydrogen atoms from most organic compounds results in
formation of a free radical with the free valence located on a carbon atom.
--H + OH — *- -• + HOH (2)
Under atmospheric conditions these carbon radicals react rapidly with oxygen
(3)
There are several exceptions to this generalization. One exception is a
carbon radical that has an adjacent -OH group. 0-H-atom transfer then gives
the corresponding carbonyl compound as shown in reaction (4) :
+ Oa — +- H02- + \=0 (4)
This alternate pathway may also be important when the 0 is replaced by N and
S; however, we do not have sufficient data to ascertain this fact.
A second type of exception to reaction (3) involves cyclohexadienyl
radicals. In this case the reaction is
H02» (5)
This exception is important in the reactions of OH with aromatic compounds
and will be considered further in Section 4.
ATMOSPHERIC REACTION OF PEROXY RADICALS
Hydrogen atom abstraction usually leads to formation of the corresponding
peroxy radicals as discussed above. The fate of the peroxy radicals is to
react with NO even at the extremely low concentrations (^ 1 x 10* molec cm" 3 ,
0.04 ppb) . The main reaction is
0« + NO - +- -io» + N02 (6)
31
-------
An alternate reaction may occur in a fraction of the time depending on the
organic structure (Darnall, 1976):
Thus, the nitrates of all corresponding peroxy radicals should be considered
as possible major products (up to 50% of the amount of peroxy radical formed).
The reaction of peroxy radicals with N02 is also fast and occurs more
readily than with NO because of the generally greater than unity value for
the NO2/NO ratio.
R00» + NO2 < > ROON02
However, to our knowledge, the reaction is totally reversable, and no products
result from the interaction. The ROON02 intermediates can build up to
significant concentrations when R equals acyl and aroyl groups, as in the
peroxyacyl nitrate (I) and peroxybenzoyl nitrate (II):
CH3C02N02 PhCOONOa
I II
Therefore, except for the fraction of peroxy radicals giving nitrate
directly from the reaction with NO, the products are determined from the
chemistry of the alkoxy radicals.
CHEMISTRY OF ALKOXY RADICALS
Under atmospheric conditions alkoxy radicals undergo both bimolecular
and unimoelcular reactions at sufficient rates to be important.
Bimolecular reactions—The important bimolecular reactions of alkoxy
radicals are with 02, NO, and N02. The reaction with oxygen is important
when o-hydrogens are present in the radical and the 3-elimination reactions
are relatively slow. The reaction
+ 02 ^)=0 + H02- (8)
is the major atmospheric reaction for CH30« and CH3CH20» and one of the
major reaction routes for sec-alkoxy radicals.
The reactions with NO and N02 occur with all alkoxy radicals to give
nitrite and nitrate, respectively, by the reactions
R0» + NO —+- RONO (9a)
R0« + NO2 *- RONO2 (10a)
32
-------
However, the fate of the alkyl nitrite formed in reaction (92) is to rapidly
(t^ < 0.5 hr) photolyze (reforming the alkoxy radical, see Section 7 ) and
thus is unimportant. In cases wheTre the alkoxy radicals have o-hydrogens a
small fraction of the time, an alternate reaction occurs:
5
-C-O- + NO —*- C=0 + HNO (9b)
-C-0- + N02 *y*0 + HN02 (lOb)
At the maximum concentrations of NO and N02 (^0.1 ppm), reactions (9a) and
(lOa) are usually unimportant but might be important in special cases depending
on the competing reactions. Reactions (9b) and (lOb) will always be less
important. Because these reactions produce the same products as reaction (8),
which will always be competitive, they can be ignored for the purposes of
identifying products that might be formed.
Unimolecular reactions—Two types of unimolecular reactions are possible.
Generally the most important are the 0-elimination reactions, which result in
major degradation of the organic structures. The general reaction is
X-C-0 »-X« + >0 (11)
where X represents various types of groups and will vary within each alkoxy
radical in most cases. The rate of reaction is determined not only by the
difference in the bond strength of the carbonyl group that is formed over that
of the X-C bond as it is broken, but also by the changes in the bond strengths
of the other X-C bonds that are not broken. Thus, the rate of eliminating one
X is determined by the other Xs that are not broken.
A simple approach to this problem is to assume that each of three X groups
in any alkoxy will 0-eliminate enough so that the products from all pathways
should be considered. In many cases, this approach would predict more products
than necessary, but would err on the safe side. We believe some generalizations
can be made that can be readily applied to this problem and that will be help-
ful in minimizing the number of products that are predicted.
Baldwin et al. (1977) have analyzed the available kinetic data on the
decompositions of simple alkoxy radicals, and the following generalizations
can be made about applying the Arrhenius equation [equation (2)] to these re-
actions.
k. = Ae~Ea/RT (12)
d
The A factor for the decomposition of primary, secondary, and tertiary alkoxy
are 101,9'6 * °'2, 101"5 ± °"2, and 1013'a ± °'2 s"1, respectively. In all
cases the activation energy is estimated from
33
-------
E = 12.8 + 0.71 AHJ (kcal/mol) if AH^ > 0
= 12.8 if AHJ S 0
Thus from knowledge of the type of alkoxy radical and the heat of reaction
(AHf) for elimination of each of the three groups estimated using techniques
discussed by Benson (1976), the rate of reaction for each elimination may be
calculated. Table 12 lists the estimated rate constants (kj) for the
elimination of X from the primary radical X-CH20», the secondary radical
X-CH(CH3)0», and the tertiary radical X-C(CH3)20«. These rate constants are
also good approximations for the cases where the methyl groups are replaced
with any group where bonding is through carbon.
As an example showing how the rate constants in Table 12 were determined
we may consider the reaction
CH3C(0)CH20- *~ CH3C(0)» + CH20
The heat of reaction AH^ is estimated to be 1.3 kcal/mole from the heats of
formation of each species. The heat of formation of the initial
radical was estimated from the heat of formation of the corresponding alcohol
(group additivity) by assuming an 0-H bond strength of 104 kcal/mole; the heats
of formation of CH3C(0)« and CH20 were obtained from standard references
(Benson, 1978). Thus using the above equation, Ea = 13.7 kcal/mole. Since
the reaction involves cleavage of a primary alkoxy radical, A = 1013*2 s~ l
and k = 1013*a e~13700/RT = 1.7 x io3 s" x at 300°K. Because of the uncertainties
in A and Ea, k has uncertainty of a factor of 10. In cases where the appro-
priate group values were not available or could not be readily approximated,
the heats of reaction were estimated by assuming that the difference in bond
strengths of X-CH3 and CH3-CH3 is the same as the difference in heat of re-
action for the elimination of X from X-CH20» and CH3 elimination from CH3-CH20«.
Also included in Table 12 are the apparent first-order rate constants for
reaction with 02 (5 x IO18 molecules cm"3, 160 torr) and N02 (2.3 x IO13
molecules cm"3, 0.1 ppm). This latter value corresponds to the maximum con-
centrations found in a large urban area; in nonurban areas, the value can be
lower by a factor of 0.01 to 0.1, in which case the reaction with N02 will be
that much smaller than the values in Table 12. The rate constants for intra-
molecular rearrangements are also included and will be discussed below.
Heteroatoms (Y) have various effects on the elimination rates. The effect
is most strongly dependent on the change in bond strength of the
carbon-heteroatom bond in the elimination process as a second group is eliminated
in the reaction
(13)
The most dramatic effect is when an oxygen atom is in an or-position, for
example
34
-------
TABLE 12. ESTIMATED RATE CONSTANTS FOR REACTIONS OF PRIMARY, SECONDARY, AND
TERTIARY ALKOXY RADICALS
• Reaction/Group
B-Elimination
I-
N02-
Br-
PhCHa-
CH2=CH-CH2-
CH3C(0)-
(CH3) 3C-
(CH3)2CH-
HOCHa-b
Cll3Cll2~
CH3-
CF3-, CH2=CH-
H-
Ph-
F-
Rx with 02 andNOa
Oa
N02(.l ppm)
6-H Rearrangements
primary 6-H
secondary 6-H
tertiary 6-H
Rate Constant, s"1
X-CH20«
1.9 x 10"
1.9 x 10"
1.9 x 10"
1.9 x 10"
1.9 x 10"
1.6 x 10
3.0 x 10l
8 x 1CT1
2 x 1(T lb
0.2
7.3 x ID"3
2.2 x 10-9
4 x 10-8
0.5 x 10-'
1 x 10-13
4.6 x 103
4.0 x 101
6.0 x 10s
3.3 x 106
3.7 x 107
X-C(H)(CH3)0»
1.5 x 103
1.5 x 10s
1.5 x 10s
1.5 x 103
1.5 x 10s
1.5 x 10s
1.5 x 10s
2.2 x 10"
6.8 x 103
6.8 x 10s
1.9 x 10*
1.0 x 10
7.8 x 10-°
1.8 x 10~"
6.5 x 10—
6.4 x 10s
4.0 x 10l
6.0 x 103
3.3 x 106
3.7 x 107
X-C(CH3)20«
7.5 x 10s
7.5 x 10s
7.5 x 10s
7.5 x 103
7.5 x 10s
7.5 x 103
7.5 x 103
5.3 x 103
1.6 x 103
1.6 x 103
4.5 x 103
1.4 x 103
2.8 x 10-*
4.6 x 10—
6.2 x 10-'
0
4.0 x 10l
6.0 x 10s
3.3 x 10*
3.7 x 107
See text for method of estimation; rate constant at 300°K uncorrected for
pressure effect; at one atmosphere, k/k (at infinite pressure) = 0.003 for 8 atom
radical, 0.5 for 11 atom radical, 0.8 for 14 atom radical and 1.0 for larger
radicals (Baldwin, 1977).
Limited experimental data suggest this rate constant is effectively 10s times
faster than indicated or competing reactions are correspondingly slower.
35
-------
CH3
Compared to methyl substituted derivatives (X-CO«), estimates of heat of re-
action favor elimination of X and H by H several orders of magnitude.
In Table 13 we estimated the rate constant for the same substituents for the
two cases
OMe OMe
X-£-0» and X-C-0-
'f0'
The same acceleration effects are expected for all ether groups and -OH. An
N-atom in the or position produces approximately the same acceleration in
P-elimination as is in -OMe elimination. Thus the data in Table 13 are good
estimates for these cases.
The effect of a-halogen atoms depends strongly on the halogen. Fluorine
is more like an oxygen substituent whereas iodine is more like a carbon group.
Chlorine and bromine are intermediate between fluorine and iodine. Thus, Tables
12 and 13 can be used in many cases to qualitatively determine the important
elimination rate constants when a-halogens are present, assuming their effect
is intermediate.
The second type of unimolecular reaction is intramolecular rearrangements
which are important for some long chain alkoxy radicals such as
J3
->S™- rti n. k^ ""vi ym QJJ (14)
,6
The rate constants for these reactions have been estimated by Baldwin et al.
(1977) for several structures, and these estimates have been included in Tables
12 and 13. The reactions are fast only when there is a hydrogen atom in the
6-position because of the favorable six-membered ring that forms on abstraction.
In cases where only five-membered or seven-membered ring rearrangements are
possible, the rate is much less and presumably other reactions will dominate.
These six-membered rearrangements result in the formation of disubstituted
groups. For the example, in reaction (14) the following reactions will result:
Cf£\Bb ^Ha
(15)
(16)
Una
(6-Rearrangements of R02» are too slow to compete with reactions with NO.)
36
-------
TABLE 13. ESTIMATED RATE CONSTANTS FOR REACTIONS OF METHOXY SUBSTITUTED
ALKOXY RADICALS*
— — — ^— ^— ^ — —
React ion /Group
P-Elimination
I
NO 2
Br
PhCHa-
CH2=CH-CH2-
CH3C-
(CH3)3C-
(CH3)2CH-
HOCHa-b
CH3CH2-
CH3-
CF3
CH2=CH-
Ph-H
H-
F-
R with 02
X
R with N02
V
(0.1 ppm)
6-H Rearrangement
Primary 6-H
Secondary 6-H
Tertiary 6-H
Rate
X-CH(OMe)0«
1.5 x 10s
1.5 x 10s
1.5 x 10s
1.5 x 103
1.5 x 10s
1.5 x 10s
1.5 x 10s
1.5 x 103
1.5 x 105
1.5 x 103
1.5 x 10s
1.5 x 10s
2.0 x 101
0.8
1.8 x 102
9.0 x 10-10
6.4 x 103
4.0 x 101
6.0 x 10s
3.3 x 106
3.7 x 107
Constants, s"1
X-C(OMe)(Me)0« !
7.5 x 10s
7.5 x 10s
7.5 x 103
7.5 x 103
7.5 x 10s
7.5 x 10s
7.5 x 105
7.5 x 10s
7.5 x 103
7.5 x 103
7.5 x 10s
7.5 x 10s
2.0 x 103
6.0 x 102
1.5 x 105
1.3 x 10-2
0
4.0 x 101
1
6.0 x 10s |
3.3 x 106
3.7 x 107
aSee text for method of estimation; rate constant at 300°K uncorrected for
pressure effect; at one atmosphere, k/k (maximum pressure) = 0.003 for 8 atom
radical, 0.5 for 11 atom radical, 0.8 for 14 atom radical and 1.0 for larger
radicals (Baldwin, 1977^.
Reaction 8 will dominate.
37
-------
CH3
H
8
Other variations can lead to CH3-C-CH2CH2CH and CHaCCHaCHaCOH; condensation
reactions of the various products, which would require a heterogeneous surface,
can lead to the following structures:
9vx«Xs CHs^xO ^
LJ L_f
These structures are generalized below in the diagram of the products from
Intermolecular hydrogen atom transfer:
Starting structure:
Products predicted:
IH
H
^TTT Jx ^WT
\H2OH
CH
W"
To illustrate how the discussion in this section is applied, we have
included Figure 2 which summarizes the reactions of 1,1-dichloroethane in the
atmosphere. From this example it is seen that the reaction of one alkoxy
radical leads to a second carbon radical, the reactions of which must also be
analyzed to determine its fate. The stable products must also be analyzed to
determine their lifetime and products of reaction if the total effect of
initial compound under consideration is to be determined.
The distribution of products may be estimated from the rate constant data,
However because of the uncertainty in the estimated rate constant, it can only
be considered qualitative and to point out the relative importance of the
possible products. The products in this example greater than 0.01% along
with their estimated relative importance are calculated from the relative
importance of the pathways as follows:
38
-------
w
(D
CH30» -H^ CH20
CH.-
0.37
(92.5%)
IH
0.03
(7.5%)
PI W
V r
H-C-C-
Cl H
I
1
i-
l H \
NO
X
(^50%)
\NO
1 X Cl
H-C-CHaONOa
Cl
NO
Cl
HC02
Cl
50%
Cl
HC-ONOa
Cl
50%
°2 > C1CC1 + H0a»
FIGURE 2. PREDICTION OF PRODUCTS FOR 1,1-DICHLOROETHANE
HCO«
Cl
-------
CH3C(Cla)ONOa = 0.925 x 0.5
= 0.46
C12CO = 0.925 x 0.5 x 0.5
= 0.23
CH20 = 0.925 x 0.5 x 0.5
= 0.23
HCC12CH2ON02 = 0.075 x 0.50
= 0.032
HCC12CHO = 0.075 x 0.50 x 0.99
= 0.038
HCC12CH2ON02 = 0.075 x 0.50 x 0.01
= 3.8 x ICT*
Thus the first three products should be considered major products (> 10%),
and the next two considered minor products (< 10% but > 1%). The last along
with HC(C12)CHO and HC(Cla)ON02 would be trace products.
PRODUCTS RESULTING FROM ADDITION OF OH TO C=C DOUBLE BONDS
The OH radical adds rapidly to carbon-carbon double bonds and is usually
the major reaction pathway with compounds having this structure. The simplest
and best studied example is the reaction of ethene:
CH2=CHa + OH —+- HOCH2CHa» (19)
The carbon radical formed follows the same chemistry as outlined in the
previous section. That is, it reacts first with Oa and then NO according to
the reactions
HOCH2CH2- + 02 —>~ HOCH2CH200» (20)
HOCH2CH200» + NO —>~ HOCH2CH20- + N0a (21)
The resulting alkoxy radical should react also as discussed in the previous
section. Niki et al. (1978) have shown that the detectable product is
formaldehyde, which indicates that B-elimination of »CH2OH dominates the re-
action
HOCH2CH20» +- HOCH2- + CH20 (22)
and then oxygen reacts with the resulting radical as follows:
HOCH2- + 02 —*- H02- + CH20 (23)
From the information in the previous section, we also expect the following
reaction to occur:
40
-------
CH +
HOCH2CH20» + Oa *- HOCH2CH + H02» (24)
The high yield of formaldehyde rules out appreciable amounts of this reaction.
Why the predictive scheme does not anticipate this effect is not clear; the
scheme may reflect a unique chemical behavior of o-OH alkoxy radicals or
possibly a more general flaw in the procedure when applied to systems containing
heteroatoms.
Although the study of the reactions of other olefins under simulated
atmospheric conditions are complicated by reactions of ozone, they further
support the idea that OH effectively cleaves the carbon-carbon double bond,
resulting in two carbonyl-containlng molecules as seen for ethene by Niki et
al. (1977). Thus, the reaction of olefins and OH is best generalized as
follows:
(25)
This generalization may suffer from over simplification, and two questions
need to be answered. First, if the following reaction of 0-hydroperoxy radical,
the intermediate formed in reaction (20), ever important?
HOC-C-00* + NO —*- HO-C-C-ON02 (26)
In the limited number of olefins studied, there is no evidence for such a
reaction although a search was not specifically made. However, because peroxy
radicals from large alkanes (> Cs) have been suggested to undergo this type
of reaction a fraction of time, it is not clear why such a reaction should
not be important for the olefin reactions where similar intermediates are
formed.
A second problem that deserves further study is what controls the cleavage
of the p-hydroxy alkoxy radicals in reaction (22). From our discussion in the
previous section the reaction,
HOCH2-C-0- —>• HOCH2-CH + R- (27)
should compete if the AHR is more favorable than the alternate pathway, which
leads to C=C cleavage x
HOCHa-C-0» *- HOCHa- + R-CH (28)
The effect should be most important when R is a group considerably above HOCH2-
in Tables 12 and 13; that is, where R = allyl, benzyl, Br, N02, and I.
Apparently no one has investigated olefins to see whether this might be the
case. In such cases, a-hydroxy aldehydes or ketones would be expected rather
41
-------
than the two carbonyls that result from C«C cleavage.
On the other hand, the chemistry of 3-hydroxy alkoxy radicals may be
unique, possibly due to the intramolecular interaction
0 0
-W
which could favor cleavage to form an HOC* radical [reaction (28)], and this
interaction would be independent of competing reactions, either reaction (24)
or (27). Further research is needed on this subject if we are going to optimize
the reliability of our methods for predicting tropospheric products for complex
compounds .
The abstraction of vinyl hydrogen is believed to compete with OH addition.
Although the amount of this reaction is expected to be small compared with
addition (^ 0.1%), the products can be sufficiently different in structure
that they should be noted. The abstraction route for ethene results in double
bond cleavage
V/*
CH2-CH °g/NO> & JJH
i2—un * i>ii2i>n —*• CH20 4- HCO* > CO + H02«
For complex structures, R-CH-CH-R1, the following intermediate is obtained:
R-CH-C-R1
which can decompose (Section 3)
R-CH-CR1 »~ R» + HC-CR'
To Illustrate the prediction of products using equation (25), we have
summarized in Figure 3 the OH reaction pathway and products for example 3 of
the tabulation on page 21 (subsection Hydroxyl Radical Reactions). In this
example, addition and cleavage of the carbon-carbon double bond is expected
to predominate (^ 80% of reactions). Hydrogen abstraction reactions will
account for the remaining products.
PRODUCTS RESULTING FROM ADDITION TO AROMATIC COMPOUNDS
The OH radical adds at a rapid rate to aromatic rings. For example, the
reaction between OH and toluene is believed to give largely ring addition with
a small fraction of abstraction from the benzylic methyl group (^ 15%). The
reactions in the atmosphere are believed to be
42
-------
CO
HOCH
42 x 10-
8
HO-CH-CH=C-COCH 3
ri H
\*JL n
,CC1 + HCCOCHa
» HO-
CH-CH=C-COCH3 *- HC-CH=CCOCH3
Cl H Cl ft
HOCHa-C=CCOCH20« -2i
HOCHa-C=CCOi
0.01 x 10-12
(0.02%)
HOGHa-C=CcOCH3 a > trace products
FIGURE 3. PREDICTION OF PRODUCTS FOR 3-CHLORO_4-HYDROXY-iCIS-BUTENOIC ACID METHYL ESTER
-------
CH2
CH3
:H2o«
CHO
02
+ isomers
0
+ H02-
+ isomers + H02-
The distribution of cresols are o-: m- : p- = 1.0: 0.06: 0.17; thus 80% of the
cresols is the ortho isomer. In smog chamber experiments designed to simulate
atmospheric conditions the amounts of cresols suggest alternate routes that
involve cleaving the aromatic ring. For example, in the case of toluene,
formaldehyde and peroxyacetyl nitrate are formed by an unknown mechanism.
CHS
+ OH
CH3C02
CH20 + C02 + H02-
(29)
CHS
N0
It is not clear at this time whether this alternative involves cresol as an
intermediate of if the formation occurs by a route parallel to cresol formation.
Ring cleavage is also assumed to form dicarbonyl compounds, such as glyoxal
and substituted glyoxals, although they have not been detected. For product
estimation, the best generalization appears to be that addition of OH to aromatic
leads to phenol formation and ring cleavage in equal amounts with the following
breakdown:
(30)
44
-------
RCCH + 2 HCCH
+ .75 RCCH + 2 HCCH (31)
RCOC
ON02
:oo-^ (32)
R0» (treat as in Section 5)
The experimentally determined rate constants for addition of OH to aromatic
rings indicate that the rate is very sensitive to the type of ring substitu-
tion. In Figure 1 we have plotted the rate constants reported by Perry et al.
(1977) as addition rate constants versus the summation of para ap-constants
(Chapman and Shorter, 1972). Because OH is an electrophilic species, a cor-
relation with the electron donating-withdrawing properties of the substrate
is a reasonable assumption. Summing of o-constants for the various
substituents oversimplifies the effect of multiple substituents because there
is no way to account for the effects of positional isomers. However, as seen
in Figure 1, a reasonable correlation of the data covers a factor of 50 in
rate constants. The least squares regression line of all data gives
log k^X = 11.578 - 2.12EO (33)
A P
The least squares regression line for only the monosubstituted structures gives
log k* = - 11.847 - 4.19a
A P
The values of kA determined from these regression lines are expected to be
reliable estimates of k. (± 20%).
Unfortunately, no data are available for strong electron-withdrawing
substituents, except for a limit value for X=-CHO where abstraction of the CHO
hydrogen is very fast. From equations (33) and (34), it is possible to estimate
the k values of compounds with such substituents, using the list of a-values
in Table 14 taken from Chapman and Shorter (1972).
Competing with addition to aromatics is abstraction of the aromatic
hydrogen. This reaction is slow compared to addition but is thought to produce
different products. Two types of products are expected: quinone and ring
cleavage products. The following are examples of the proposed chemistry for
benzene:
45
-------
TABLE 14. SELECTED VALUES OF HAMMETT or
Substituent
H
D
B(OH),
He
Et
Pr
Bu
CHCH
C=CPh
cyclo-CsHr
Ph
CH,Ph
CH,CN
CHaOR
CHjSOjCFj
CH,C1
CF,
CClj
CN
CHO
COMe
C(0)NH,
CO,H
CO,R
SIMe,
NB,
NMe3
NHAc
NHBz
NHC(0)CF,
N,
o
p
0
0
0.12
-0.16
-0.14
-0.15
-0.18
0.23
0.16
-0.21
0.02
-0.09
0.18
0.03
0.31
0.12
0.55(0.
0.46
0.69
0.43
0.44
0.38
0.42
0.43
0.00
-0.57
-0.60(-0
-0.09
-0.19
0.12
0.78
SubaCicuent
NCS
N-NPh
N,
NO,
PPh,
POPhi
PO(OEt)2
OH
OMe
OCF,
OPh
OC(0)Me
SH
SMe
SCF,
S(0)Me
54) S(0)3Me
S(0),CF,
S(0),NHa
SO,F
SF,
SeMe
SeCF,
SC(0)Me
F
Cl
.83) Br
I
10,
NMe,
SMe,
°p
0.38
0.33
0.15 !
0.80 1
I
0.19 I
i
0.53 j
0.60
-0.36 (-0.37)
-0.27 (-0.27)
0.35
0.32
0.32
0.14
0.16
0.15 .
-0.05(0.00) '
0.42
0.57(0.49)
0.73
0.93
0.91
0.60
0.91
0.68
0.0
0.38
0.42
i
0.06(0.06) ,
0.23(0.23)
0.27
0.96
"Chapman and Shorter (1972).
46
-------
+ OH
Oa/NO
(35)
Oa/N0
Oa
CHO
(36)
ring cleavage
products
(37)
The ring cleavage products are expected to be similar to those in reaction (31)
To demonstrate the application of these techniques in Figure 4, we have
analyzed compound 4, 3-chloroethylbenzene, from the tabulation on page 21.
From equation (33), when k0^ = 1.5 x 10" 1S molec cm"3 s"1, we see in Figure 4
that attack at the benzylic -CH2- group accounts for ^ 93% of the reaction;
thus m-chloro benzaldehyde and n-chloroacetophenone are expected to be major
products. Ring addition, which is relatively unimportant because of the de-
activating effect of Cl, gives ^ 1% of the phenol isomers and similar amounts
of products from ring cleavage. Attack at the CHS groups also gives
m-chlorohenzaldehyde (4%). Abstraction from the ring will also lead to 1% of
products from ring cleavage.
47
-------
»CH,
Cl
1.5 x 10" '
5 x 10-"
0.2
CHaCH,
(Isoners)
CHjCHaCOO- + CH,CHaCCH +
\
CH.CHjoONOa
CH,CHaO,« -j-- CH,
,CH
O,NO
o,
CHj-CHjO»
Oj/NO
CH
Uia
-H« | «- ring cleavage products
"Cl
+ CH,'
CH20
+ CHaO
tOj/NO
CHaO*
CHO
ci
FIGURE 4. PREDICTION OF PRODUCTS FOR 3-CHLOROETHYLBENZENE
48
-------
SECTION 6
PRODUCTS FROM THE REACTIONS OF OZONE
The most important reactions or organic compounds with ozone in the atmo-
sphere are the reactions of olefins. The reaction is believed to involve the
initial formation of molezenide
X
--
(37)
which, at atmospheric temperatures, readily decomposes by the following re-
actions:
»
0-
-Xc-c-
III
'\/N ozonide
0
IV
Reactions (39) and (40) are believed to be the major pathways, but reaction
(38) cannot be ruled out. The initially formed products are one mole each of
carbonyl compound and the percarbony.1 intermediate, referred to as the switterion
or methylene peroxy radical; the electronic structure of this latter intermediate
appears to be singlet biradical. (Wadt and Goddard, 1975).
Under special conditions the ozonide (IV) has been reported (Niki et al.,
1977) but at typical atmospheric concentrations is not expected to form.
O'Neal and Blumstein (1973) have suggested that the diradical intermediate
(III) might have a lifetime-long enough so that intramolecular reactions, such
as the following one, could be important:
(43)
49
-------
However, new data on alkoxy cleavage (Baldwin et al., 1978) suggest that the decom-
position by reaction (41) would proceed too fast under environmental conditions
for reaction (43) to be important. Experimental results explained by this re-
action should be reconsidered to determine if reactions of OH with the alkene
can account for the same products.
There now appears to be no question as to the intermediary of the per-
carbonyl intermediates but their chemistry is not well understood. Unimolecular
rearrangements of percarbonyl intermediates appear to dominate any bimolecular
reactions with NO, N02, 03, or even Oa. The simplest and probably the most
rapid reaction is collapse of the biradical to the three membered ring structure,
dioxirane:
p
^C-00« *- -C-fo (44)
dioxirane
The simplest dioxirane intermediate has been detected in the reaction of
ozone and ethene by Martinez et al. (1977), using photoionization mass spectro-
metry and by Lovas and Svenram (1977), using microwave spectroscopy. However,
above -100°C this intermediate decomposes rapidly even at low pressures, which
should dramatically stabilize molecules of such low molecular weight; therefore,
the intermediate is expected to be extremely labile at atmospheric conditions.
The pathway open to the dioxirane structure is cleavage of the extremely weak
(^ 12 kcal/mole) 0-0 bond to form the dialkoxy radical, dioxymethylene.
-CO) ^ -C-0- (45)
Wadt and Goddard (1977) have suggested that in the simple case, CH2(0»)2,
hydrogen atom migration to form formic acid is the favored reaction:
HCOH + 95 kcal
The overall conversion of the peroxycarbonyl to formic acid is exothermic by
at least 140 kcal/mole. Because each step can proceed rapidly relative to
collision at one atmosphere, the formic acid can be formed with most of this
excess energy. The molecules of hot formic acid have been suggested to de-
activate and decompose in several ways (Herron and Huie, 1977; Golden, 1978):
(46)
(47)
(48)
(49)
50
-------
H
C0
(50)
HOH + CO
(51)
Determination of the products of the reaction of 03 + CH2=CHa under
atmospheric conditions is complicated by the formation of radicals (H02« and
OH), which react with 03, CH2=<:H2 and the products, especially formaldehyde.
Thus, an accurate quantitative description of the products under atmospheric
conditions even in this simple case, has not been obtained. Less data are
available for more complicated olefins and only a qualitative evaluation can
be made. Table 15 reviews some examples of products formed under atmospheric
conditions.
TABLE 15. PRODUCTS FROM REACTION OF OZONE AND ALKENES
UNDER ATMOSPHERIC CONDITIONS
Alkene
ethene
propene
cis-2-butene
t rans- 2-b u t ene
cis-or trans-l,2-dichloroethene
Products
0
CH20, CO, C02, HCOH
CH3CH, CH20, CO, C02, CH2CO
CH3OH, CH<,
n 0
CH3CH, C02, CH<,, CH20, HCOH
/I
CH3CH, CO, CH20, CH3OH
HCC1, HC1, CO, HCOH, CC120
Reference
Scott et al. (1957)
Hanst et al. (1959)
Niki et al. (1977)
Walters et al. (1977)
Blume et al. (1976)
For predictive purposes the reaction of olefin and ozone may be generalized
as follows:
Ra
li-C-Ra +
R3-C(02»)R<,
(52)
U-C-Rft + Ri-
C(02»)Ra
(53)
51
-------
if R or R. = H
Ra = Ra(Ri)
Rb - R«(Ra)
/-T-\
a(b)
Ra(b)°2'
R ,. ,
a(b)
i
R
R
/u\
a(b)
R ,. ,C
a(b)
a(b)
IOH (54)
H« + COa (55)
C0a
see Section 5.
(56)
(57)
In some cases the epoxide of the corresponding olefin has been reported
in small yields along with decomposition of the epoxide. Therefore the fol-
lowing reactions may also be important:
03
Ri
Ri
^^^fc •
^1 f^^St
Ra^
tl
An
11
^R3
HaO
surface
Ra - H
surface
, R^-C-C-R.
RaR<»
OIIOH
RaR<«
9 ?a(H
R*
(58)
(59)
(60)
This reaction pathway generally does not account for more than 10% of the re-
action olefin with ozone. Some workers have reported much higher conversion
to these products (Vrbaski and Cvetanovic, I960), but these results may be
caused by the secondary reactions during analysis.
52
-------
SECTION 7
PHOTOCHEMICAL REACTION PRODUCTS
GENERAL
The predictive model for estimating products of tropospheric photochemical
reactions represents a compromise between accuracy, complexity, and general
applicability. The model encompasses photochemical reactions that are well-
characterized under conditions similar to those in the lower atmosphere as well
as some reactions that are less completely understood. Certain types of re-
actions readily excluded from the scheme are those that cannot occur in the
absence of solvent or in the presence of oxygen, reactions that require bi-
molecular interactions with species other than oxygen, and reactions that require
more energy than is available from light in the 290 to 800 nm region of the
spectrum. Many important reactions are also excluded from consideration because
existing literature data are insufficient to permit reaction mechanisms to be
confidently postulated.
These limitations necessarily restrict the number of novel compounds to
which the scheme can be applied. However, the compounds for which little
literature data exist are frequently those that do not absorb light or are
generally known to be photochemically inert in the 290 to 800 nm region. More-
over, the chromophores that can be treated by the present model represent
potentially large numbers of new pollutants. Thus, the scheme can be applied
to enough molecular types to be of some practical value and also to those types
of molecules likely to undergo facile photochemical transformations in the lower
atmosphere.
In the following sections the limitations and applicability of the pre-
dictive model are described more fully, the general types of reactions that
are Included in the scheme are discussed, and the specific approach to be used
in estimating photochemical products is described in detail.
LIMITATIONS
Although spectral absorption curves cannot be accurately predicted, some
general observations can be made regarding chromophores that either do or do
not absorb in the 290 to 800 nm spectral region. Clearly, the predictive model
only applies to those compounds that absorb strongly in the solar region because re-
actions with OH or 03 will generally predominate over photolysis for weak
absorbers. Table 16 lists some approximate absorption regions for various
molecular types. Because absorption curves typically do not extend to wave-
lengths as high as 290 nm for aliphatics, nonconjugated olefins, alcohols,
ethers, halides (except iodides), amines, disulfides, and carboxylates (acids,
esters, and amides), these types of molecules can be disregarded.
53
-------
TABLE 16. APPROXIMATE ABSORPTION REGIONS FOR
ORGANIC MOLECULES
Category
Aliphatics
hydrocarbons
fluorides, chlorides
bromides
iodides
ethers, alcohols
aldehydes
ke tones
acids
esters
amides
amines
azines
azo
nitro
nitroso
nitrite
nitrate
sulfide
disulfide
Non-con j ugated
Olefinics and Acetylenes
hydrocarbons
aldehydes
ke tones
nitro
Example
n— CfcH 10
CHBr3, CHC13
C3H7I
C2H3OH, C2H9OC2H3
CHaCHO
C2H3C(0)CH3
CH3C(0)0
CH3C(0)OCH3
CaH5C(0)NHta
(C2H3)3N
(CH3)2C=N-N=C(CH9)2
CH3-N=N-CH3
CH3N02
(CH3)3CNO
(CH3)3CONO
C2H3ON03
CaH 5802^3
CHaSSCHs
(CH3)2C=C(CH3)2
CH2=CHCHO
CH3CH=CH-C(0)CH3
€H3C(N02)=C(CH3)2
Region
nm
120-170
180-260
180-320
150-200
240-340
240-320
200-230
200-240
180-220
180-240
200-290
200-230, 320-400
200-220, 260-320
200-300, 540-740
200-240, 320-400
200-300
200-220
200-240
170-240
200-250, 290-400
200-240, 260-330
180-380
Reference
a
a, b
b
a
a.
a, b
a, b
a
a, b
a, b
b
a
a, b
a
a, b
a
a, b
a, b
a, b
b
b
b
aCalvert and Pitts (1977).
bPerkampus (1971).
54
-------
On the other hand, aldehydes, ketones and functional groups that feature
a double bond to nitrogen (such as nitro and azo compounds, nitrites and
nitrates) do absorb in the solar region and must be considered. Note that the
table lists only aliphatic and nonconjugated olefins. Compounds that feature
extended TT systems (such as aromatics, conjugated olefins, and unsaturated
heterocycles) normally absorb in the 290 to 800 run region; these types of
molecules (with and without substituent groups) are not excluded on the basis
of molecular spectra.
A significant amount of data pertains to the photochemistry of functional
groups that do not absorb in the solar region. This data has necessarily been
obtained using wavelengths < 290 nm and is not directly relevant to the present
case. A question that arises is how to treat those functional groups that
normally do not absorb in the solar region but that can absorb when incorporated
adjacent to or into a IT system. Is the photochemistry of an aliphatic carboxylic
acid ester the same as that of an aromatic ester? In some instances, the energetics
of the processes will prohibit photochemical reactions with wavelengths > 290 nm,
but reasonable mechanisms cannot be excluded. To extend the applicability of
the model to as many molecular types as possible, we include in the model certain
photochemical reactions that have been studied using only wavelengths < 290 nm.
In these instances, the estimated products are classified as possible rather than
probable for cases where the photochemistry has been investigated under more
relevant conditions.
Because we are required to consider the photochemical reactions that occur
under tropospheric conditions, the processes that can be treated by the present
scheme are limited. Certainly it is unlikely that photosolvolysis, photoion-
ization, or electron-transfer reactions will occur in the gas phase. Similarly,
concentrations of reactive atmospheric species (except oxygen) are low, and bi-
molecular photochemical reactions such as intermolecular atom abstractions,
photoadditions, and photosensitized reactions, are improbable. The effect of
oxygen on photofragmentation reactions will be large because the carbon radical
fragments from the photolytic reactions will be scavenged by oxygen to form
peroxy radicals. This is an important consideration since products derived
from peroxy radicals are predictable, using the approach given in Section 5.
Another limiting factor is the energy available in solar radiation. The
lower wavelength limit of 290 nm represents a maximum of 98.6 kcal/mole, and
any reaction with an activation energy greater than this amount cannot proceed
under tropospheric conditions. Table 17 (Benson, 1976) lists bond dissociation
energies for some typical organic molecules. These values can be used to deter-
mine whether or not a photofragmentation reaction can proceed using solar radiation.
For example, most C-F bond homolyses require > 105 kcal/mole and could not occur
with 290 nm irradiation. Similarly, cleavage of bonds to substituents (other
than halogens) on aromatic rings is unlikely. When considering the photochemistry
of aromatics then, we should include reaction pathways other than scission of
bonds involving aromatic carbons.
The problem of multiple reaction pathways must also be addressed. As will
be seen in the following section, several photochemical reaction mechanisms
can be postulated for some chromophores. Although the relative importance of
various pathways can be assessed for specific cases, such assessment is not
55
-------
TABLE 17. BOND-DISSOCIATION ENERGIES FOR SOME ORGANIC MOLECULES R" - R*
in
(34±l)kCH, *
(26±DC2Hs
(21±l)n-C,H7
(18±l)i-CjH7
(8.0±l)r-Bu
(78.5±DC6H5
(45±DC6H5CH2
(40±l)allyl
(-5)CHjCO
(-4.5±1)CH3CH2O
(-LSJCHsCHjOj
(67.5±2)CHj-CH
(52.1)
H
104
98
98
94.5
92
110.5
85
87
87
104
90
108
(19.8)
F
109
108
108
105
—
125
—
—
118
—
—
—
(28.9)
a
83.5
81.5
81.5
81
80
—
—
—
82.5
—
—
91
(26.7)
Br
70
68
68
68
64
78
—
—
—
—
—
—
(25.5)
I
56
53.5
53.5
53
51
64
40
43.5'
51.6
—
—
60
(9.5)
OH
91.5
91.5
91.5
92
92
112
77
80
108
43
(28)e
—
(46±1)
NH2
85
84
84C
84
84
105
—
—
101
—
—
—
(4±1)
OCH3
82
82
82
82
81
100
70
(68)c
97
38
21
95
(34±1)
CH3
88
85
85
84
81
100
72
74.5
81
81
72
99
(26±1)
C2H,
85
82
82
80
78
97
69
71.5
78
81
72
96
(18±1)
i-C3H7
84
80
80
77.5
74
95
67.5
69.5
80
82
(71)
94
(8.0±1)
r-Bu
81
78
78
74
70
92
65
66
—
82
(69)
92
(78.5 ±1)
C«H5
100
97
97
95
92
116
77
78
95
99
91
112
• All values are in kcal/mole.
* Values in parentheses near radicals and atom are AH?300.
1 Values in parentheses are estimates by the author.
-------
generally possible because of the dependence of product distributions on re-
action conditions, such as wavelength, reactant concentration, and solvent
versus gas phase. Also, from an environmental standpoint, it is better to
predict too many products because products featuring high biological activity
(for example, carclnogenlcity, and phototoxicity) can have important ecological
ramifications if formed even in trace amounts. Given these considerations,
no provision has been made for estimating product ratios, and multiple reaction
mechanisms must be considered to be equally probable.
As a final note on limitations, we should consider the accuracy and reli-
ability of the model. Many of the reactions described here have been studied
in solution, in the absence of oxygen, or using wavelengths < 290 nm; in many
cases the assumption that demonstrated reaction mechanisms would be operative
in the atmosphere represents an extrapolation of the facts. Also there is no
guarantee that an individual photochemical reaction has not been inadvertently
excluded from the model. The user should bear these points in mind when assessing
the reliability of the method. Only the most straightforward transformations
should be regarded as highly probable. Other reactions and mechanisms should be
regarded as indicative rather than conclusive. The purpose of the model is not
to provide absolute judgements about photochemical reactions but rather to identify
and recommend for further study those new compounds that could have a potentially
adverse effect on the environment.
PHOTOCHEMICAL REACTIONS
Carbonyl Compounds
Photochemical reactions of aldehydes and ketones are well-characterized
and many pertinent examples are known (Calvert and Pitts, 1966; Coyle and Carless,
1972; Turro, 1978). To predict products of photolysis of carbonyl compounds,
we need to identify structural features that either permit or disallow one or
more of the various reaction mechanisms that can occur.
The first important photochemical reaction to consider is photofragementa-
tion [reactions (61a) and (61b)] to produce free radicals.
RiC(0)R2 • R,C(0)» + R2« (61a)
-------
Similarly, for aliphatic aldehydes, reaction (63a) predominate* over re-
action (63b)
R1C(0)H + hv
a
^•MB
b,
R
HC(0)
(63a)
(63b)
and direct production of acyl radicals is unimportant (Borkowski and Ausloos,
1962). For aromatic aldehydes process (63b) is favored over (63a) on the basis
of the bond dissociation energies involved (Berger et al., 1973), but experi-
mental verification of this is lacking.
In the special case of cyclopentanones and cyclohexanones, photofragmenta-
tion yields a diradical, 1, that is prone to Intramolecular disproportionation
via a 5- or 6-center cyclic transition state, yielding an unsaturated aldehyde
[reaction (64a)] or a ketene [reaction (64b)j (Coyle, 1971; Srinivasan, 1963).
+ hv
•i
o
a
U f* ^mf* * ^P
fiat*-—if ^
u
CH,-» fcH (64a)
c-oo
(64b)
Other cyclic ketones are less likely to undergo intramolecular dispropor-
tionation because entropic and enthalpic factors retard the rate for this process
to the point where combination of the diradical fragments with molecular oxygen
[reaction (52)] is the predominant process.
•CHa(CHa)nC(0)» + 20a
•OOCHa(CHa)C(0)00«
2
(65)
Of course, for acyclic ketones no Intramolecular disproportionation is available
to the radicals formed by reactions (61a and b), and these species will necessarily
be scavenged by molecular oxygen to form peroxy radicals [reaction (66)].
58
-------
RtC(0)» + RaC(0)« +
RiC(0)00» + R8C(0)00»
Ra
(66)
Except for the cyclopentanones and cylcohexanones, photofragmentation of aliphatic
ketones will lead to formation of peroxy radicals. The ultimate atmospheric
products to be derived from the peroxy radicals will be those predicted by the
model described in Section 5 above. Thus, to predict the atmospheric products
of the photofragmentation reactions it is necessary to identify only the possible
species RjOO" and R.C(0)00« and then refer to the scheme in Section 5 to analyze
estimated products.-1 The scheme in Sections cannot precisely accommodate re-
actions of diperoxy radicals, such as _2, and for this reason, products of photo-
fragmentation of cyclic ketones, other than cyclohexanones and cyclopentanones,
cannot be estimated beyond the generalization that multifunctional oxygenated
species will probably be formed.
An important, and distinctive pathway is available to acyclic aldehydes
and ketones that feature a hydrogen atom in an aliphatic carbon y to the carbonyl
group. For these types (Wagner, 1971) an intramolecular hydrogen atom abstraction
via a 6-center cyclic transition state yields a diradical, as in reaction (67).
RiCHCHaCHa
(67)
Generally, two reaction pathways are available to 3: (1) elimination to
form an olefin and an enol (which typically tautomerizes to the carbonyl com-
pound) , and (2) reaction (68a) or cycllzatlon to form a cyclobutanol, as in
reaction (68b).
(68a)
RxCH-CHa + CH,C(0)Ra(from CHa-CRa)
OH
Ra -
(68b)
Ri
H
59
-------
Usually, both reactions (68a) or (68b) cannot occur unless electronic, steric, or
ring-strain considerations favor one pathway. Reaction (68b) is especially
favorable for a-diketones, (Urry and Trecker, 1962) Urry et al., 1962), pre-
sumably because the biradical is resonance stabilized, as in reaction (69)
Ri
CCCH2C-R2 + hv-»-
H9fi
R!C-0
CH2CR2
HO 9-
&aC-CC
H
OH
Ri-C-
0
I
(69)
Another aspect of photochemical reactions of carbonyl compounds concerns
intramolecular skeletal rearrangements of cyclohexanones and cyclohexadienones,
as in reactions (70) and (71) (Kropp, 1967).
+ hv
R R
(70)
R R
Reactions of carboxylates (acids, anhydrides, esters, and amides) are
considerably less certain than those of ketones and aldehydes. Most work has
been done with simple aliphatic species that only absorb at short wavelengths.
For wavelengths less than 290 nm, sufficient energy is available to produce
fragmentation of most bonds in the aliphatic carboxylates. This will not
necessarily be the case at higher wavelengths, and to assess the probability
of various reactions occuring we must estimate heats of reaction of alternative
pathways. Such an estimation can be performed using established methods and
available heats for formation of various radical species (Benson, 1976). For
example, in the case of acetic acid, the following processes occur at wave-
lengths < 210 nm (Ausloos and Steacie, 1955):
CHaC(0)OH
CH9» + «C(0)OH
CH3C(0)« + OH
CH3C(0)0- + H
(72a)
(72b)
(72c)
60
-------
To determine the likelihood of reactions (72a) through (72c) occurring
with wavelengths > 290 nm (< 98.6 kcal/mole), we need to calculate heats of
reaction for each process. Only those processes with AIL < 98.6 can occur.
Using the available data, we estimate the heats of reaction for reactions (72a),
(72b), and (72c) to be +88, +109, and +108 kcal/mole, respectively. Thus, only
reaction (72a) is likely to occur. For esters, exemplified by CH3C(0)OCH3, the
possible processes are reactions (73a) through (73b).
CH3C(0)OCH3 » CH3« + «C(0)OCH3 (73a)
^ CH3C(0)« + -OCH3 (73b)
* CH3C(0)0» + CH3- (73c)
The estimated heats of reaction are +82, +85, and +96 kcal/mole, for reactions
(73a), (73b), and (73c), and all must be considered possible. Anhydrides,
exemplified by CH3C(0)OC(0)CH3, primarily undergo dissociation as in reaction
(74) (Ausloos, 1956):
CH9C(0)OC(0)CH3 * CH3C(0)0« + CH3C(0)« (74)
The estimated heat of reaction (74) is +72 kcal/mole; therefore proceeds the
reaction at wavelengths > 290 nm. For amides, such as CH3C(0)NH2, reaction
(75) is dominant at lower wavelengths (Booth and Norrish, 1952).
CH3C(0)NH2 • CH3- + •CWNHz (75)
Because the estimated heat of this reaction is +83 kcal/mole, it is expected
to occur at longer wavelengths. If aromatic moieties are present in the carbo-
xylates, it is unlikely that photofragmentation will produce phenyl (or substi-
tuted phenyl) radicals at longer wavelengths because of the high bond strengths
of phenyl-carbonyl bonds.
Possible photofragmentation reactions of carboxylates that absorb at wave-
lengths > 290 are summarized in the next section. As before, carbon radical
fragments will be scavenged by molecular oxygen to produce peroxy radicals,
which will give products according to the scheme presented in section 5.
A final category of reactions involving carboxylates is the intramolecular
hydrogen atom abstraction and elimination analogous to reactions (67) and (68a)
for aldehydes and ketones. For amides with an aliphatic hydrogen y to the
carbonyl, olefins and a shorter chain amide (from the enol tautomer) are pro-
duced, as in reaction (76) (Booth and Norrish, 1952).
HaNCCHaCHatR —^ H2NCCH3[HaNC=CH2] + RCH=CH2 (76)
For esters with y-aliphatic hydrogens in the acid moiety or B-hydrogens in the
alcoholic moiety, reactions (77a) and 77b) can occur (Ausloos, 1958).
61
-------
f 9 ?
RiC-CH2OCCH2CH2CR2
R1CH=CH2 + HOCCH2CH2R2
RiCH2CH2OCCH3 + R2CH=CHS
(77a)
(77b)
The analogous reaction for anhydrides with a 0-hydrogen produces an acid and
a ketene [reaction (77)] (Ausloos, 1956).
?9 8?
RiC-C-O-C-C-f
RiCH=C=0 + R2CH2C02H
(78a)
RiCH2C02H 4- R2CH=C=0
(78b)
Analogous reactions for carboxylic acids can be postulated but have not been
demonstrated.
Olefins
Conjugated olefins undergo photofragmentation when irradiated with short-
wavelength uv light and at low pressures in the gas phase. In the troposphere
with irradiations > 290 nm, such reactions are improbable because of the high
bond dissociation energies involved. The atmospheric reactions of conjugated
olefins will thus be largely confined to molecular rearrangements such as ring-
opening, ring formation sigmatropic rearrangements, and cis-trans, trans-cis
isomerizations (Woodward and Hoffman, 1970; Turro, 1978; Coyle, 1974).
Ring closure of conjugated dienes, trienes and cis-stillbenes is repre-
sented by reactions (79) through (81).
R
R "
(79)
(80)
(81)
62
-------
^Ring-closure in 7- and 8-membered ring polyenes provides bicyclic olefins,
as shown in reactions (82) through (86).
(82)
(83)
(84)
(85)
(86)
The ring-opening of cyclic olefins is simply the reverse of the corresponding
ring-closures, that is
+ hv
(-72)
Nitrites, Nitrates, Nitro and Nitroso Compounds
This category of compounds is important because they normally feature strong
absorbance in the visible range and because they are quite photolabile. The
nitro group is particularly effective in causing bathochromic shifts in aromatic
compounds.
63
-------
The primary reaction of aliphatic nitroso compounds is photofragmentation,
as in reaction (87).
R-N=0 + hv »- R» + NO (87)
As for photofragmentation of carbonyl compounds, the radicals formed initially
are rapidly scavenged by molecular oxygen, and the atmospheric products are
derived from the corresponding peroxy radicals. For compounds with hydrogen
atoms on the carbon 3 to the nitroso group, an intramolecular elimination
[reaction (88)] is also considered possible (Calvert and Pitts, 1966).
RiO
:CCH2N=0 + hv • R!CH=CH2 + HNO (88)
For analogous situation exists for aliphatic nitro compounds, where photo-
fragmentation [reaction (89)] yields carbon radicals and N02 (Paszyc, 1974;
Glasson, 1975) and cyclic elimination [reaction (90)] produces nitrous acid
and an olefin from nitro compounds with hydrogens 3 to the N02 group (Gray et
al., 1955).
RN02 • R» + N02 (89)
RiCHCHaNOa »- R1CH=CH2 + HONO (90)
For aromatic nitro compounds, a different mechanism prevails. In nitro-
benzene, photofragmentation [reaction (91)] yields involve N-0 homolysis rather
than C-N cleavage (Hastings and Matsen, 1948).
C6H3N02 + hv —*- C6H3N=0 + 0 (91)
This work was performed using wavelengths < 290 nm, but the estimated heat of
reaction (+94 kcal/mole) is low enough so that the process should be considered
possible at < 304 nm which is just in the solar region. The actual computed
photolysis rate constant at 30° N latitude (Section 4) for nitrobenzene using
the absorption spectra reported by Calvert and Pitts (1966) and assuming a
quantum yield of 1.0 below 305 nm and 0.0 above is 7.6 days"1 at the summer
solstice and 0.7 day"1 at the winter solstice. Thus the lifetime of nitro-
benzene in the troposphere will be less than one day except near the winter
solstice where it will be slightly greater. This example illustrates how
effective a narrow region of the solar spectrum even at the solar cut off can
be in causing reaction, if the compound absorbs strongly in this in region
such as nitrobenzene does and if the quantum yield is favorable.
Aliphatic nitrites undergo a facile photofragmentation into alkoxy radicals
[reaction (92)] (McMillan et al., 1969; Wiebe and Heicklen, 1973).
RON=0 + hv »• RO- + NO (92)
The fate (and products) of the alkoxy radicals in the atmosphere would be as
described in Section 5.
The analogous reaction of aliphatic nitrates is less well understood.
64
-------
Fragmentation into alkoxy radicals [reaction (93a)] seems likely, but N-0
cleavage [reaction (93b)] and cyclic elimination of nitrous acid [reaction (93c) ]
are also possible (Gray and Style, 1953; Gray and Rogers, 1954; Rebbert, 1963).
hv - *~ RCH20« + N02 (93a)
- ^ RCHaONO + 0 (93b)
• RCHO + HONO (93c)
Halides
Most photochemical studies on organic halides have used wavelengths < 290 nm.
Yet, Table 7 shows that carbon halogen bond strengths (except for fluorides)
are typically less than the 98.6 kcal/mole maximum energy available in the solar
spectrum. This is true even for aromatic and olefinic halides. Therefore, photo-
chemical transformations of organic halides that absorb in the visible region of
the spectrum should be considered. The only important photochemical reaction of
organic halides is fragmentation into radicals [reaction (94)].
R-X + hv • R- + X (94)
When more than one type of halogen substituent is present, cleavage involves
only the weakest bond (that is, I > Br > Cl > F in order of preferential cleavage)
A good example of this type of reaction is photolysis of chloroacetone (Strachen
and Blacet, 1955). When 313.0 nm irradiation is used, the principle reaction
is C-C1 bond cleavage [reaction 95)]:
ClCHzC(0)CH3 • Cl» + •CH2C(0)CH3 (95)
If the halogen substituent is remote from the ch.romoph.ore, however, the halogen
bond is not broken. Thus, for 4-chloro-2-butanone, the reactions (96a) and (96b)
proceed to the exclusion of reaction (96c) (Taylor and Blacet, 1956).
ClCHaCHaC(0)CH3 + hv - ^ ClCH2CHa- + -C(0)CH3 (96a)
C1CH2CH2C(0)« + »CH3 (96b)
CHaCHaC(0)CH3 (96c)
This example is a good illustration of the point that the photolabile group
must be in proximity to the chromophore to permit a reaction to proceed.
Aromatic halides (except fluorides) also apparently undergo C-X cleavage
to yield halogen atoms or phenyl radical [reaction (97)] (Shoma and Kharasch,
1968; Hee and Sutherland, 1979).
Ar-X - ^ Ar- + X- (97)
PROCEDURE
Estimating atmospheric products of photochemical reactions is facilitated
65
-------
by referring to Table 18. The table summarizes reactions for the various
chromophores and functional groups described in the previous section. The
general procedure for using the table is as follows. First, identify all chromo-
phores or functional groups in a candidate molecule. Photochemical reactions
for each chromophore or functional group listed in the table must be considered.
For chromophores or functional groups not listed, no products can be predicted.
For a given functional group or chromophore, more than one photochemical re-
action may be possible (e.g., for aliphatic ketones, fragmentation and cyclic
elimination may be possible). In these cases, all possible reactions are cited.
The photochemical products for the reactions to be considered are given in the
table. If molecular products are formed directly from the photochemical reaction,
they are the atmospheric products. If carbon radicals are formed from the initial
photochemical act, combination with molecular oxygen to form peroxy radicals is
assumed, and estimation of the atmospheric products is as described in Section 5.
When oxy radicals are formed, the atmospheric products are also predicted using
the procedure in Section 5.
The photochemical reactions in Table 18 are identified as either probable
or possible. The possible reactions are postulated on the basis of work per-
formed on compounds that do not absorb in the solar region.
If a particular functional group is in proximity to or incorporated into
a chromophore that absorbs in the 290 to 800 run, a photochemical reaction
involving the functional group may occur. If so, the most probable reaction
pathways are those listed in the table. It is important to note that if a
functional group is remote from an absorbing chromophore, it is like that photo-
chemical transformation will not occur (see example on the previous page for
chloro-substituted ketones).
Care must be taken in using the table when situations involving unusual
geometric configurations arise. Steric or ring strain in potential products
or ring systems (e.g., aromatic ring substituents located ortho to one another)
that place reactive functionalities in close proximity are potentially important
effects that cannot be treated in a general fashion. Such occurrences must be
considered on a case by case basis, and the user must rely on previous experience
in assessing the impact of such effects on product distributions.
As a final work of precaution in applying Table 18, it should be clear
that the presence of a structure in the table does not mean the structure
photolyzes in the solar spectrum. The table should be only applied to compounds
that absorb sufficiently for photolysis to be important. In those reactions
which are indicated as "possible," additional chromophores are required to
make it possible for these structures to absorb solar light inorder to react.
66
-------
TABLE 18. PRODUCTS OP PHOTOCHEMICAL REACTIONS
Functional Croup
L
Ketone: RjCRa
fl
A,
U.
I 8
RiCHCH.CH.CRa
u n
¥ V
RiCCHCHaCRa
0
r*i
^j
X
6
R R
Q
Udehyde: RiCH
*L 8.
R.CHCHaCHiCH
Car boxy lie acid:
R,C(0)OH
Carboxyllc acid
ester:
R,C(0)OR,
"L 8
RiCHCHiCHiCORi
8 L_
R,COCHaCHR,
^rboxyllc acid
inhydride:
R,C(0)OC(0)Ra
t 2
R,CHC(0)OCR,
Type
RiR. - alkyl
R, • alkyl,
Ra " aryl
Ri, HI - Aryl
Cyclic: RiR, -
-CH,(CHi) CHj-;
n - 2,3 "
n«,3
o-Hydrogen on
saturated
carbon
o-Hydrogen on
carbon
Cyclohexa-
dlenone
Cyclohexenone
R, - alkyl
Ri • aryl
a-Hydrogen on
saturated
carbon
R, - alkyl
R..RJ - alkyl
R, - alkyl,
R, - aryl
Ri - aryl.
R, - alkyl
R, • aryl,
R, • aryl
Y "Hydrogen on
saturated
carbon In
acid moiety
8-Hydrogen on
saturated car-
bon In alcohol
moiety
R»,R» • Aryl
or alkyl
B-Hydrogcn on
saturated
carbon
Reaction
Fragmentation
Fragmentation
Fragmentation
Fragmentation
Fragmentation
Cyclic elimina-
tion
a-Hydrogen
transfer
Rearrangement
Rearrangement
Fragmentation
Fragmentation
Cyclic elimina-
tion
Fragmentation
Fragmentation
Fragmentation
Fragmentation
Fragmentation
Cyclic elimina-
tion
Cyclic elimina-
tion
Fragmentation
Cyclic •llmina
tion
Likelihood
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Possible
Probable
Possible
Feasible
Possible
Possible
Possible
Possible
Possible
Possible
Possible
67
Initial
Photochemical
Products
RlC(O)', R,', KI',
RaC(O)'
R,', R,C(0)>
No reaction
•CH,(CHa)nCHaC(0)«
•CH,(CH,)nCHaC(0)'
a
[enol]
R1CHCH,CH,C(OH)R1
i?
f/L
^\£T~*
R
cX
R
R,.(0), HC(0)
RiC(O), H-
p
R.CH-CH,. CHaCH [enol]
R,., *C(0)OH
R.CW)", RaO«, Ri',
•C(0)ORa, R,C(0)0«, Ka'
R,C(0)«, RaO', Ri',
•C(0)ORa
R,C(0), RaO, R,C(0),
Ra'
R,C(0), R,0
R,CH-CHa, CHaC(0)ORa
[enol)
R,C(0)OH -I- CHa-CRRa
R,C(0). RaC(O), R,C(0)
RaC(0)0
R.CH-C-0. R,C(0)OH
Atmoipherlc
Products
b
b
CHa-CH(CHa) .COaH
CH.CCH.j^jCH'C-O
Uncertain0
As shown
OH
!Ra
— H
.
As shown
As shown
b
b
s shown
b
b
b
b
b
As shown
As shown
b
As shown
-------
TABLE 18 (concluded)
!arboxylic acid
unlde:
R,C(0)HH,
\
,
)lefln:
R
CH,-CH-|-CH.CH,
R
CH,-CH-CH>
CH-CH-CHi
R
C.Hj-CHliCH-C.H,
R-C.H,,
R-C.H,
R-C.H,
R-C,H,
R-C,H,
R-C.H,
R-C..H,
Vitroao com-
pounds:
R-NO
RCllCHjN-0
litro compounds:
RNO,
u
RCHCH,NO,
Nitrites:
RONO
Nitrate*:
RONO,
RCH.ONO,
ialides: (except
fluorides)
R-X
R, - alkyl
Ri - aryl
Y-Hydrogen on
saturated
carbon
Conjugated
dlenea
Conjugated
trienes
Cia-Stllbenes
1,3-Cycloocta-
dlenea
1,3.5-Cyclo-
octatrlenea
Cyclooctata-
traenea
1, 3-Cyclohepta-
dlene
1,3,5-Cyclo-
beptatrlene
1,3-Cyclohexa-
dienes
Cyclobutenes
Alkyl
B-Hydrogen on
aliphatic
carbon
Allcyl
Aryl
B-Hydrogen on
aliphatic
carbon
Alkyl
Allcyl
Alkyl
a-Hydrogen on
aliphatic
carbon
alkyl, aryl
Fragmentation
Fragmentation
Cyclic elimina-
tion
Ring cloaure
Ring cloaure
Ring cloaure
Ring cloaure
Ring cloaure
Ring cloaure
Ring cloaure
Ring cloaure
Ring closure
Ring closure
Fragmentation
Cyclic elimina-
tion
Fragmentation
Fragmentation
Cyclic elimina-
tion
Fragmentation
Fragmentation
Fragmentation
Cyclic elimina-
tion
Fragmentation
Possible
Possible
Possible
Probsble
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Probable
Possible
Probable
Feasible
Probable
Probable
Possible
Possible
Possible
Possible
R,., .cwnra,
No reaction
R.CH-CH,, CH,C(0)NR,
[enol]
Cyclobutenea
1 , 3-Cyclohexadlenea
Dihydrophenanthrenea
Bicyclo[4.2.0]oct-7-ene
Bicyclo[4.2.0]octa-
2,7-diene
Blcyclo[4.2.0)octa-
2,4,7-triene
Bicyclo[3.2.0)hept-
S-enes
Bicyclo [ 3 . 2 . 0 ) hepts-
2,5-dlenes
1,3, S-Hexatrlenea
1,3-Butadlenes
R> + NO
RCH-CH, -I- HNO
R> + NO,
RN-0 + 0
RCH-CH, + HONO
R0> + NO
R0> + NO,
RONO + -0
RCHO + HONO
R. + X»
b
As shown
As shown
As shown
As shown
As shown
As shown
As shown
As shown
As shown
As shown
As shown
b
As shown
b
As shown
As shown
b
b
As shown
As shown
b
The word "Probsble" under the heading Likelihood Indicates that photolysis in the solar has been demostrated,
although not necessarily In the presence of oxygen and In the gaa phase; the word "Possible" Indicates that
photolysis has been demonstrated only below 290 nm but the energetics would be favorable for reaction in the
solar region if structure Is associated with an adjacent chromophore that permitted absorption in the solar
region.
Derived from corresponding peroxy radicals; see Section 5.
Multifunctional oxygenated apecies derived from >OOCH,(CH,) CH,C(0)00-.
n
68
-------
SECTION 8
PREDICTIVE SCHEME
The following is an outline of the predictive procedure.
STEP 1: ESTIMATION OF RATES OF ENVIRONMENTAL PROCESSES
To determine the relative contribution of the various atmospheric chemical
processes to the formation of products, we must first determine the lifetimes
of the processes. The rate of loss (R ) of compound C is
\j
Rc = -dC/at = [ROH[OH] + k03[03] + kp} [C]
or expressed in lifetimes, the rate of loss is
The contribution of each process is proportional to the reciprocal lifetimes.
However, l/t_ is only an upper limit, and therefore the importance of the photo-
reaction can be much less, depending on the quantum yield.
To limit the amount of effort used in estimating products from reactions
that will be unimportant relative to the other processes, we suggest the following:
• If I/T--. > 100/Trt , then products from ozone reactions may be
On Oa
ignored; conversely, if 1/T_ > 100/Tnu, then the OH reactions
t»3 Utl
may be ignored. If neither of these inequalities apply, then
both processes must be considered.
• If 1/T..-. > 100/Trt > 100/T , then the photo products need not
OH Oa p
be determined. However, if the inequalities do not apply, the
photo products should be determined.
The procedures for estimating I/T for each process are given below.
Estimation of the Rate Constant of Reaction with OH (kQII) and T QH
The rate constant for the reaction with OH may be calculated from the
expression
i j *
kOH
69
-------
The first summation accounts for reaction of all reactive hydrogen atoms
within the molecule. The term kg^ is the reactivity of the hydrogen atoms from
Table 4, taking into account the position and degree of substitution on the
adjacent atom by alkyl, vinyl, phenyl, C(0)-, 0-, S-, and N- groups. The terms
am and PJJ^ account for the effect of groups in the cr-position and (J-posltion
on the carbon-hydrogen bond and are also defined in Table 4. These terms are
applied for each o and 0 substituent; thus,
The term n accounts for the repetition of the i carbon-hydrogen bond.
The second summation accounts for the reactivity of all carbon-carbon
double bonds in the molecule. The term kE is the intrinsic reactivity of each
double bond, taking into account the degree of substitution, whereas og accounts
for the effect of any halogen substituted directly to the double bond. These
terms are defined in Table 5, n. accounts for the repetition of each unique
double bond.
The third summation accounts for addition of OH to aromatic rings in the
molecule. The intrinsic reactivity of the ring, k^, accounts for the degree
of alkyl substitution. The OTA accounts for the halogen substituents in the
ring. Both kA and OA are defined in Table 6. In addition, k^ may be estimated
from Figure 1 using valves of Q_ from Table 14.
The sum of the three terms determined ICQJJ (in cm3 molec ) for the com-
pound. Examples of this calculation are given on page 21.
The lifetime for reaction with OH is
TOH = 1/kOH[OH]
Assuming an average environmental concentration of OH = 10a molec/cm3 (1.7 x 10 or
4.1 x 10~8 ppm) the halflife is
T (seconds) = 1.0 x 10~'/k'
w£l Oil
Estimation of the Rate Constant for Reaction with Ozone (k_ ) andT.
QS 03
To estimate the rate constant for the reaction with ozone, the structure
of the compound is analyzed for the reactive groups identified in Table 7. Thus,
the method involves determining if the compound has a carbon-carbon double; if
it does have such a bond, the degree of substitution is determined. Then, from
data in Table 7, the rate constant for that specific type of group is determined.
If the compound contains an aromatic ring, the degree of substitution is deter-
mined and the estimated rate constant is determined also from Table 7. If the
molecule contains more than one reactivity group, the molecular reactivity is
the sum of the rate constant for each group.
70
-------
The half-life for reaction with 03 is
Since the average tropospheric ozone concentration is 1 x 1012 molec cm~3
(1.7 x 10~9 M or 0.041 ppm), the half-life may be expressed as follows:
T (seconds) = 1.0 x 10~12/k_
\Ja (}3
Estimation of the Rate Constant for Photolysis (k ) and T
2 P_
From the absorption spectrum of the compound, the average values of the
cross section for the 10 nm intervals used in Tables 8 through 10 are deter-
mined and then combined with the corresponding J* values from these tables.
The upper limit for the photolysis rate constant (k ) is
k = Ect.J'
p XX
This calculation assumes the photo reaction occurs with quantum yield of unity.
The limit for the lifetime is
t = 1/k
P P
The J' values are in units of day, but for comparison of t with the other
values they must be converted to second units:
T (seconds) = 8.64 x 104T (days)
STEP 2. ESTIMATION OF PRODUCTS FROM OH REACTIONS
All C-H bonds, C=C bonds, and aromatic rings were identified for estimation
of the OH rate constant. The products formed from reaction of each of these
groups must be determined.
C-H Bond Reactions
Each C-H bond that reacts with OH is assumed to be converted to an alkoxy
radical as discussed in Section 5.
-£-H + OH —^ -C-0-
The fate of the alkoxy radical determined from Tables 12 and 13 allows one to
estimate (1) the rate of 0-elimination for each group substituted on the central
C atom, (2) the rate of reaction with 02, (3) the rate of reaction with N02,
and (4) the rate of 6-H rearrangement if the reaction is possible. The necessary
data are listed in the appropriate column in the tables corresponding to the
structure of the alkoxy radical.
71
-------
If the reactions generate additional carbon radicals, then the products
from the corresponding alkoxy radicals must also be determined.
C=C Bond Reactions
For each C=C bond in the molecule, the following general reaction may be
assumed:
Ri
, -
R2
•*- R
0 0
iCR? i i\3Vfixit
If one or more of the R groups are hydrogen, it is possible that some hydrogen
abstraction may occur, but the amount will be extremely small (0.1%), and this
route may be ignored.
Aromatic Ring Reactions
For each aromatic ring the following distribution of products is assumed:
OH (minus one R,
general H)
Ring Cleavage Products
+0.5 Ring Cleavage Products
00 00
RiCCR2 + % R2CCR3
09
R
The contribution of hydrogen abstraction from the ring is expected to be small
and the products may be ignored.
Relative Amounts of Products
The predominance of each product will depend on the yield in each reaction
in which it is formed as well as the relative importance of each process. The
relative importance of the various OH reactions may be estimated from the values
of k.., k_, and k. which were determined in estimating
72
-------
STEP 3: ESTIMATION OF PRODUCTS FROM 03 REACTIONS
For each carbon-carbon double bond reacting with ozone, the following
generalization can be made:
R3-C(02«)R<,
R = R3(Ri)
a(b)
if one R
H
a(b)
ff
R f, vCOH
a(b)
,. ,
a(b)
/v\
a(b)
X^K
R
f. v
a(b)
+ C02
see section 5.
H» + C02
Small amounts ('v* 1%) of the following products may be generally expected:
OH OH
2<,
R!-C-CH
(H)
R3
STEP 4. ESTIMATION OF PRODUCTS FROM PHOTOLYSIS
To estimate the possible products from photolysis, we identified the various
chromophores in Table 18. The pathways and products are obtained from the table.
For chromophores or functional groups not listed in the table, no products can
be predicted.
All the processes are assumed to occur with equal facility in the absence
of more detailed data. In cases where carbon radicals or alkoxy radicals are
formed, the products of these intermediates are determined as in Step 2.
73
-------
INDEX
Acids (Carboxylic acids)
OH reactions, 14-21, 30-40
Photolysis reactions, 54, 60, 67
Aldehydes (Alkanols)
OH reactions, 14-21, 30-40
Photolysis reactions, 54, 58, 67
Alkadienes (Dienes, conjugated olefins)
OH reactions, 8, 14-21, 40-42
03 reactions, 10, 49-52
Photolysis reactions, 62, 63, 68
Alkanes
OH reactions, 7, 14-21, 30-40
03 reactions, 10
Photolysis reactions, 54
Alkanols (Alcohols)
OH reactions, 7, 14-21, 30-40
Photolysis reactions, 54
Alkenes (Olefins)
OH reactions, 8, 14-21, 40-42
03 reactions, 10, 11, 49-52
Photolysis reactions, 54, 62, 63, 68
Alkynes (Acetylenes)
OH reactions, 9, 14-21, 40-42
03 reactions, 11, 49-52
Amides
OH reactions, 14-21, 30-40
Photolysis reactions, 54, 61, 68
Anhydrides (Carboxylic anhydrides)
OH reactions, 14-21, 30-40
Photolysis reactions, 54, 61, 67
Aromatic Compounds (Aralkanes)
OH reactions, 4, 14-21, 42-48
09 reactions, 11, 49-52
Halides
OH reactions, 7, 14-21, 30-40
Photolysis reactions, 54, 65, 68
Cycloalkenes (Cyclic olefins)
74
-------
OH reactions, 8, 14-21, 40-42
03 reactions, 10, 11, 49-52
Photolysis reactions, 54, 62, 63, 68
Esters (Carboxylic esters)
OH reactions, 8, 14-21, 30-40
Photolysis reactions, 54, 61, 67
Ketones (Alkanones)
OH reactions, 7, 14-21, 30-40
Photolysis reactions, 54, 57-60, 67
Nitrates
OH reactions, 14-21, 30-40
Photolysis, 54, 63, 64, 68
Nitrites
OH reactions, 14-21, 30-40
Photolysis reactions, 54, 63, 64, 68
Nitro Compounds
OH reactions, 14-21, 30-40
Photolysis reactions, 54, 63, 64, 68
Nitroso Compounds
OH reactions, 14-21, 30-40
Photolysis reactions, 54, 63, 64, 68
Terpenes
OH reactions, 8, 14-21, 40-42
03 reactions, 10, 11, 49-52
75
-------
REFERENCES
Ausloos, P., and E. W. R. Steacie. 1955. The Vapor Phase Photolysis of Acetic
Acid. Can. J. Chetn. 33:1530-1535.
Ausloos, P. 1956. The Gas Phase Photolysis of Acetic and Propionic Anhydride.
Can. J. Chem. 34:1709-1713.
Ausloos, P. 1958. The Photolysis of Alkyl Esters. Can. J. Chem. 36:383-392.
Baldwin, A. C., J. R. Barker, D. M. Golden, and D. G. Hendry. 1977. Photo-
chemical Estimates and Computer Simulation. J. Phys. Chem. 81:2483-2492.
Benson, S. W. 1976. Thermochemleal Kinetics, Second Edition. John Wiley
and Sons, Inc., New York.
Berger, M. T., L. Goldblatt, and C. Steel. 1973. Photochemistry of Benzaldehyde.
J. Amer. Chem. Soc. 45:1717-1725.
Blume, C. W., 1. C. Hisatusune, and J. Heicklen. 1976. Gas-Phase Oznolysis of
cis- and trans-Dichloroethylene. Intern. J. Chem/ Kinet. 8:235-258.
Booth, G. H., and R. G. W. Norrish. 1952. Photochemical Studies of Some Organic
Nitrogen Derivatives. J. Chem. Soc. 188-198.
Borkovski, R. P., and P. Ausloos. 1962. The Vapor Phase Fluorescence and its
Relationship to the Photolysis of Propionaldehyde and the Butyraldehydes.
J. Amer. Chem. Soc. 84:4044-4048.
Bouchy, M. and Andre, J. C. 1977. Photooxldatlon of Blacetyl Kinetic Study of
Formation of Biacetyl Peroxide and Carbon Dioxide. Molec. Photochem.
8:345-357.
Cadle, R. D. and C. Schadt. 1952. Kinetics of the Gas Phase Reaction of Olefins
with Ozone. J. Amer. Chem. Soc. 74:6002-6004.
Cadle, R. D., and Schadt, C. 1953. Kinetics of the Gas Phase Reaction Between
Acetylene and Ozone. J. Chem. Phys. 21:163.
Calvert, J. G., and J. N. Pitts, Jr. 1966. Photochemistry. John Wiley and
Sons, Inc., New York.
Carter, W. P. L., A. C. Lloyd, J. L. Sprung, and J. N. Pitts, Jr. 1979. Com-
puter Modeling of Smog Chamber Data; Progress in Validation of a Detailed
Mechanism for the Photooxidation of Propane and n-Butane in Photochemical
Smog. Intern. J. Chem. Kinet. 11:45-101.
76
-------
Chapman and Shorter. 1972. Advances In Linear Free Energy Relationships.
Plenum Press, New York.
Coyle, J. D., 1971. Product Ratios in the Ring-opening Photoreactions of
Cycloalkanones. J. Chen. Soc. (B):1736-1740.
Coyle, J. D., and H. A. J. Carless. 1972. Photochemistry of Carbonyl Compounds.
Chem. Soc. Rev. 1:329-353.
Coyle, J. D. 1974. The Photochemistry of Olefinic Compounds. Chem. Soc. Rev.
3:329-353.
Darnall, K. R., W. P. L. Carter, H. C. Lloyd, A. M. Winers, and J. N. Pitts
Jr. 1976. Importance of R0a + NO in Alkyl Nitrate Formation from C*-C«
Alkane Photooxidation Under Simulated Atmospheric Conditions. J. Phys.
Chem. 80:1948-1950.
Darnall, K. R., R. Atkinson, and J. N. Pitts, Jr. 1978. Rate Constants for
the Reaction of OH Radical with Selected Alkanes at 300K. J. Phys. Chem.
82:1581-1584.
Demerjlan, K. L., J. A. Kerr, and J. G. Calvert. 1974. The Mechanism of Photo-
chemical Smog Formation. Adv. Environ. Sci. and Tech. 4:1-262.
Glasson, W. A. 1975. Methoxyl Radical Reactions in Atmospheric Chemistry.
Environ. Sci. Tech. 9:1048-53.
Golden, D. A. 1978. Organic Free Radicals. Workshop on Chemical Kinetics
Data Needs for Modeling the Lower Troposphere. Reston, Virginia.
Gray, J. A., and W. G. Style. 1953. The Photolysis of Ethyl Nitrate. Trans.
Faraday Soc. 49:52-57.
Gray, P., A. D. Yoffe, and L. Roeelaar. 1955. Thermal Decomposition of the
Nitroalkanes. Trans. Faraday Soc. 51:1489-1497.
Gray, P., and G. T. Rogers. 1954. The Explosion and Decomposition of Methyl
Nitrate in the Gas Phase. Trans. Faraday Soc. 50:28-36.
Greiner, N. R. 1970. Hydroxyl Radical Kinetics by Kinetic Spectroscopy. VI
Reactions with Alkanes in the Range of 300-500°K. J. Chem. Phys.
53:1070-1076.
Hampson, R. P., Jr. and D. Garvin. 1978. Reaction Rate and Photochemical Data
for Atmospheric Chemistry-1977. National Bureau of Standards Special
Publication 513.
Hansen, D. A., R. Atkinson and J. N. Pitts, Jr. 1977. Structural Effects on
the Chemiluminescence from the Reaction of Ozone with Selected Organic
Compounds. J. Photochem 7:379-404.
Hanst, P. L., E. R. Stephens, W. E. Scott, and R. C. Doerr. 1959. Atmospheric
77
-------
Ozone-Olefin Reactions. Preprints Petrol. Div., Amer. Chem. Soc. 4:A7-A16.
Hastings, S. H., and F. A. Matsen. 1948. The Photodecomposition of Nitrobenzene.
J. Amer. Chem. Soc. 70:3514-3516.
Hee, S.S.Q., and R. G. Sutherland. 1979. Vapor and Liquid Phase Photolysis
of the n-Butyl Ester of 2,4-Dichlorophenoxyacetic Acid. Arch. Environ.
Contain. Toxicol. 8:247-254.
Heicklen, J. 1966. The Reaction of Ozone with Perfluoroolefins. J. Phys. Chem.
70:477-480.
Hendry, D. G., A. C. Baldwin, J. R. Barker, and D. M. Golden. 1978. Computer
Modeling of Simulated Photochemical Smog. EPA-600/3-78-059.
Hendry, D. G. and W. R. Mabey. 1978. Unpublished data.
Hendry et al., D. G. 1979. Test Protocols for Environmental Processes: Air
Chemistry (Draft). EPA Contract 68-03-2227.
Herron, J. T., and R. E. Huie. 1977. Stopped-Flow Studies of the Mechanism
of Ozone-Alkene Reactions in the Gas Phase Ethylene. J. Amer. Chem. Soc.
99:5430-5435.
Japar, S. M. and H. Niki. 1975. Gas-Phase Reactions of the Nitrate Radical
with Olefins. J. Phys. Chem. 79:1629-1632.
Japar, S. M., C. H. Wu, and H. Niki. 1974. Rate Constants for the Reaction
of Ozone with Olefins in the Gas Phase. J. Phys. Chem. 78:2318-2320.
Kenley, R. A., J. E. Davenport, and D. G. Hendry. 1978. Hydroxyl Radical
Reactions in the Gas Phase. Products and Pathways for the Reaction of
OH with Toluene. J. Phys. Chem. 82:1095-1096.
Kropp, P. J. 1967. Organic Photochemistry. 0. L. Chapman, Ed. Marcel Dekker,
New York.
Leighton, P. A. 1961. Photochemistry of Air Pollution Academic Press, New
York.
Lovas, F. J., and R. D. Suenraun. 1977. Identification of Dioxirane in Ozone-
Olefin Reactions via Microwave Spectroscopy. Chem. Phys. Letters. 51:453-456.
Martinez, R. I., R. E. Huie, and J. T. Herron. 1977. Mass Spectrometric De-
tection of Dioxirane and its Decomposition Products from the Reaction of
Ozone and Ethylene. Chem. Phys. Letters. 51:457-459.
McMillen, G. R., J. Kumari and D. L. Snyder. 1969. Chemical Reactions in the
Urban Atmosphere. C. S. Tuesday, Ed. Elsevier, New York.
Morrissey, R. J., and Schubert, C. C. 1963. The Reactions of Ozone with Propane
and Ethane. Combust. Flame 7 263-268.
78
-------
Nakagawa, T. W., L. J. Andrews, and R. M. Keefer. 1960. The Kinetics of
Ozonization of Polyalkylbenzenes. J. Amer. Chem. Soc. 82:269-276.
Niki, H., P. D. Maker, C. M. Savage, and L. P. Breitenbach. 1977. FTIR Spectro-
scopic Observation of Propylene Ozonide in the Gas Phase Reaction of Ozone-
.cis-2-Butene-Formaldehyde. Chem. Phys. Let. 46:327-330.
Niki, H., P. D. Maker, C. M. Savage, and L. P. Breitenbach. 1978. Mechanism
for Hydroxyl Radical Initiated Oxidation of Olefin-Nitric Oxide Mixtures
in Parts Per Million Concentrations. J. Phys. Chem. 82:135-138.
O'Neal, H. E. and C. Blumstein. 1973. A New Mechanism for Gas Phase Ozone-
Olefin Reactions. Intern. J. Chem. Kinet. 5:397-413.
Paszyc, S. 1974. Photochemistry of Nitro Alkanes. J. Photochem. 2:183-198.
Perkampus, H. H. et al. 1971. UV Atlas of Organic Compounds Vo; I-V. Plenum
Press, New York.
Perry, R. A., R. Atkinson, and J. N. Pitts, Jr. 1977. Kinetics and Mechanism
of the Gas Phase Reaction of OH Radicals with Aromatic Hydrocarbons Over
the Temperature Range 296-473K. J. Phys. Chem. 81:296-304.
Pitts, J. N., Jr., L. D. Hess, E. J. Baum, E. A. Schuck, J. K. S. Wan, P. A.
Leermakers, and G. F. Vesley. 1965. Recent Progress in Photobiology.
E. J. Bowen, Ed. Blackwells, Oxford.
Pitts, J. N., Jr. and B. J. Finlayson. 1975. Mechanisms of Photochemical
Air Pollution. Angew Chem. Intern. Ed. 14:1-74.
Pitts, J. N., Jr. A. M. Winer, K. R. Darnall, A. C. Lloyd, and G. J. Doyle.
1977. Hydrocarbon Reactivity and the Role of Hydrocarbons, Oxides of
Nitrogen, and Aged Smog in the Production of Photochemical Oxidants.
International Conference on Photochemical Oxidant Pollution and Its Control.
Proceedings: Vol. II p. 687-704.
Porter, G. B. 1960. Photooxidation of Biacetyl. J. Chem. Phys. 32:1587-1588.
Rebbert, R. E. 1963. Primary Processes in the Photolysis of Ethyl Nitrate.
J. Phys. Chem. 67:1923-1925.
Ripperton, L. A. and H. E. Jeffries. 1972. Formation of Aerosols by Reaction
of Ozone with Selected Hydrocarbons. Advances in Chemistry Series.
113:219-231.
Robbins, W. K., and R. H. Eastman. 1970. Photodecarbonylation in Solution. I.
Quantum Yields and Quenching Results with Dibenzyl Ketones. J. Amer. Chem.
Soc. 92:6076-6077.
Schere, K. L., K. L. Demerijian. 1977. Calculation of Selected Photolytic Rate
Constants over a Diurnal Range: A Computer Algorithm. EPA 600/4-77-015.
79
-------
Schubert, C. C., R. N. Pease. 1956. The Oxidation of Lower Paraffin Hydro-
carbons. I. Room Temperature Reaction of Methane, Propane, n-butane, and
Isobutane with Ozonized Oxygen. J. Am. Chem. Soc. 78:2044-2048.
Scott, W. E., E. R. Stephens, P. L. Hanst, and R. C. Doerr. 1957. Further
Developments in the Chemistry of the Atmosphere. Proc. Amer. Petrol.
Inst. 37:171-183.
Shoma, R. K., and N. Kharasch. 1968. The Photolysis of lodoaromatic Compounds.
Angew, Chem. Intern. Ed. Eng. 7:36-44.
Srinivasan, R. 1963. Photochemistry of Cyclic Ketones. Adv. in Photochem.
Vol. I. W. A. Noyes, G. S. Hammond, and J. N. Pitts, Eds. Wiley-Inter-
sicence, New York.
Strachan, A. N., and F. E. Blacet. 1955. The Photolysis of Chloroacetone at
3130 L J. Amer. Chem. Soc. 77:5234-5257.
Taylor, R. P., and F. E. Blacet. 1956. The Photolysis of Chloroketones. J.
Amer. Chem. Soc. 78:706-707.
Turro, N. J. 1978. Modern Molecular Photochemistry. Benjamin/Cummings.
Publishing Co., Inc. Menlo Park, CA.
Urry, W. H., and D. J. Trecker. 1962. Photochemical Reactions of 1,2-Diketones.
J. Amer. Chem. Soc. 84:118-120.
Urry, W. H., D. J. Trecker, and D. A. Winey. 1962. Photochemistry of 1,2-
diketones-l,2-cyclo-decanedione and 2,3-pentanedione. Tet. Lett. 609-613.
Vrbaski, T. and R. J. Cvetanovic. 1960. A Study of the Products of the Re-
actions of Ozone with Olefins in the Vapor Phase as Determined by Gas-
Liquid Chromatography. Can. J. Chem. 38:1063-1069.
Wagner, P. J. 1971. Type II Photoelimination and Photocyclization of Ketones.
Accounts Chem. Res. 4:168-177.
Wadt, W. R. and W. A. Goddard III. 1975. The Electronic Structure of the Criegee
Intermediate. Ramification for the Mechanism of Ozonolysis. J. Amer. Chem.
Soc. 97:3004-3021.
Walter, T. A., J. J. Bufalini, and B. W. Gay, Jr. 1977. Mechanism for Olefin-
Ozone Reactions. Env. Sci. Technol. 11:382-386.
Wiebe, H. A., and J. Heicklen. 1973. Photolysis of Methyl Nitrite. J. Amer.
Chem. Soc. 95:1-13.
Woodward, R. B., and R. Hoffman. 1970. The Conservation of Orbital Symmetry.
Verlay Chemie, Weinheim.
Williamson, D. L. and R. J. Cvetanovic. 1968. Rates of Reactions of Ozone
with Chlorinated and Conjugated Olefins. J. Amer. Chem. Soc. 90:4248-4252.
80
-------
TECHNICAL REPORT DATA
(I'lcasc read Instructions on the reverse before completing)
W-W/T2-79-001
2.
4. TITLE AND SUBTITLE
Atmospheric Reaction Products of Organic Compounds
Final Report
3. RECIPIENT'S ACCESSION-NO.
5.
6. PERFORMING ORGANIZATION CODE
7. AUTHOR!*'
Dale G. Hendry and Richard A. Kenley
8. PERFORMING ORGANIZATION REPORT NO.
PYU-8395
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Chemical Control
Office of Toxic Substances
U.S. Environmental Protection Agency
Washington, D.C. 70460
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-01-5123
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Chemical Control
Office of Toxic Substances
U.S. Environmental Protection Agency
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final Technical Reoort
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
the environment.
In this procedure, the relative importance of the three dominant reaction pathways-
photolysis, reaction with OH radical, and reaction with ozone—are first determined
for each compound. Then the products from each pathway that play a major role for
that compound are estimated using the techniques outlined in the methods.
The methods are applicable to a wide variety of compounds; however, as the structure
of the compounds differ from the structures on which the procedures were based, the
conclusions become tentative. It is believed, however, that this will be a minor
fraction of the cases to which it is applied.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Ticld/Group
18. DISTRIBUTION STATEMENT
Unlimited distribution through
National Technical Information Service
Sorinnfield, Va. 22151
19. SECURITY CLASS (Tllil Report)
21. NO. OF PAGES
20. SECURITY CLASS (Tills page)
22. PRICE
EPA Form 2220-1 (9-73)
81
------- |